The Effects of Firm Maturity: IPO and Post-IPO Performance, Growth, Efficiency, Profitability and Returns;
& The Rational Part of Momentum
Jorge Alberto Murillo Garza
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy
under the Executive Committee of the Graduate School of Arts and Sciences
COLUMBIA UNIVERSITY
2008
UMI Number: 3305253
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI UMI Microform 3305253
Copyright 2008 by ProQuest LLC.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC 789 E. Eisenhower Parkway
PO Box 1346 Ann Arbor, Ml 48106-1346
©2008 Jorge Alberto Murillo Garza
All Rights Reserved
ABSTRACT
The Effects of Firm Maturity: IPO and Post-IPO Performance,
Growth, Efficiency, Profitability and Returns;
& The Rational Part of Momentum
Jorge Alberto Murillo Garza
This thesis analyses the relation of firm maturity (age) and firm's performance at
their IPO and Post DPO returns and fundamentals. The first chapter analyses post-issuance
performance utilizing a sample of 9,400 IPOs spread from 1935 to 2002 and shows that
young firms (under 9 years old) are the most likely to underperform and be delisted; three
and five year cumulative abnormal returns range between -20% and -75% for this age
group. Between 8% and 18% of young firms delist before reaching their third IPO
anniversary, in contrast only 2% of old and mature firms delist. The increasing number of
young firms listed during the 80's and 90's, both on the Nasdaq and NYSE, accounts for
the strong underperformance during that period. Given the small supply of young-small
IPOs it is very plausible that investors seeking "the next Big Thing" overprice these
"lottery like" securities and underestimate their failure probabilities, resulting in overall
underperformance from this group.
The second chapter establishes an important link between industrial economics
and finance by exploring the effects of a firm's age on realized returns and firm
fundamentals. Over four decades, Mature firms generated an excess return of 20 to 30
basis points after controlling for industry, size and book-to-market characteristics. A
simple zero cost strategy that takes long positions in mature firms and short positions in
young and old firms yields annualized returns of 5.38% and 5.60% or between 7.33% and
3.71% in excess to size and value portfolios. We decompose age, listing cohort, and year
effects to analyze key firm fundamental characteristics related to growth, innovation,
efficiency, liquidity, default risk, and profitability. Maturity decreases firms' default
probability, earnings uncertainty, market illiquidity, and shortens the investment horizon.
While innovation and growth opportunities decrease with time, profitability, dividend
yield, and process efficiency increase with firm age. Firms in their mature stage enjoy
growth, profitability, lower risk and offer higher returns.
Finally, the third chapter presents arguments supporting rationality in the well
known momentum effect. We find that the returns of different momentum deciles closely
track a measure the rate of change in fundamental value calculated from analysts'
earnings estimates. We also find that while past changes in fundamental value predict
future stock returns, stock rates of return appear to predict subsequent changes in
fundamental value, up to a year in the future. The ability of past returns to predict both
future returns and future changes in fundamental value is consistent with heterogeneous
expectations models of capital market equilibrium, where the expectations of informed
investors create the apparent predictive ability of past returns. Since heterogeneous
expectations models are consistent with rational behavior, if there is a significant
irrational component to momentum, it is likely to deal with biases in the way analysts and
investors form estimates of earnings and fundamental value.
Table of Contents
Chapter 1 1
1.1 Introduction 1
1.2 Data 7
1.3 Methodology 13
1.4 Results 15
1.4. A Event-Time Returns 15
1.4.B Calendar-Time Returns 22
1.4.C Regression Results 26
1.5 Conclusions 27
1.6 References 53
Chapter 2 56
2.1 Introduction 56
2.2 Firm Age Models and the Relevance of Maturity 58
2.3 Data 62
2.3.A Measuring Firm Age 62
2.3.B Sample Selection 63
2.3.C Fundamental Characteristics 64
2.4 Firm Maturity and the Cross-Section of Returns 65
2.4.A The Firm Maturity Spread 65
2.4.B Time Series of Returns 70
2.5 What is Age? 72
2.5.A Firm Age and Growth 74
i
2.5.B Firm Age and Innovation Perspectives 77
2.5.C Firm Age and Process Efficiency 79
2.5.D Liquidity and Cash Flow Risk 81
2.5.E Default and Debt Structure 85
2.5.F Profitability and Cash Flow Distribution 88
2.6 Firm Age and Investment Opportunities 89
2.7 Conclusions 95
2.8 References 128
Appendix 2.A 131
Appendix 2.B 132
Chapter 3 133
3.1 Introduction 134
3.2 Representing Changes in Fundamental Value with Analyst's Estimates 141
3.3 Data and Variable Construction 145
3.4 An Initial Look at the Evidence 146
3.5 The Time Path of Momentum Deciles 148
1.5.A Returns 148
1.5.B Fundamental Value by Momentum Deciles 152
3.6 Ranking The Time Path of Momentum Deciles 155
3.7 Do Prices Predict Fundamentals or Do Analysts Chase Prices 158
3.8 The Prediction Horizon of Informed Investors 161
3.9 Conclusions 164
3.10 References 177
n
Tables and Graphs
Chapter 1
Table 1.1 Sample of Initial Public Offerings 30
Figure 1.1 Age Behavior Across Time 31
Table 1.2 IPO Characteristics by Firm Age 32
Table 1.3 IPO Characteristics by Industry 34
Table 1.4 EPO Characteristics by Size and Value 36
Table 1.5 First Day of Trading Returns and Survival 37
Table 1.6 Cumulative Abnormal Returns by Age Group 39
Table 1.7 Cumulative Excess Buy and Hold Returns by Age Group 44
Table 1.8 Wealth Relatives by Age Group 46
Figure 1.2 Three Year Cumulative Abnormal Returns by Age Group 47
Figure 1.3 Three Year Buy and Hold Abnormal Returns by Age Group 47
Figure 1.4 Three Year Wealth Relatives by Age Group 47
Table 1.9 Three Year Holding Period Return Distribution 48
Figure 1.5 Monthly Return Distribution by Age Group 48
Table 1.10 Distribution of IPOs per Year Cohorts and Listing Exchange Market.. .49
Table 1.11 Calendar Portfolio Returns for IPOs by Age Group 50
Table 1.12 Regression Coefficients for IPOs Three Year Returns 52
Chapter 2
Figure 2.1 Annualized Firm Age Portfolio's Returns 97
Figure 2.2 Annualized Firm Age Portfolio's Standard Deviation 97
Figure 2.3 Firms Book to Market Ratio by Firm Age 97
iii
Figure 2.4 Firms Sample as a Percentage of Total CRSP Firms 97
Table 2.1 Firm Sample and Summary Statistics 98
Table 2.2 Expected Monthly Returns by Age Group and Industry 99
Figure 2.5 Firm Age Decile Portfolios Returns 103
Figure 2.6 Firm Age Returns in Excess of Size and Book to Market 103
Figure 2.7 Firm Age Returns in Excess to Industry 103
Table 2.3 Time Series of Portfolio Returns and Fama French Factors 104
Figure 2.8 Mature Firms Alpha Estimate to Yearly Regressions 105
Table 2.4 Firm Age and Growth Rates 106
Table 2.5 Firm Age and Innovative Edge 109
Table 2.6 Firm Age and Process Efficiency I l l
Table 2.7 Liquidity and Cash Flow Risk 114
Table 2.8 Default and Debt Structure 118
Table 2.9 Firm Age and Profitability 121
Table 2.10 Industry Concentration and Firm Age 123
Table 2.11 Capital Invested Turnover and Firm Age 125
Table 2.12 "Value Creation" and Firm Age 126
Table 2.13 New Duration Estimates 127
Chapter 3
Table 3.1 Sample Size and Distribution 167
Table 3.2 Returns and Change in Value, Cross-Sectional Evidence 169
Figure 3.1 Hypothetical Returns for Momentum Deciles 171
Figure 3.2 Returns for Momentum Deciles 171
iv
Figure 3.3 Change in Value for Momentum Deciles 171
Table 3.3 Returns and Change in Value for Momentum Portfolios 172
Table 3.4 Returns and Change in Value for Past Value Portfolios 173
Figure 3.4 Change in Value for Past Value Deciles 174
Figure 3.5 Returns for Past Value Deciles 174
Table 3.5 How Fundamental Value Evolves When Momentum Works 175
v
Acknowledgements
The author is very grateful to Wei Jiang, Andrew Ang, Bruce Greenwald, James Scott,
and Tano Santos for all of their support, advice and mentorship. Also, the author
appreciates being a researcher at the Heilbrunn Center for Graham & Dodd Investing
working with Bruce Greenwald and James Scott. The author values comments from
Daniel Paravisini, Francisco Perez, and seminar participants at Columbia Business
School Ph.D. Seminars. Finally, the author thanks Gustavo Grullon and James Weston
from Rice University for initial discussions on this topic and sharing an initial compiled
age dataset.
vi
To my loving wife, Christina,
My always encouraging parents, Jose Antonio and Maria de los Angeles,
My dear parents-in-law, Bob and Jean,
My Brothers, Jose Antonio and Eric,
And all of my beloved family members.
vn
1
Chapter 1: Firm Maturity and IPO Underperformance
1.1 Introduction
The long run underperformance of Initial Public Offerings (IPOs) documented by
Ritter (1991) and Loughran and Ritter (1995) suggests that investors may systematically
be too optimistic and posses challenges to market efficient hypothesis. This paper
addresses two primary issues related to IPO underperfomance. First, the paper analyses
IPO performance in terms of firm maturity (age) at the time of the firm's IPO, if
investors are systematically overoptimistic firm maturity should not matter in the
observed performance. We find that IPO post-issuance long-run (three to five years)
underperformance comes primarily from young firms (less than 9 years old), while old
firms (more than 40 years) show no underperformance. Furthermore, young firms have a
higher probability of being delisted before their third year anniversary.
Second, the paper examines the pre and post 1970 IPO performance and sheds
light into the reason for the unobserved and observed underperformance pre- and post-
1970 as shown by Gompers and Lerner (2003). Prior to 1970, the percentage of young
firms listed was small. Post 1970, with the introduction of the Nasdaq, the percentage of
young firms listed grew dramatically to account for 50% of public offerings. An
expansion of firms with low short-term profitability and high growth expectations would
increase the number of failures and overall IPO underperformance. Hence, the
proliferation of young firms going public during the 70's, 80's and 90's (which more than
doubles every decade) caused lower survival rates and more underperformance than in
previous decades.
2
This paper shows that firm age is related to both size and book-to-market ratio,
and that age monotonically increases as size and book-to-market increase. If investors are
indeed overoptimistic, most likely they would be towards young firms which offer
potential high growth rates, positively skewed returns, and could be the "next big thing."
However, it is important to consider that young firms also will take longer to show
profitability, have high uncertainty on future cash flows, and high probability of being
delisted in their early years.1 This paper identifies the group of firms (young) which are
the main source of IPO underperformance. To my knowledge this is the first paper that
connects firm maturity to IPO performance using such a large dataset covering 9,400
IPOs from 1935 to 2002.
Proponents of efficient markets argue that once an IPO is publicly traded it is the
same as any other stock and, thus, aftermarket stock price should appropriately reflect the
share's intrinsic value. Consequently, risk-adjusted post-IPO stock price performance
should not be predictable. Miller (1977) proposes that investors have heterogeneous
expectations regarding the valuation of a firm, however given constraints on shorting
IPOs at the time of offering, only the most optimistic investors will buy the IPO; over
time, as the variance of opinions decreases, investors' valuations will converge towards
the mean valuation and its price falls. This is consistent with the drop in share price at the
end of the lockup period. An important question is whether IPO underperformance occurs
because of institutional constraints, such as short sale restrictions in the IPO market
which prevent the expectations of pessimistic investors from being reflected in the
offering price; or because of systematic over-optimism on behalf of investors.
The paper presented in chapter 2 decomposes firm age effects, listing cohort, and year effects for several fundamental characteristics associated with a firm.
3
Barberis and Huang (2007) propose that under cumulative prospect theory
positively skewed securities in small supply can be "overpriced" and therefore earn
negative average excess returns. It appears that investors could think of these stocks as
"lotteries," a heterogeneous holdings equilibrium can be supported as long as they are in
small supply and have sufficient positive skewness. We find young and small firm IPOs
are indeed in small supply relative to the market and that these firms tend to have higher
skewness than older and/or bigger firms. It is very plausible that investors are overpricing
young-small IPOs thinking that these might be the next "Ebay" (which quintupled its
market value three years post IPO) while underweighting their failure probability.
Rajan and Servaes (1997) show that IPOs have better long-run performance when
analysts ascribe low growth potential rather than high growth. They study earnings
growth forecasts and find that within six months of the IPO analysts predict that these
firms will grow approximately six percentage points faster than their industry peers.
Long-run growth predictions decline substantially over the following months, suggesting
that analysts eventually realize the predicted growth cannot be attained.
Because young firms have very short histories, investors may find it more
difficult to asses their growth potential and current value contributing to potential over-
optimism. Pastor and Veronessi (2002) argue that young firms have higher uncertainty of
future profitability and growth perspectives than older firms, and as firm age increases
this uncertainty decreases due to a learning process. However, it seems there may be
more than learning involved, young firms are innovation driven and hence face a high
probability of failure due to barriers of entry and volatility in the demand for their "new"
4
products. For the average investor, it is difficult to differentiate the next 'Yahoo' from
other young firms that might perish soon after their IPO.
Several other papers have proposed different explanations of why the observed
underperformance occurs. Teoh, Welch, and Wong (1998) propose that firms are eager to
'look good' when they conduct their IPO, and the market has difficulties in disentangling
carefully hidden warning signals. This would suggest that at least part of the poor long-
run performance is due to a market that is unduly optimistic and unable to properly
forecast tougher times ahead. This argument would point towards overconfidence by both
entrepreneurs (Bernardo and Welch, 2001), and investors (Daniel, Hirshleifer, and
Subrahmanyam, 1998). However, as mentioned earlier, investors might be more
optimistic towards younger firms (which could potentially give misleading information)
rather than well established mature firms which have a long track record.
In addition, Mikkelson, Partch, and Shah (1997) show that poor long-run return
performance is also accompanied by poor financial accounting performance post-IPO
relative to pre-IPO performance and industry conditions. Similarly Jain & Kini (1994)
show a significant decline in operating performance subsequent to the IPO, which is
inconsistent with the fact that EPOs are initially priced at high price-earnings (P/E)
multiples, implying high future growth expectations on part of the investors. IPO firms
start out with high market to book and P/E ratios relative to their industry counterparts
and experience a decline in these measurements post-IPO. In general, results suggest that
investors appear to value firms going public based on the expectations that earnings
growth will continue. However, in actuality, the pre-IPO profit margins, on which
expectations are formed, may not be sustained. This paper proposes that these
5
expectations may be higher and more uncertain for younger firms than for mature firms.
Additionally, some of the poor performance could be attributed to optimistic accounting
early on in the life of the firm.
Brav and Gompers (1997) suggest that venture backed IPOs outperform non-
venture backed IPOs when portfolios are equally weighted and that value weighting
significantly reduces performance differences and substantially reduces under-
performance for non-venture IPOs. They also show that IPO underperformance is a
characteristic of small and low book-to-market firms regardless of whether they are IPO
firms or not. In addition, Brav and Gompers, propose that VCs may be able to entice
more and higher quality analysts to follow their firms, thus, lowering potential of
asymmetric information between the firm and investors, implying that markets may not
discount the shares of non-venture backed companies enough. Consistent with Brav and
Gompers, this paper shows that both size and B-M characteristics are related to EPO
underperformance; however, we show that age is key characteristic that explains IPO
underperformance above size and book-to-market. Since young-small IPOs represent a
very small percentage of the overall IPO market value (small supply) it is not surprising
that by value weighting performance differences would be mitigated. It would be
interesting to analyze whether having a VC helps differentiate young IPO winners from
losers by lowering the information asymmetry between investors and firms, and whether
VCs might let a firm mature longer before taking it public.
On the other hand, Gompers and Lerner (2003) convincingly show that in an
earlier sample from 1935 to 1972, IPOs do not underperform benchmarks on the
aggregate, arguing that IPO underperformance observed in the last three decades may be
6
just a small sample effect. Thus, there may be no IPO underperformance ex-ante, but
during the post-1970 period we may have drawn a small sample where too many IPOs
perform very poorly ex-post. In contrast we show that post 1970 underperformance is
attributed to failure of young firms. Schultz (2002) presents arguments that point towards
market timing, where we would observe that more IPOs follow successful IPOs and
hence the last large group of IPOs would under-perform and be a relatively large fraction
of the sample. However, Ang, Gu, and Hochberg (2005) develop a model to test whether
IPO underperformance may result from observing too few star performers ex-post rather
than expected ex-ante. After allowing for returns to be drawn from mixtures of
outstanding, benchmark, or poor performing states, they find that small sample biases are
extremely unlikely to account for the magnitude of the post-1970 IPO underperformance
observed in data. The evidence in this paper suggests that the post-1970
underperformance is due to the large quantity of young firms listed in the Nasdaq. Young
firms listed on Nasdaq account for 43% of the sample of IPOs or 4,070 firms. When
Nasdaq was introduced, the volume of EPO's (particularly young firm IPOs) increased
dramatically, allowing for weaker firms to list on a more lenient exchange which may
have made it harder to differentiate good firms from bad ones, increasing the "lottery"
behavior of these group.
The rest of this paper is organized in the following way: Section 1.2 presents the
data and its characteristics. Section 1.3 describes the methodologies used to measure IPO
performance. Section 1.4 presents results. Section 1.5 concludes the chapter.
7
1.2 Data
The main focus of this chapter is to introduce firm maturity as a key characteristic
to explain the firm's post IPO performance. The sample is comprised of 9,400 initial
public offerings from 1935 to 2002. To be included in the sample, an IPO firm must have
an offer price greater than one dollar and must be subsequently listed on the Center for
Research in Securities Prices (CRSP) tapes within six months of the offering date. In line
with previous literature, all unit offerings, REITs, ADRs, limited partnerships, and public
offerings of close-end funds are excluded from the sample. IPOs are identified using two
methods: (1) IPOs that occurred between 1970 and 2002 are identified using the
Securities Data Corporation (SDC) Global New Issues database. (2) IPOs that occurred
before 1970 are identified following the data sources and methodology described in
Gompers and Lerner (2003).
Firm age, the key piece of information, is computed as the difference between the
year of IPO and the earliest of either its founding or incorporation date. The founding
dates and/or incorporation dates are obtained from the compilation of two datasets. (1)
The extended Jovanovich and Rousseau (2001) dataset. This dataset was extended by
Fink, Fink, Grullon and Weston (2005); it contains the date of first incorporation and/or
founding date for a sample of publicly traded firms between 1925 and 2005. (2) The
Field-Ritter dataset of company founding dates, used in Field and Karpoff (2002) and
Loughran and Ritter (2004). This dataset contains the founding dates for 8,309 firms that
went public in the U.S. during 1975-2005. All where the founding dates were different
the data was verified by hand; we did 280 additions and about 350 corrections.
8
The resulting dataset was then extended by hand using various sources. Kelley's
"Business Founding Date Directory" was used to identify extremely old firms that were
founded before 1915. The data sets mentioned in the previous paragraph contained
several firms that were classified as having between zero and three years of age at the
time of listing. Some of these firms (especially those with very high asset values or sale
levels) seemed unlikely to be "new" firms. Therefore, these dates were verified by hand
and in most cases reclassified. The "International Directory of Company Histories" was
used to obtain these additional founding dates, and also to assign the correct founding
dates for spin-offs, rollups, and reverse LBOs, which in most instances had received the
date of issuance as its founding date. Several dates were also verified through electronic
databases and the firm's own webpage. It was difficult to determine the founding date
for some of the reclassified firms. In those cases a judgement about the most accurate
date was based on the whole history of the firm, its subsidiaries and/or divisions that went
public.
In order to explore the effect of firm age on IPOs aftermarket performance in
detail, I have designated twelve age groups based on the age of the firm at the time of the
IPO: 1) zero years; 2) one year; 3) two to three years; 4) four to five years; 5) six to eight
years; 6) nine to twelve years; 7) thirteen to eighteen years; 8) nineteen to twenty seven
years; 9) twenty eight to forty years; 10) forty one to fifty five years; 11) fifty six to
seventy five years; and, 12) firms greater than 75 years. Given that age is computed as the
difference between two years, a firm can be classified as a one year firm even though it
might be some months older or younger than this age; for example if the firm is
2 Business and Company Resource Center, Business Insights, D&B Million Dollar Database, Hoover's
Company Capsules & Profiles, and Thomson ONE Banker.
9
incorporated in January of 1990 and listed on December of 1991 it will be classified as a
one year old firm, however, effectively it would be almost two years old. These
differences in age should average out to the corresponding age group given the size of the
sample. Particular attention should be placed on the zero age group, many of these firms
(especially those listed on the NYSE) might not be as young as presumed. Even though
these firms were revised by hand it was not possible to find more information that would
change their current classification, but it is very likely that these firms are either spin
offs, roll-ups or consolidating firms. This group was included in the sample for
completeness, but all conclusions will be derived from the other eleven groups.
Table 1.1 presents the IPO sample distribution over time and includes number of
IPOs per year, median and average firm age. The sample of IPOs is mostly concentrated
in the mid 80 's and throughout the 90's. It can also be observed both in Table 1.1 and
Figure 1.1 that firm age decreases dramatically after 1970, both on the mean and the
median. Previous to 1970, average age fluctuated between 40 and 50 years old. Post-
1970, firm age dropped significantly below 30 years, reaching a 10 year low in 1999.
Moreover, table 1.1 presents the average, median, 25th percentile, and 75th
percentile rankings of market value and the ratio of book to market equity value at the
close of the first day of trading for each year. The table shows that market value varies
across time, in general the average market value is high during the period of 1945 - 1972,
then drops considerably during the 70's and 80 's, and not surprisingly rises during the
late 90's. It is important to note the behavior of the book to market ratio both on the
median and the average, which previous to 1956 is usually above 1.0, declines to 0.51 in
10
1968, then bounces back to ratios above 1.0 in the late 70's and decreases consistently
during the 80's and 90's to hit a minimum of 0.5 in 2000.
Market value is computed using the closing price of the first day of trading
multiplied by the number of shares outstanding given by the CRSP tapes. In order to be
able to compare between years, market value is normalized to 2002 dollars, adjusting by
the Consumer Price Index. Since at the date of the IPO book value has not yet been
reported for any firms, the book to market ratio is set to the average of the three digit SIC
industry book to market ratios. Book value is computed using Compustat North America
data tapes, and is defined as the sum of shareholders equity (data60), balance sheet
deferred taxes (data74), and investment tax credit (data51). In addition to the book value
computed with Compustat, the dataset is enhanced by including the hand-collected book
equity values from the Moody's Industrial, Public Utility, Transportation, and Bank and
Finance Manuals provided by Davis, Fama and French (2000). The book to market ratio
is computed using concurrent book value and market values.
Table 1.2, panels A and B, present the sample distribution, average, median, 25th
percentile, and 75th percentile rankings of market value and the ratio of book to market
equity value at the close of the first day of trading for each of the twelve age categories.
Table 1.2, panels C and D present statistics on the sample one year after the first day of
trading. The book to market ratio of each firm is updated as information from Compustat
becomes available. Hence, the descriptive statistics presented in panel D correspond
mostly to the book to market ratio of the firm and not the three digit SIC industry average
of book to market ratios. It can be noted that 60% of sample IPOs can be classified
between 2 and 18 years old and that market value increases as firm age increases,
11
implying that older firms are also bigger firms. At the same time, book to market ratios
increase as firm age increases. The average book to market in panel D revolves around
0.5 for firms below 18 years and climbs to 1.01 for firms above 75 years old, implying
that young firms behave as growth firms and old firms as value ones.
Table 1.3, panels A, B, and C, present the sample distribution, average, median,
25th percentile, and 75th percentile and rankings on the mean of firm age, the median for
market value, and the ratio of book to market equity value at the close of the first day of
trading for each of the twelve industry group categories defined in Fama and French
(2002). Table 1.3, panel D, presents statistics on the book to market ratio one year after
the first day of trading.
In panel A, it can be observed that 73% of the IPO sample is comprised of firms
in the following sectors: Business Equipment, Shops, Finance, Health, and Other.
Business equipment, firms which make up almost 24% of the sample, are ranked as the
second youngest industrial sector with an average age of 10.89 years; health firms rank
first and telecommunication rank third with average ages of 9.61 and 11.48 years
respectively. Not surprisingly, these industry sectors are considered to be high growth
and high risk ones, with potential high outperformers as wells as several firms that will
perish. On the other hand, utility firms which only make for 1.36% of the sample are
ranked as the oldest sector with an average age of 36.36 years; utility firms are followed
by finance, non-durables, and manufacturing industry sectors with average ages of 34.84,
32.04, and 31.19 years respectively.
In panel B, it can be observed that median rankings of market value realized after
the first day of trading for finance, durables, shops, and health sectors receive the lowest
12
ranking or market value; the sectors of energy, chemicals, telecommunications and
utilities receive the highest rankings. In panel C, it can be noted that median rankings of
book to market value ratios realized after the first day of trading for health,
telecommunications, and business equipment industries can all be considered growth
sectors; meanwhile the industry sectors of manufacturing, utilities, and finance are
classified as value sectors. This classification is maintained one year after the first day of
trading as it is observed in panel D.
Table 1.4 presents the sample distribution and firm age median and average based
on the Fama and French (1992) quintile classification on size and book to market ratio.
Panel A classifies each IPO based on the information available on the first day of trading;
most of the sample is classified on the second, third and fourth book to market quintiles.
Panel B classifies IPOs based on information available one year after the first day of
trading. This is a more representative descriptor of firms characteristics, the majority of
firms are classified in the upper left triangle (small growth firms). Furthermore, it should
be noted that in both panels A and B it is observed that age increases monotonically as
size increases well as book to market ratio increases, which classifies old firms on
average as big value firms and young firms as small growth firms. When Brav and
Gompers noted that underperformance is a characteristic of small, low B-M firms, they
did not consider that a key characteristic of these firms is that they are extremely young
and hence have higher uncertainty and probability of failure. We will address this point in
detail in Section 1.4.
13
1.3 Methodology
Following standard practice, the long-run performance over t months, particularly
3 and 5 years (36 and 60 months) of initial public offerings is measured by constructing
benchmark-adjusted returns for stock /' relative to benchmark m in month t as the
difference between raw returns on the IPO and benchmark returns for the corresponding
period, rt t (m) = Rit -Rmt where R{, is the raw return of firm i in event month t and Rmt
is the benchmark return in event month t. Returns are computed from the first listing on
the CRSP daily return tapes. Event months are defined as successive 21 trading-day
periods relative to the IPO date. The first day of trading (month 0) is set apart and is not
included in the analysis. Thus, returns for the first month comprise of returns listed on
days 2 to 22, the second month listed on days 23 to 43, and so on. Because several of the
EPOs are delisted before reaching their 60th month of trading, it is important to adjust for
the delisted return. Furthermore, it will be shown that the probability of being delisted,
which is higher for young firms, plays a key role in the observed underperfomance.
Delisted EPOs are assigned the delisting return provided by CRSP event tapes. This
paper presents results using six different value weighted benchmarks:4 1) matching Fama
and French quintile portfolios for size and book-to-market; 2) matching industry
portfolios using Fama and French twelve industry classification; 3) all stocks traded on
the NYSE, Amex, and Nasdaq; 4) stocks traded on the Nasdaq; 5) NYSE small stocks;
and 6) S&P500.
In addition to the tests presented in this paper using CRSP event tapes delisting return, tests were also run using a delisting return of-0.3 for all non-Nasdaq firms and -0.55 for Nasdaq firms as suggested by Shumway and Warther (1999). These results were very similar and consistent with the presented results.
Results for equal weighted benchmark portfolios were also computed. These results were stronger than the ones presented of value weighted portfolios; in that sense, these results are more conservative.
14
Following Ritter (1991), cumulative average adjusted m benchmark excess returns
(CAR) to event month horizon s for each age group are defined as:
5 nage,t
CARages (m) = j ] ARagel (w) where ARagel (m) = •£- £ rit (m) and naget is the number of
stocks in the IPO age group portfolio in event month t. Thus, ARaget (m) is the average
benchmark-adjusted return across all IPO firms in the pertaining age group category in
event month t.
In addition to cumulative average adjusted returns, two other measures are
implemented: (1) Cumulative excess (abnormal) holding-period returns (CHP) and (2)
wealth relatives (WR) of stock i relative to benchmark m until the earlier of horizon event
month s or its delisting per age group: CHPis (m) = ]~|(1 + rit) - Y[ 0 + rm,t) where rit is
the raw return of stock i and rm t is benchmark m return as defined previously. This
represents the excess return of a zero-cost strategy that goes long on an IPO and shorts
the benchmark m portfolio. Cumulating these returns for each age group provides the
long-horizon return to this zero-cost strategy.
;'=1
For ease of comparison between cumulative excess holding periods across
different horizons, annualized CHP statistics are computed using the following
transformation: CHPaa™a,bed (m) = (1 + CHPages (m))u,s -1
Wealth relatives (WR) are computed using the month s holding period returns (HPis)
defined as:
15 age
1 + — YHR, "age,* i-t ' ' J
WRavp, = ^ age.s age
^ + — YHPms nage,s i-J m . s
1=1 5 5
where HPis =Y\(l + Ri,,)~l ^^ Hpm,s = E I 0 + R*,>)~*with ^.•,«(,n) b e i n 8 t h e r a w
r=l (=1
return on stock / in event month t and R, the benchmark return in event month t for the
same period and age group. HPis measures the total return from a buy and hold strategy
where a stock is purchased at the first closing market price after going public and held
until the earlier of horizon event month s or its delisting. Wealth relatives greater than
1.00 can be interpreted as IPOs outperforming the benchmark. A wealth relative less than
1.00 indicates that IPOs underperformed the benchmark.
1.4 Results
1.4.A Event-Time Returns
One key finding of this paper is that IPO overall underperformance is
concentrated around young firms, this section presents the analyses of IPO returns in
event time. We start by examining the first day of trading of an IPO and its survival
expectation given its age at the time of the IPO. Delisted firms are classified using
CRSP's delisting code (dlstcd) where 400-499 is assigned to firms that have been
liquidated and 500-599 is assigned to firms that are dropped from the CRSP lists for no
clear reason. Firms that are merged or acquired codes 200-299 are classified as such.
Table 1.5 shows that between 8% and 18% of the IPOs within the age range of 0 and 8
years (young firms) will be delisted by the third year of trading; while for IPOs over 40
16
years {old firms) delisting probability is only between 0.9% and 2.2%. By the fifth year,
between 15% and 31% of young firms will have been liquidated or dropped from the
exchange; in comparison, only 5% of old firms will have such fate. At the same time, the
probability of being acquired or merged is slightly higher for young firms than old ones.
It can be noted that, with the exception of the zero-age group, the probability of a firm
being delisted decreases monotonically as age increases.
It is no surprise that average EPO underperformance is deeply linked to all the
firms that are being delisted and that these firms extremely negative returns weight down
overall perfomance. However, before this study, it was hard to disentangle which firms
would have an ex-ante lower survival rate. Young firms have new products with high
growth potential but at the same time have to prove themselves in the market and break
barriers of entry in a competitive environment which lowers their survival probability.
This paper shows that young firms have a high probability of being liquidated and also
comprise a large percentage of the IPO sample (especially after 1970 when Nasdaq was
introduced). Therefore it is clear that a large source of the observed underperformance
can be attributed to young firms.
The high uncertainty on returns and survival related to young firms should be
reflected to an extent in the underwriters price discount of the offering, the IPO's first day
of trading return (jump) measured as the difference between the bidding price and the
closing price of the day can be used as a proxy for this discount. Panel C of Table 1.5
shows that young IPOs have an average jump of more than 5.5% on the first day while
old firms jump less than 2.7%. We would expect for young firms to have a larger
discount than old firms because of the uncertainty related to these firms. At the 75th
17
percentile young firm's discount is about 7.5% while old firms discount is about 2.7%.
However, it may be that this discount is not large enough for young firms given their risk
profile. It is worth emphasizing that this is not the total first day return that an investor
could attain by buying the shares at the offer price from the underwriter which tends to be
lower. The reason for using the bidding price is that for firms not recorded on SDG
(before 1970) the offer price is not available.
The three and five year cumulative abnormal returns computed for each of the
twelve age categories, as well as for all IPOs is presented in Table 1.6, panels A and B.
Delisted IPOs receive the delisting return as explained previously, and thereafter, the firm
and its benchmark are excluded from all computations. The left columns present the
CAR's for the six selected benchmarks as delineated previously, the right columns report
t-statistics computed following Ritter (1991).5 In general, by the third year firms have a
20% underperformance relative to the diverse benchmarks. However, once we observe
the different age categories, it is clear that the underperformance is attributed to younger
firms rather than older ones. Underperformance of IPOs relative to benchmarks
monotonically decreases from an average 50% for extremely young EPOs (zero and one
years old) to a mere 2% (statistically insignificant) for firms older than 56 years old. It is
also important to note that by the fifth year, old firms increase their underperformance up
to 20% in some benchmarks. Young IPOs remain the major losers underperforming up to
75%, but the firms between 13 and 40 years old {mature firms) seem to improve their
performance after surviving the third year. It would seem that once the mature EPOs are
able to establish themselves as good firms (having removed all bad EPOs) these firms
These t-statistics must be interpreted with care, since Barber and Lyon (1997) show that the small sample distributions for the CAR statistics are severely skewed compared to the Ritter (1991) asymptotic distributions.
18
offer better growth opportunities and performance than older firms which are not be able
to grow at the same pace nor outperform the benchmarks.
Since a high percentage of young firms suffer the fate of being delisted
(liquidated), panels C and D present the three and five year cumulative returns, excluding
firms that will be delisted. The goal is to test whether the out-performance of old firms
relative to young firms would hold for those firms that do not fail during the first years of
the IPO. By excluding delisted firms, the three year underpeformance relative to
benchmarks becomes insignificantly different than zero for all age groups, except the
zero-age firms. Five years after the IPO, mature firms that would be classified between
10 and 20 years old are outperforming the benchmarks. Meanwhile, the performance of
old firms is not strong.
In order to investigate the importance of firm age in IPO performance in contrast
to size and book-to-market characteristics as suggested by Brav and Gompers, we
compute cumulative abnormal returns sorting firms into seventy five groups based on
size, book-to-market & age. Panel E sorts firms into size and book-to-market quintiles
based on the information available on the first day of trading, as in Table 4, and
subsequently into three age groups6 and then computes the cumulative abnormal returns
relative to the NYSE, Amex & Nasdaq value weighted benchmark. The results show that
IPO underperformance is mostly attributed to firm age rather than just pure size and
book-to-market characteristics. All of the young firm's twenty five size & book-to-
market groups have a negative three performance ranging from -5.9% to -85.5% (sixteen
of the twenty five groups are statistically significant). In contrast mature and old firms
6 1) Finns between 0 and 8 years (young); 2) Firms between 9 and 40 years (mature); 3) Firms 41 and above (Old).
19
show mixed results of underperformance (some are positive and some negative) and with
a lower magnitude, only seven of the mature firms groups that show underperformance
are statistically significant and only three of the old firms groups have significant
underperformance.
Panel F classifies firms into Fama and French groups based on the information
available seven months after the first day of trading, thus we have a better classification
in the book-to-market ratio as we incorporate firm's book information rather than the
industries average as explained earlier in Section II. The results are consistent with Panel
E, most of the statistically significant underperformance is concentrated in young firm's
category (fifteen groups), particularly in the first three small and mid size groups with no
difference in the book-to-market ratios. These results seem to indicate that EPO
underperformance can be attributed to young-small firms rather than being a size and
book-to-market characteristics effect.
In Panel G we present total market value after the first day of trading adjusted to
2002 dollars for all IPOs within their size and book-to-market category. The panel also
presents the percentage that each group represents of overall IPO market value and the
average IPO market cap for each category. Young small & mid cap IPOs represent about
8% of the total EPO market value captured on the first day by the entire sample, while
they comprise 36% of the sample firms. Consistent with Barberis and Huang young firms
are positively skewed, offering potential high growth rates with a small probability of
extremely high returns, given their small supply it is very plausible that investors
overestimate their success probabilities and overprice these stocks. Hence it may not be
20
surprising to find that IPO underperformace is linked with the vast number of "lottery
like" young firms which impact is magnified in equal weighed portfolios.
Table 1.7, panels A and B, present three and five year annualized cumulative
excess holding-period returns for each age category. On average, IPOs underperform
relative to their benchmarks by an annual 4.5% and 3.5% on a three year and five year
buy and hold strategy, respectively. However, when underperformance per age group
over a three year period is closely examined, again, most is concentrated among young
firms ranging from -4.9% to -15.5% annual excess returns, depending on the benchmark
and age group that is taken. Buy and hold returns for old firms over a three year period
range from -1.2% to 3.8% (all of which are statistically insignificant), depending on the
benchmark that is observed; in effect, firms above 40 years show a zero
underperformance. For all benchmarks, underperformance decreases monotonically as we
increase the age of the firm. As was observed in the previous table, when considering a
five year holding period, mature firms start to show better performance as older firms'
performance declines; meanwhile, young firms keep showing strong underperformance,
particularly in the extreme young firms group. Figures 1.2, 1.3 and 1.4 graph the
cumulative abnormal returns, buy and hold, and wealth relatives for each age group in
relative to the benchmarks.
Table 1.8 presents additional evidence of underperformance for young firms vs.
old firms. This table shows the three and five year wealth relatives for each of the twelve
age categories. Young firms show a loss in investors' wealth between 20% and 35% over
the first three years and between 23% and 51% during a five year period. Clearly,
investors would have done much better by investing in the benchmarks. At the same time,
21
old firms range in a wealth loss of 7% up to a gain of 4%, depending on the benchmark
and age group during the three year period, and a loss between 13% and 2% on a five
year period. Again, investors' wealth relative measures increase monotonically as firm
age increases.
The previous results show that young firms will on average underperform. Such
underperformance is attributed mostly to the high percentage of young firms that delist
prior to their 5l year anniversary. In order for investors to invest in these firms that have
higher uncertainty and lower survival probability, these firms must also offer high growth
perspectives and returns when they are successful. As shown earlier, most of the young
firms are classified in technology intensive industrial sectors, such as health, where
innovation carries high rewards for the firms that are able to ensure themselves market
share. Table 1.9, presents the three year holding period return distribution (percentile
breakpoints) per age group. As expected, we observe that in the bottom 2nd and 5th
percentile points (the major losers) young firms have lower returns than older firms,
which decreases monotonically. In the 2 percentile, young firms have lost up to 98% of
their value, while old firms only present a loss between 76% and 90%. In the 5th
percentile, young firms have a loss of 93%, while old firms will have lost only between
63% and 76%. In the 40th percentile, young firms show a strong loss of 40% to 49%,
while old firms start showing a small out-performance. When we look at the top winners
(star IPOs) in the 95th and 98th percentiles, in Figure 1.6, we can observe that the returns
for young firms are much higher than for older firms. In the top 95th percentile, young
firms will have returns between 350% and 400%, while star old firms will have more
moderate returns around 240%. In the top 98th percentile of extreme success stories (such
22
as Yahoo, Microsoft, and Google), three year holding period returns for young firms is
between 584% and 737%. Older super star firms also show amazing returns between
300% and 395% but these returns are significantly lower than those offered by the supper
star young firms.
1.4.B Calendar-Time Returns
In addition to the previous event-time performance tests, this paper presents
calendar-time tests. Schultz (2003) argues that there is no evidence of IPO
underperformance in calendar-time. Consistent with the results presented by Ang, Gu,
and Hochberg, the results presented in this paper show that contrary to Schultz's claims,
IPO underperformance is also observed in calendar time. Furthermore, by classifying
IPOs according to the listing exchange, it is apparent that most of the underperformance
is attributed to the vast number of young firms listed in the Nasdaq.
The paper by Schultz focuses only on one-month holding-period returns. In order
to examine calendar-time returns, this paper looks at holding periods longer than one
month, by reporting 12, 36 and 60-month holding periods. In month t, IPO portfolios are
formed placing an equal amount of money in all IPOs that have gone public over the last
year. This portfolio will be held from time t to M-12, 36, or 60-months and will be
rebalanced to only hold IPO's that have gone public over the last year. Therefore, the
calendar-time IPO portfolio returns represent the returns on an equally-weighted portfolio
of IPO's, with each IPO occurring no later than one year. After computing the calendar-
time IPO raw returns, the NYSE, Amex, and Nasdaq value weighted returns are
subtracted to obtain the benchmark-adjusted holding-period returns in calendar time.
23
Table 1.10, panel A, presents the total number of IPOs realized during five year
periods across time, the sample is also broken into three main age groups: young (firms
under 9 years), mature (firms from 9 to 40 years), and old (firms above 40 years). Each
age group is subsequently broken into three groups depending upon which exchange
market lists each (NYSE, Amex, or Nasdaq). A very small fraction of firms are listed in
other exchanges. Panel B presents the same distribution as a percentage of IPOs realized
during the time period. The table shows how at least in this sample, all IPOs listed
previous to 1960 occurred in the NYSE. During these 25 years, old and mature firms
dominated the IPO market as a percentage of listed IPOs. However, in the 60s the
American Exchange became a key market in which to list; with more lenient rules than
the NYSE, the number of IPOs more than doubled, consisting mainly of young and
mature IPOs that listed on the Amex. The percentage of listing young firms increased
from an average of 10% previous to 1960 to 20%. Although mature firms still comprised
43%> of the listings, half of the listings occurred on the Amex.
Post-1971, when the Nasdaq was introduced, the number of IPOs in the Amex
decreased significantly as firms opted to list on the Nasdaq. Young firms' listing
percentage rose to become 50% of the listings by the end of the 70s. During the 80s, the
IPO market quadrupled as even more firms began to list in the Nasdaq: young firms made
up 52% of all listings, mature firms composed 30% to 36%, and old firms only accounted
for 12% to 17%. During the 90s (the dot-com era), the IPO market once again almost
doubled, averaging 2,350 IPOs every 5 year period: young IPOs made up between 52%
and 59% of all IPOs, mature firms were 34% to 38% of the IPOs listed, and old firms
shrunk to less than 10%. Post Bubble-Bust in 2001, the IPO market shrunk severely: the
24
percentage of old firms rose to 18%, and the percentage of firms listed on NYSE doubled
the previous decade percentages.
Calendar-time IPO portfolio returns observed over 12, 36 and 60-month holding
periods are reported in Table 1.11, Panel A. The portfolios are formed based on the firm's
age and the exchange in which it is listed. The left side presents the mean using
overlapping observations of annualized portfolios raw returns, the right side presents
annualized excess returns. The reported f-statistics are computed with Newey-West
(1987) standard errors, using a lag length of one less than the holding-period horizon.
Thus the statistic corrects for moving average errors induced by the overlapping
observations. Panel B presents portfolio returns for the time period from 1935 to 1970;
Panel C presents returns for the post-1971 period; Panel D presents returns for portfolios
formed from 1962 to 1970 when Amex was a very popular stock exchange.
Inspecting Panel A, it can be observed that across the entire time period, excess
returns are positive and it would seem that there is no underperformance relative to the
market. Nevertheless, it can be noted that, on average for all firms, returns increase as
firm age increases, and especially for firms listed on the NYSE; a less clear relation can
be established for firms listed on the Nasdaq and Amex, which will be explained in the
other three panels.
In Panel B, IPOs listed before 1971, in general all stocks listed on the NYSE
performed similarly regardless of their age group. However, stocks listed on the Amex
show outstanding performance, with average excess returns of 15.65%, 19.25%, and
15.88% for 12, 36, and 60-month holdings, respectively. Since IPOs listed on the Amex
occurred only after 1962, Panel D, compares NYSE and Amex over the exact same
25
period. It can clearly be seen that Amex IPOs outperform NYSE firms. Further,
investigation of the firms listed on the Amex over the same period, sheds light onto why
these firms outperformed others so impressively: during the 1964-1966 time period, some
young and mature firms backed with big government contracts were listed in the Amex
and as a result significantly outperformed the market given the special conditions of their
projects.
Panel C shows that, with the exception of the particular period of the 60s, post-
1971 when the Nasdaq is introduced, young firms underperform significantly and old
firms outperform the benchmark. Nasdaq & Amex listed firms perform better than NYSE
IPOs, during this period even NYSE mature firms show underperformance. However, its
important to keep in mind that NYSE stocks represent a very small percentage of the
post-1971 sample and particularly during the 70's and 80's there are very few young and
mature firms; hence, a few bad firms might be down-skewing results.
Calendar-timer results show that IPO underperformance is stronger in the post-
1971 era mainly because of two factors: 1) A dramatic explosion of young firm IPOs
which inherently carry a higher potential for growth and returns when successful; when
not successful, however, there is a higher probability of failure and delisting, and also a
tendency towards less reliable information being conveyed to investors. 2) The
introduction of the Nasdaq stock exchange allowed younger firms to be listed in a less
regulated exchange. Along with Brav and Gompers (1997), it would be interesting to
consider these young firms and analyze them based on being backed by VCs or not, it
may be that VC backed firms wait longer to perform the IPO and hence, are more
successful.
26
1.4. C Regression Results
The event-time and calendar-time patterns documented above are not necessarily
independent of each other. Table 1.12 reports the results of a multiple regression using
the raw three year total return on IPOs as the dependant variable. The explanatory
variables are the logarithm of one plus age, the three year total return on the market
(NYSE, Amex, and Nasdaq value weighted), the three year return on matching Fama and
French size and book-to-market quintile portfolio, the three year return on matching
Fama and French twelve industry portfolios, the logarithm of market value and matching
average three SIC industry book-to-market ratio after the first day of trading, dummy
variables to identify the three main stock exchanges, dummy variables to identify each of
the twelve industry categories, and dummy variables to identify five year time periods.
These results support the conclusions from earlier tables. The intercept shows an
underperformance of 40% to 60%. The first regression, using only age and the market as
explanatory variables, shows IPOs underperform by 43%; however, firm age significantly
improves performance (12.43% times each logarithmic age year) which implies that firms
older than 32 years would show a zero underperformance. The other regressions show
that firm age improves IPO performance regardless of the firm's size and book-to-market
characteristics. Firm size only marginally increases the after market performance,
however, the firm book-to-market ratio does improve performance significantly (about
10% times the B-M ratio). By looking at the exchange dummy variables we note that
NYSE and Nasdaq underperformance was very similar (about 60% on average), but
Amex shows less underperformance (about 42%).
27
Additionally, the industry dummy variables revealed that in most cases the
industry was not relevant to the IPOs performance, except for business equipment, health,
and finance sectors which showed a significant better performance. Finally, the five year
time period dummy variables show that with the exception of the time period 1976 to
1980, in general the year when the IPO occurred did not contribute significantly to its
performance. It is worth noting that IPOs realized during the height of the dot com era
(1996 to 2000) decreased in performance by 28%. Those realized after the bust increased
in performance by 23%. In all the regressions, age remains an ex-ante known variable
that significantly improves the ex-post IPO performance.
1.5 Conclusions
Beginning with Ritter (1991) researchers have found that on average IPO's appear
to be overpriced and tend to underperform relative to multiple benchmarks subsequent to
their issuance for three and five year horizons. This paper analyzes IPO post-issuance
performance in terms of firm maturity {age) at the time of IPO. Expected IPO
performance has never before been thought of in terms of firm maturity; to my
knowledge this is the first time that such link is established. The data suggests that most
of the IPO underperformance can be primarily attributed to young firms (less than 9 years
old). These firms also have a high probability of being delisted before their fifth year
anniversary, up to 32%. The dramatic difference in underperformance of young firms
relative to old firms (above 40 years) is significant and robust to several benchmarks:
market indexes, size and book-to-market matched portfolios, and industry portfolios.
28
Brav and Gompers (1997) suggest that IPO underperformance occurs primarily in
small non-venture-backed IPOs; additionally, underperformance is related to small firms
with low book-to-market rations regardless of being an IPO. This paper establishes a
clear connection between firm age, size, and book-to-market ratios, showing that age
monotonically increases as size and book-to-market increases. However, our results show
that underperformance is mainly attributed to young firms, particularly to small and mid
sized ones within the young age category. Mature and old firms show very few cases of
significant underperformance, and the instances where underperformance is identified the
magnitude is significantly smaller and cannot be attributed to any particular size or book-
to-market quintile.
Gompers and Lemer (2003) show in an earlier sample, from 1935 to 1972, EPOs
do not underperform benchmarks on the aggregate, implying that observed
underperformance may simply be the result of a small sample. The results presented are
consistent with their results. Prior to 1971 there is no statistically significant
underperformance. However, prior to 1970, the percentage of young firms was very
small; post-1971, with the introduction of the Nasdaq, the percentage of young firms
listing grew dramatically to account for 50% of all new listings. Therefore, it seems that
post-1971 IPO underperformance is linked to the explosion of young firms listed both on
the Nasdaq and NYSE. Fama-French (2004) argue that in the 80's and 90's profitability
becomes more left skewed and growth more right skewed; the result leads to a sharp
decline in new list survival rates due to delisting caused by poor performance. An
expansion of weaker firms with low short-term profitability and high growth would
increase the number of failures and overall IPO underperformance. Indeed, young firms
29
have both high growth rates and longer profitability terms than older firms. Hence, the
increase during the 80's and 90's in the number of young firms going public would cause
a higher probability of being delisted. Consistent with the Fama-French findings, there
was a significant decline in firm survival rates. Our results show that IPO
underperformance is mainly attributed to a large group of young firms that in contrast to
older firms, offer very high returns in their star state (highly successful firms) making
them as a group highly positively skewed while at the same time have very high failure
incidence.
This effect may exist because young-small firms become "lottery-like" stocks to
investors who are overoptimistic about their growth perspectives and overprice these
stocks while undermining the loss probabilities. Given their small percentage of the
overall IPO market value, under cumulative prospect theory these stocks might support a
heterogeneous holdings equilibrium as proposed by Barberis and Huang (2007) and
therefore would not be socially wealthfare destroying. Additionally, analysts may
contribute to the overall over-optimism by systematically ascribing extremely high
growth rates to young firms, given that they do not have enough information and/or
history to differentiate good firms from bad ones.
Our results suggest that investors might overprice young-small IPOs thinking that
they may hold the next "Ebay" while underweighting their failure/delisting probabilities;
implying that the observed ex-post high IPO underperformance, which can be attributed
to young firms, may not be so "puzzling". Particularly once we account for the large
number of young firms which constitute more than 36% of the sample, we would expect
to observe overall IPO underperformance.
CO CO
CO
CO CO
0
010)0
0)
CO 00
-si O
Ol
CO
CO
CO
CO
o o o o
A
W
M
-*
01
CD
O
l C
O
Ol
CO
J
* ho
CO
-
J *
• o
M
CO
CO
Ol
Ol
-s|
O
£ "
to
Ol
u
CO
en
N
J
CO
en
-k
CO
8 C
O
•t*
CO
CO
^ CO C
O
J»
^J
CO
A
O
) s
a
N>
-»•-*-*•
W
ho
O) cn O) CD o -P*. CO
^
oi p
-si gi
A
w
£. CO
CO
O)
Ol
CO
o o
6 8
CO
CO
CO
hO
hO
OO
Ol
OO
OO
Ol
OO
M
hO
CO
N)
5 6
CO
CO
£>
.
CO
Ol
NJ
O
bi
b
N)
0>
co
ro
40.0 42.1
42.0 40.0
43.0 44.8
34.0 41.5
40.5 52.0
_k
_k
-*
{sj
CO
ro
304 728 740 o
CO
O
l o
en
o
Ol
£3
§ 00 ro
CO
l>0 o
CO
C
O
o
Ol
lo
o 8
8 ro
o
CO
2 -si o
o
Ol o
o
00 o
o
00 co
-J
CO
Co
CO
Ol
O
Ol
00
O
O
O
O
-si
Ol
o
CO
o
Ol s CO
o
Ol
CO
o
^1
CO
o
CO
2 w
Ol
A R
o
Ol
^1 o
o
CO
0
0
a ro
w
CO
N
)
Ol
Ol
o
o
o
CD
-s
)
o
CO
o
S
0)
-sj
o o
Ol
N) o
o
Ol
o
-si
00 o
00
CO
2 CO
C
O
Ol
-si o
-si
-sj
CO
C
O
o
o
o
-s|
o
CO
o
-si
ro
ro
00 ro
CO
ro
o
Ol o
o
CO
o
CO
o
CO
52
Ol 8 C
O
o
Ol
-s|
o
Co o
o
CD
C
O
* o
-s(
-si
ro
-si
-J o
Ol
CO
o
Co o
CO
O
l
ro
-si
CO
-s
i
CO
8 CO
C
O
o
o
o
CO
O
l o
CO
ro
o s C
O1
Ol o
00 ro
o
CO
C
O _
o
ro
8 ro
Ol
CO
8 o
00 o _ Jk
o
0>
_
^
8 CD
CD
Ol
ro
Ol
o
Co o
o
CO
ro _
o
o
ro
ro
ol
CO
o
o
Ol
o
-•J o
o
00 o
CO
8 s ro 8 C
o
o
-si
o
CO
ro
o
CO
8 CO
Ol
-si
-si o
CD
O
CO
-s
i _
o
-si _
ro
CO
s to 3 C
O 8 C
O
o
CO
o SI
_
o
CO
_ k
ol
o
o
o
o
CO
C
D
Ol
£
CD
C
D
CO
C
O
CO
C
O
CO
C
D
CO
CO
C
O
CO
C
O
CO
C
O
-»•
ro
o
i -»
• o
-*
a>
cf
lo
cx
oo
o-
Ni
-s
j-
*c
5o
ro
->
-K
w
ro
o
co
co
ro
o
co
° °
o oi
>
P vP
^p
cy
s tf
s tr
s
CO
CO
ro
ro
co
CD
o
co
o
co
b
en
en
b
o
en
en
8 8
4*.
A
co
ro
rp
co
oi
ho
Ol
00
-*
00
-»•
CD
Ol
CO
8 S
O)
-*
O
to
o)
s ro
5
380 543 o 8
ro
00 353 o
Ol
249
902 o
0
0 C
O
157 300 o
Ol
ro
680 4431 _
ro
o
en
ro
o>
_
CO
-
o
191 190 o
CO
ro
152 227 . ro
ro
3 142 o
CO
112 156 o & 245
092
o
CO
-
J
404 288 o
ro
0 o
o
o
-*
o
v
| (D
N
N
b
(D
01
Ol
O
Ol
00
-J
CO
->•
bo
*.
co
5
_*.
_k
p
p
r*
r*
r*
ro
b)
co
co
'-*
'-*•
co
-»•
-*
co
_».
_*
o
-*
^ ;-
* co
co
b
-s
i b
ho
b
b
-*
ro
co
-^
co
co
co
8S
8 co
_
»._
»._
» -*
ro
ho
O
l O
l A
-*
O
-*
H
cr
CO
65
3 TEL
er>
n •<
n S
i 1 o
o
cr
ort
_ hS
&
S*
o
s?
fV
rin 79
us
» D a C/J
e 3 3 M
•5
i73
<-•
•
» !-••
(/)
«• r»
c/>
o
2 do'
s n >
ore
n W
cr
5'
n
"i
O
v>
VI re
8 8
8 5
o o
o <o
8*
§
C0
(0
(0
(O
(0
(D
CD
(D
(0
CD
tD
(D
(D
(0
(D
(0
tO
<D
(0
tD
(C
(O
(0
(D
(0
CO
(D
(O
CD
t
0(
D(
0(
D(
DC
D(
0t
0t
DO
CD
C0
CD
00
Ca
aC
0C
0C
0S
SS
N^
S'
JS
SS
C
0-
sJ
0>
Cn
-^
Wr
0-
iO
<D
00
-s
J0
)O
l4
^W
N3
-»
O(
DC
0-
Nj
0)
ai
AW
KJ
-»
.O
to
*j
-g
yi
*w
5 to
-»
->
• s
N
-*
A
Ol
O
CO
C
7I
O
O)
*.
co
Ol
to
-t»
o N
>
<o
o t;
i a>
o o
co
en co
en
en
en
o>
**
co
SN
)K
)f
l)
^Q
-'
CO
g)
A
->
O)
Ul
CB
C
D
N
b b
b
b
b
b
b
S
S
^
oo
oo
co
co
oo
-*io
op
o b
bb
bc
nb
bb
-i
CO
0s
_i s •t
* .24%
Ol .76%
ro
.82%
N) .82%
0) b to
5?
o .89%
M 48%
o to
CO
as
o 85% o 60%
o O
63%
o CO
o o
o
o
o
-NJ
CO
C
O
CD
-^
J
b b
b
b
b
C
CC
OO
lO
OC
OC
OC
OC
OO
l
bb
bb
bb
bc
nb
o o
o
o
o o)
en
j*
bo
'-*•
co
K
J
ro
to
ro
-*
-*
-*
-*
-*
ro
r?
-'
M-
'C
OO
NJ
IO
tD
Ol
Ci
CO
co
cn
o>
o
co
o> e
n
sb
)C
OC
O(
OC
ob
lU
SN
3C
OO
«-
»M
N-
* CO
fv
j •S
C
O
•vj
-si
-pk
fO
O
-»
M
W
N)
oi
-»
ce
(o
o)
ao
^u
A
sc
ow
-i
sJ
.-
»S
S«
ta^
^K
oi^
bto
i&ii
go
sy
sto
sa
s&
Jb
igs
co
-f»-
en
en
co r
o
M
01
(0
^ _
>.
-A
_
i.
_*
ro
m
ro
ro -
»•
eo
ro
ro
- —
-s
i C
D
CD
O
A
ffl
O
S
Ol
-*
* 8
-*
co
*»
-*
co
ro
ro
ro
-*
*»
oc
oc
no
)o
oe
n-
*k
Q-
gA
-*
tD
^o
p-
sj
cD
co
io
9s
iD
-i
wc
oc
oo
os
cn
o)
A-
*^
cn
^M
co
o
ro
o
ro
ro
ro
en
en *
».
*.
N
CO
4
i,
CD
-^
-*
-s
jr
o-
^.
-*
or
oc
oc
n-
i^-
Nic
o-
Nj
co
o
co ro
ro
o co
o ^
-*•
en
4*.
o> -*
co
pp
oo
op
pp
po
oo
pp
pp
pp
pp
pp
pp
pp
^^
pp
pp
p
tn
ww
Au
ww
AA
cn
AW
tn
wb
AO
iy
iA
Ac
jA
ff
is
co
bM
N^
S0
)-
'N
)S
yi
WU
l0
3-
»(
0^
-*
01
-»
tD
ON
3W
0)
0)
(0
OI
O>
JS
-l
OC
a0
1*
C0
O
o o
o
o
o
o
o
NS
AO
li
-U
lj
i.
. _
. o
p
©
p
p
oi
en
co
-J
-J
op
pp
pp
pp
p o
iN
'v
ic
nc
nb
cn
oi
cn
OO
NJ
Ol
SM
OO
CD
AS
o o
o
o
ro
en
to
-u
oo
oo
pp
oo
oo
op
pp
pp
pp
co
co
cn
wb
ib
ib
ib
)O
is
s(
0^
^(
DS
>J
s W
(D
-»
0)
WO
)M
OC
O-
A(
OO
I>
CO
-'
NM
N
pp
pp
pp
^^
^r
o^
pp
p b
NC
o^
co
iD
^k
jc
no
bc
oN
CD
c
oc
Dr
oC
DC
oe
nc
oc
nc
oc
Dc
or
oc
oc
D
p p p p p
en o> en o oi
-j oi en co o) p p p p ;-*
bi b> CD
'-
*
co co -* en o ->>
co co b
o -* o -»•
_». _» _* -*. ro
§2UK8 -k
O
->•
co '
-* o
en o eo
* O
0
O
s o>
-»•
oo
<o
w
'
• •
• i
• i
• •
••»
oo
en
w
O
l O
l O
^
1
00 —
'
Wh
0^
.J
>.
h0
OO
->
--
^^
l-
vl
O
U1
^l
tD
<3
5N
)W
OO
O-
»-
CO
Oi
a>
en
OO
-S
*•
On
oo
-S
-&
•
03
en
oo
en
*•§
^1
^1
O
on
^ 1 *.
ro
ro
ai
en
eo
a>
N
)
oo
oo
oo
oo
oo
oo
c
oc
no
no
n'
.P
'-
'-
t'
.*
.^
*.
*.
**
MO
S-
»0
10
)*
MU
(S
O)
2 PP
PP
PP
PP
PP
PP
c
ob
o^
^c
^e
nr
oe
ne
ne
nr
ob
) c
oo
oc
oc
»-
>i
co
o^
ic
ji
on
oJ
&.
-^opopppooooo
bi
Bi
ob
sb
ib
ib
ib
ic
nb
is
O*
>O
SU
l0
0f
fl
S«
Ol
D^
-'
-'
-'
-'
OO
OO
OO
OO
_>
. _
»
_i
O
CO
-i
M
->
(D
O)
CO
-g
^>
IO
OO
) 0
->
0(
00
1W
M(
J1
-n
3 >
(a
CD
%
o -*
TJ
O
</)
00
01 3 •a
CD
s?
to
en s s?
3 <D a a 3 3 CD
01
3 ~J cn
3"
s?
TJ
Dl
3 (D
00
CO
o ok- o
3 0) •n
*•
<D
i-* o
3 *+
3"
(0
31 3 *• a ID
«<
o -*
B) a
3 (O
o sr I to
Si «
+
S
ill
t N
(I
I U
O
S
—
»
M
-t
O)
_k
00
(O
o
(O
O)
*>.
K)
• i
ii
•>
C9
ID
U
ro
~
- —
—
OI
*>
UI
SO
-'
U-
'N
)U
IM
w
eo
en
en
eo
-'
Oe
o-
^(
0~
40
^i
oo
-^
oo
oo
.u
ro
on
oo
o-
vi
en
A
•&•
en
-
j ^
on
^
b
o b
o
o
-J
oo
-s
_*
^1
-t*
_k
ro
en
o
eo
ro
O
en
-8
ro
ro
o
->•
-J
CO
O)
O
SO
NO
lU
lA
^J
VU
IO
M-
' O
-'
UU
I-
l'
-'
-'
OC
OU
lW
IO
wc
oM
bi
ki
ui
Mu
bi
ui
Mo
UM
SJ
kM
OI
dl
O-
'-
'-
'O
114 o
202
o
175 CD
eo
158 £
137 CO
k
ro
oo
o
CD
-J
cn
106 ro
o
105 ro
eo
eo
00
•r>
C5 en
en
365 -><
CO
905 en
534 ro
ro
435 ro
o>
308
385 00
250
234 -o.
o
283
05
309 o
363 05
230 en
o
149 o en
oi
^e
oe
or
or
or
or
or
or
o
05 tn co eo CD
en eo
'•>i
b * o oi ro o
CJ1
eo oo eo
M
CO S
_i CD
ro CD
00 O
OO O
eo co -
en * 0
•+I
T Q
in S?
3 •o »
%
ro
°L
5
5? £ <D a u 3 3 CD
»>
•o
3 £.
>
3 S! a- a < e.
c (D
* o 3 #
•*
3"
efi 3 r*
a B>
«< o e*
3 a.
5'
(Q
—*
H
cr
bJ I n o
a"
>
ara n o
o
B
•a
*« o
0. — aft-k
. CO
Q
)
>. %
o
-^ a = re > 4^ •{£ CO
2 6 o
c CO
a.
s?
£ to
t^
c « fl) S
£ n 1 i S?
£-. u>
CM
s?
o
a.
1 O
T
.«> b
9: >
0 * 4) o> < I E
in 107 o
CM
— CM
Si CM
s 3 CM
CO
CO
CM
O
o co CM
CM
O
co CM
co CO
O
CO
O)
CM
s
CO
in
05
o
CM
0>
O
s co
CO
1^
CO
•St
in
O
CM
CD
CM
a>
co
•* C
M
O
CO
io
t-
ns
oi
mi
o<
ji
ii
)(
>)
commcnin-*co-<-
CM CO CM CM CM CM •*
O
r-CM in
o o
-, — r
an
•«t
co co co co
O
(D
CO CM
LO oS
co to co
o>
r o \f CD CO
CO O) T-
s in
co
co
r- CM
OCOCOCMOOTT-I^OCO'"-'"-
^i
ni
no
io
iN
NO
Oi
s^
o c
dc
d'<
-:i^
cM
'T)
:co
cd
-<
-:c
Mo
S-
>-
:
T-
T-
CM
CM
cO
CO
CO
-d
-C
DC
DC
Or
-
o CM
CM
CO
CD
CO
CO
CO
1^
CM
co
•«t
co r>-
"-co o
co co l-~
CJ> in
v—
CO
•*
CD lO
•
*
co
W
«)
>-
0)
rO
Nl
D(
I>
OI
Ot
n
OI
"»
CO
<O
CM
O"
>O
CM
mi
J-
CM
CO
C
Mi
nC
MT
-c
OO
OI
^l
f)
'*
'*
lO
£ £ 8. 5. S £ I
I CM
T
f co
a>
^ •
- "
. ?
•*. y
».
i—
^—
%-••
^ ••»
_ |
^
O)
I n o
o QL I
.
1 n to o
>> o
c Q
. 4-<
0) •s ro
S
6 k-t o
o
CO
Q 0)
c ro D
.
ss £ in h
-C
CO
a> £ c CO T5
S v?
o^ £ m
CM
S?
<D
a E
CO
OT
m
b a.
»
o
* o
o> < UJJ •
—
1
"-I lo
oc
oi
oc
Do
mi
or
-n
-i
o'
^-
co
co
ioin
mc
oc
oi"
-c
oo
oc
M
cz
io
oc
io
ci
ci
do
-r
^^
^
i-
«(
B<
tl
fl
»N
NS
(!
>'
t'
-c
ot
o^
'^-
'^t
tr
io
io
co
t-
co
o o
dd
dc
id
cJ
od
cJ
dT
-:
§ 5
co
co
Wr
^c
Mt
om
co
co
oo
oo
oo
oo
oo
oo
oo
(O
't
oo
rn
in
co
t C
MC
MC
MC
MC
MC
MC
MC
MC
O
CO
T-
lO
CO
-<t
•<*• O
OO
OO
OO
OO
OO
O
O
CM
CM
m
CD
co CD co
CM
O)
"* r»
*-
co CO
O
00 r
e'
en
«
81%
•*
54%
•*
CO
r-m
ST
- CO
O
CO
S
ID
U)
r C
B
O
CM
lO
CM
t- CO
O
O
(O
CD
O
) f-
CM
CM
CM
^
- •*
CM
CO
f-
in *
•*• m
^,
a
N
o in
CO
IO
00
T-
'J
.J
L'
Pi
S.
co 0>
T-
lis
34
Table 1.3: Initial Public Offerings Characteristics by Industry
The sample is composed of 9,400 identified IPOs from 1935 through 2002, grouped in 12 industry categories following Fama and French classification: 1) Consumer Non-Durables - food, tobacco, textiles, apparel, leather, toys; 2) Consumer Durables - cars, tv's, furniture, household appliances; 3) Manufacturing - machinery, trucks, planes, office furniture, paper, commercial printing; 4) Energy - oil, gas, and coal extraction and products; 5) Chemicals and allied products; 6) Business Equipment - computers, software, and electronic equipment; 7) Telecommunications; 8) Utilities; 9) Shops -wholesale, retail, and some services; 10) Healthcare - medical equipment, and drugs; 11) Finance; 12) Other - mines, construction, building materials, transportation, hotels, business services, and entertainment. Panel A presents the number of IPOs, sample percentage, average, median, 25th percentile, and 75th percentile, and ranking on the mean of the firm's age at the time of the IPO for each industry category. Panel B presents the number of IPOs, sample percentage, average, median, 25th percentile, and 75th percentile, and ranking on the median of market value at the close of the first day of trading (in constant 2002 dollars adjusting for the Consumer Price Index) for each industry category. Panel C presents the number of IPOs, sample percentage, average, median, 25th percentile, and 75th percentile, and ranking on the median of the ratio of book to market equity value at the close of the first day of trading for each industry category. For each firm the ratio is set to the average of the three SIC digit industry book to market ratios. Panel D presents the number of IPOs, sample percentage, average, median, 25th percentile, and 75th percentile, and ranking on the median of the ratio of book to market equity value for surviving firms one year after the IPO for each industry category. For each firm the ratio is computed using Compustat data for the book value when available or set to the average of the three SIC digit industry book to market ratios at the time.
Panel A: Firm Age (On the first day of trading)
Industry
Non-Durables
Durables
Manufacturing
Energy
Chemicals
Business Equipment
Telecommunications
Utilities
Shops
Health
Finance
Other
# of IPO's
666
214
801
238
156
2,240
313
128
1,231
988
1,055
1,370
Sample %
7.09%
2.28%
8.52%
2.53%
1.66%
23.83%
3.33%
1.36%
13.10%
10.51%
11.22%
14.57%
25th %
7.00
5.00
7.00
3.00
5.00
4.00
3.00
10.00
4.00
3.00
4.00
3.00
Median
18.00
12.50
19.00
9.00
15.00
7.00
5.00
36.00
10.00
6.00
15.00
8.00
Mean
32.04
24.82
31.19
18.17
26.42
10.89
11.48
36.36
19.90
9.61
34.84
15.44
75th %
52.00
31.00
49.00
23.00
36.00
13.00
13.00
53.00
27.00
10.00
58.00
18.00
Ranking
10
7
9
5
8
2
3
12
6
1
11
4
Total/Avg 9,400 100% 4.83 13.38 22.60 31.92
35
Panel B: Market Value (On the first day of trading)
Industry # of IPO's Sample % 25th % Median Mean 75th % Ranking
Non-Durables
Durables
Manufacturing
Energy
Chemicals
Business Equipment
Telecommunications
Utilities
Shops
Health
Finance
Other
666
214
801
238
156
2,240
313
128
1,231
988
1,055
1,370
7.09%
2.28%
8.52%
2.53%
1.66%
23.83%
3.33%
1.36%
13.10%
10.51%
11.22%
14.57%
37.09
41.36
43.19
61.64
45.49
46.24
73.98
169.68
33.54
33.39
33.65
40.53
98.34
83.35
104.85
130.61
131.06
127.70
203.98
359.17
83.41
84.81
80.38
104.96
314.98
245.64
214.56
403.59
425.57
457.00
614.11
712.99
194.21
229.21
448.56
297.63
305.04
182.47
216.68
330.03
441.70
335.52
539.93
645.71
195.03
188.27
249.06
252.18
5
2
6
9
10
8
11
12
3
4
1
7
Total/Avg 9,400 100% 54.98 132.72 379.84 323.47
Panel C: Book-to-Market (On the first day of trading)
Industry # of IPO's Sample % 25th % Median Mean 75th % Ranking
Non-Durables
Durables
Manufacturing
Energy
Chemicals
Business Equipment
Telecommunications
Utilities
Shops
Health
Finance
Other
665
214
801
238
156
2,240
313
128
1,225
985
1,053
1,313
7.13%
2.29%
8.58%
2.55%
1.67%
24.01%
3.35%
1.37%
13.13%
10.56%
11.28%
14.07%
0.60
0.59
0.67
0.48
0.55
0.42
0.41
0.68
0.58
0.37
0.72
0.46
0.77
0.70
0.84
0.68
0.68
0.50
0.49
0.84
0.71
0.44
0.98
0.61
0.88
0.80
0.93
0.71
0.73
0.54
0.57
1.02
0.79
0.46
1.06
0.72
1.05
0.95
1.07
0.88
0.87
0.61
0.60
1.11
0.90
0.51
1.22
0.87
9
7
10
6
5
3
2
11
8
1
12
4
Total/Avg 9,331 100% 0.54 0.69 0.77 0.89
Panel D: Book-to-Market (One year after IPO)
Industry # of IPO's Sample % 25th % Median Mean 75th % Ranking
Non-Durables
Durables
Manufacturing
Energy
Chemicals
Business Equipment
Telecommunications
Utilities
Shops
Health
Finance
Other
656
214
792
238
156
2,212
308
127
1,215
983
1,050
1,340
7.06%
2.30%
8.52%
2.56%
1.68%
23.81%
3.32%
1.37%
13.08%
10.58%
11.30%
14.42%
0.32
0.27
0.31
0.33
0.30
0.18
0.23
0.54
0.26
0.22
0.52
0.23
0.53
0.43
0.54
0.52
0.49
0.30
0.40
0.68
0.43
0.32
0.87
0.41
0.66
0.54
0.64
0.57
0.51
0.38
0.58
0.79
0.55
0.36
1.00
0.54
0.85
0.69
0.88
0.71
0.66
0.49
0.63
0.88
0.66
0.45
1.17
0.67
9
6
10
8
7
1
3
11
5
2
12
4
Total/Avg 9,291 100% 0.31 0.49 0.60 0.73
36
Table 1.4: IPO characteristics using Fama French 25 portfolio classification
The sample is composed of 9,400 identified IPOs from 1935 through 2002, IPOs are sorted based on Fama French size and book-to-market quintiles. Panel A presents the number of IPOs, sample percentage, firm age median and average using information available at the close of the first day of trading. For each firm the book to market ratio is set to the average of the three SIC digit industry book to market ratios. Panel B presents the number of IPOs, sample percentage, and firm age median and average using information available one year after the IPO. For each firm the ratio is computed using Compustat data for the book value when available or set to the average of the three SIC digit industry book to market ratios at the time.
Panel A: Fama French Portfolios Sorted based
Low B/Mkt
2
3
4
High B/Mkt
Low B/Mkt
2
3
4
High B/Mkt
Panel B
Low B/Mkt
2
3
4
High B/Mkt
Low B/Mkt
2
3
4
High B/Mkt
Small
46
494
519
381
138
Small
7
5
6
7
12
Number of IPO
2
95
684
735
610
182
3
92
908
898
638
205
I'S
4
78
669
697
422
160
Median Age of the IPO
2
9
7
9
11
20
3
8
7
10
13
15
4
17
7
9
15
23
on Information Available after the First Day of Trading
Big
35
208
236
152
49
Big
38
8
9
45
49
: Fama French Portfolios Sorted based on
Small
473
521
409
305
199
Small
5
5
7
7
11
Number of IPO
2
595
641
442
358
211
Median
2
6
8
11
13
24
3
841
735
427
261
140
I'S
4
953
470
262
180
90
Age of the IPO
3
8
9
12
13
38
4
8
13
14
24
42
Big
349
127
124
68
58
Big
8
26
25
51
61
Small
0.49%
5.29%
5.56%
4.08%
1.48%
Small
10.85
7.79
11.33
15.52
30.61
Percentage of Sample
2
1.02%
7.33%
7.88%
6.54%
1.95%
3
0.99%
9.73%
9.62%
6.84%
2.20%
4
0.84%
7.17%
7.47%
4.52%
1.71%
Average Age of the IPO
2
16.12
11.50
17.22
22.20
34.02
3
15.73
13.24
18.48
28.15
32.44
4
27.55
14.96
19.61
28.59
35.04
Big
0.38%
2.23%
2.53%
1.63%
0.53%
Big
43.11
22.22
27.12
51.51
55.51
Information Available One Year After the IPO
Small
5.07%
5.58%
4.38%
3.27%
2.13%
Small
7.85
9.16
12.37
17.61
23.45
Percentage of Sample
2
6.38%
6.87%
4.74%
3.84%
2.26%
3
9.01%
7.88%
4.58%
2.80%
1.50%
4
10.21%
5.04%
2.81%
1.93%
0.96%
Average Age of the IPO
2
10.66
13.63
18.95
25.70
37.15
3
14.44
17.24
22.08
29.41
47.06
4
14.83
25.30
27.99
36.61
46.53
Big
3.74%
1.36%
1.33%
0.73%
0.62%
Big
21.29
37.37
45.73
54.51
64.53
37
Table 1.5: First Day of Trading Return and Survival Analysis
The sample is composed of 9,400 identified IPOs from 1935 through 2002, grouped in 12 age categories. Panel A and B present the number of delisted and merged firms as well as delisted, merged, and survival percentages after three and five years respectively. Panel C presents the average, median, 25th percentile, 75th percentile, maximum, and standard deviation for the first day of trading return per age category, the return is computed using the low and closing prices recorded in CRSP.
Panel A: Firm Survival 3 Years After IPO
Firm Initial # Delisted Merged Delisted Merged Survival
Age
0
1
(2-3)
(4-5)
(6-8)
(9-12)
(13-18)
(19-27)
(28-40)
(41-55)
(56-75)
75+
Total
of IPO's
207
578
1290
1175
1332
1104
1013
733
551
450
430
537
9400
Firms
26
104
212
130
109
71
56
30
15
10
4
12
779
Firms
14
54
124
141
138
110
87
66
50
31
34
40
889
%
12.56%
17.99%
16.43%
11.06%
8.18%
6.43%
5.53%
4.09%
2.72%
2.22%
0.93%
2.23%
8.29%
%
6.76%
9.34%
9.61%
12.00%
10.36%
9.96%
8.59%
9.00%
9.07%
6.89%
7.91%
7.45%
9.46%
%
87.44%
82.01%
83.57%
88.94%
91.82%
93.57%
94.47%
95.91%
97.28%
97.78%
99.07%
97.77%
91.71%
Panel B: Firm Survival 5 Years After IPO
Firm Initial # Delisted Merged Delisted Merged Survival
Age
0
1
(2-3)
(4-5)
(6-8)
(9-12)
(13-18)
(19-27)
(28-40)
(41-55)
(56-75)
75+
Total
of IPO's
207
578
1290
1175
1332
1104
1013
733
551
450
430
537
9400
Firms
51
184
357
217
197
134
104
60
29
31
19
27
1410
Firms
29
105
211
236
262
203
175
130
85
64
68
75
1643
%
24.64%
31.83%
27.67%
18.47%
14.79%
12.14%
10.27%
8.19%
5.26%
6.89%
4.42%
5.03%
15.00%
%
14.01%
18.17%
16.36%
20.09%
19.67%
18.39%
17.28%
17.74%
15.43%
14.22%
15.81%
13.97%
17.48%
%
75.36%
68.17%
72.33%
81.53%
85.21%
87.86%
89.73%
91.81%
94.74%
93.11%
95.58%
94.97%
85.00%
00
E
3 4
-* a> a. O
) c
•a ra l-»•-o >. ra a *
•*
!2 iZ
6 0) c (0 a.
> a> a +•* 00
X
cs 2 s?
£ in
i»-
c n
0) 2 c <a •o
a> 2 s?
£ m
CSI
a> T
O
< § LL
CO
CO
T-
00 CO
»
-•*
o
N^
Sy
"st* N
^ o^
o^ 0
s* o^
in
•* o
o
(O
ID
M
01
N
O)
O)
CM
CD
in co
co in
si N
.'
•<t C
N
CSI •*
in
to
to
Vg
vP
CO
O
T
-W 00
co csi
sP V
P co
m
CN
C
N
co o> csi
m
co CSJ
m
co
oo
m
co
•** co
co
co in
in in
oi
as co
o oo
C
N co
CM
i-
i- CM
T
-
O
CO
O
"*
so
v"
>*P 0
s- 0
s" 0
s" co
a> o
CO
CN
T
J-
CN
00
CN
r-
in in
p cq
CO
C
D
CD
T
f
in co
CM
C
\J
o
in in
in
o o
in in
csi csi
m
oo co
a>
co
N-
co co
•* co
in co
CN C
si
Is-
Is-
CM
CO
* w
to
in CN
vo
^p
0s-
0s
CO
O
CD
I
s-csi
csi
CD
C
O
CN
CN
csi
csi
^p
so 0
s- 6
s-in
co
05 05
00 CO
«
t
d
d
o
o
d
d
6s-
CO
d
d
o
o
^-
oo co
25" 0
?
Is-
CM
in
in
m
r>-
.~ i
i i
i n
o> oo
T
- io
.i il
o i-
£S
2-
S£
.2J
.CC
$N
].2
.l£-
rs-
39
Table 1.6: Cumulative Abnormal Returns by Age Group
The sample is composed of 9,400 identified IPOs from 1935 through 2002, grouped in 12 age categories. Panel A presents the three-year cumulative abnormal returns relative to six value weighted benchmarks: 1) Fama and French size and book-to-market quintiles portfolio; 2) Fama and French 12 industry portfolios; 3) NYSE, Amex and Nasdaq stocks; 4) Nasdaq stocks; 5) NYSE small stocks; and, 6) S&P500 stocks. The left columns present the cumulative abnormal returns and the right columns the corresponding t-statistics which are computed following Ritter's 1991 paper. Panel B presents the five-year cumulative abnormal returns. Panel C and D present three and five year cumulative abnormal returns excluding the firms that ex-post will be delisted.
Panel A: Three Year Cumulative Abnormal Returns
Firm Age
0 1
(2-3)
(4-5) (6-8)
0-12) (13-18)
(19-27)
(28-40)
(41-55)
(56-75) 75+
All IPO's
FF5x5
-54.3% -49.4%
-40.0% -30.4%
-28.3%
-13.2%
-15.5%
-9.6%
-7.8%
-13.1%
-9.1% -6.2%
-23.2%
IND12
-53.8%
-54.5%
-37.7%
-22.3%
-24.9%
-13.4%
-13.3%
-8.9%
-3.0%
-9.0%
-4.8% -2.1%
-20.5%
NY/A/N
-50.8%
-47.6%
-34.2%
-19.7%
-19.7%
-6.2%
-8.5%
-3.0%
0.9%
-6.2%
-1.8%
0.7%
-15.9%
Nasdaq
-53.9%
-52.9%
-34.5%
-22.0%
-24.8%
-13.6%
-12.5%
-6.4%
-7.0%
-12.8%
-0.1%
-3.9%
-21.1%
Small
-42.3% -41.7%
-37.1%
-26.1% -23.8%
-9.5%
-10.6%
-4.0%
-6.2%
-8.5%
-4.1%
0.8%
-18.5%
S&P500 -54.6%
-51.5%
-37.3%
-22.2%
-22.2%
-8.8%
-11.3%
-5.6%
-0.9%
-8.6%
-4.0% -1.7%
-18.6%
FF5x5
-4.6
-7.1
-8.2
-5.9 -7.2
-3.4
-4.0
-2.4
-2.2
-3.5
-2.5
-1.9
-16.7
IND12 NY/A/N
-4.5 -7.7
-7.8
-4.3 -6.4
-3.4
-3.4
-2.2
-0.8
-2.4
-1.3 -0.6
-14.7
-4.2
-6.8
-7.0
-3.8 -5.0
-1.6
-2.2
-0.7
0.3
-1.6
-0.5 0.2
-11.4
Nasdaq -4.1
-7.0
-6.9 -4.2
-6.1
-3.2
-2.9
-1.3
-1.3
-2.2
0.0
-0.9
-13.3
Small -3.6
-6.0
-7.6
-5.1 -6.1
-2.4
-2.7
-1.0
-1.7
-2.2
-1.1 0.2
-13.3
S&P500 -4.5
-7.3
-7.6
-4.3 -5.6
-2.2
-2.9 -1.4
-0.2
-2.2
-1.1
-0.5
-13.2
Panel B: Five Year Cumulative Abnormal Returns
Firm Age 0
1
(2-3) (4-5)
(6-8)
(9-12)
(13-18) (19-27)
(28-40)
(41-55)
(56-75)
75+
All IPO's
FF5x5
-75.5%
-71.8%
-50.6%
-23.1%
-27.9%
-12.5%
-16.1%
-16.2%
-14.5%
-19.3%
-23.8%
-20.1%
-28.1%
IND12
-70.1%
-70.5%
-42.6%
-11.8%
-22.0%
-8.5%
-6.4% -8.7%
-5.2%
-9.6%
-15.9%
-9.9% -20.5%
NY/A/N -63.1%
-60.4%
-36.0%
-2.0% -10.5%
3.9%
1.9% -0.6%
1.4%
-5.2%
-9.7%
-6.2%
-11.9%
Nasdaq -70.4%
-72.4%
-39.7%
-8.3%
-18.8% -7.1%
-4.6%
-5.9%
-11.0%
-19.5%
-16.8%
-20.2%
-20.7%
Small S&P500 -51.0% -56.1%
^2.8% -13.4%
-17.5%
-3.8%
-6.1% -5.4%
-10.5%
-12.6%
-16.4%
-8.5% -18.4%
-69.1%
-65.3%
-40.0% -4.9%
-14.0%
0.1%
-1.8%
-4.4%
-1.5%
-9.0%
-13.1% -9.7%
-15.6%
FF5x5 -5.0
-7.3
-7.1
-3.3 -4.9
-2.2
-3.3
-2.8 -2.9
-3.3
-4.0
-4.2 -14.4
IND12 NY/A/N
-4.8 -7.3
-6.1
-1.8 -4.0
-1.8
-2.1
-1.8
-1.3
-1.6
-2.5 -2.2
-11.3
-4.4
-6.3
-5.2 -0.7
-2.1
0.4
-0.4
-0.4
0.0
-0.9
-1.5
-1.6 -7.1
Nasdaq -4.3
-6.7
-5.4
-1.5
-3.5
-1.6
-1.5
-1.3
-1.6
-2.0
-1.7
-3.1
-9.9
Small
-3.8
-5.8
-6.0 -1.9 -3.1
-0.8
-1.6 -1.0
-2.0
-1.8
-2.4
-1.8
-9.6
S&P500 -4.7
-6.8
-5.7
-1.1 -2.7
-0.3
-1.1
-1.0
-0.5
-1.6
-2.2
-2.3
-8.9
40 Panel C: Three Year Cumulative Abnormal Returns (Excluding Firms that will be Delisted)
Firm Age 0 1
(2-3) (4-5)
(6-8) (9-12) (13-18) (19-27)
(28-40)
(41-55) (56-75) 75+
FF5x5 -34.9%
-5.1%
5.0% -3.6%
-4.8% 1.9%
-3.7%
1.9%
-1.1% -3.6% -1.8% 1.7%
IND12 -34.2%
-11.0%
6.3% 3.6%
-1.8% 1.7%
-1.6% 2.6%
3.3%
0.9% 2.2%
5.3%
NY/A/N -31.4%
-3.3%
9.9% 6.4% 3.7%
9.1% 3.5% 8.5% 7.2%
3.8% 5.6% 8.6%
Nasdaq -32.7%
-5.5% 12.4%
5.8% -0.6% 3.4% 1.6% 8.3%
2.7%
3.5% 12.0%
6.9%
Small -24.2%
1.9%
5.6% -0.4% -1.4%
4.5% 0.7% 6.7%
-0.2% 0.4% 2.8%
8.2%
S&P500 -35.1%
-7.1%
6.9% 4.0% 1.4% 6.7% 0.8%
5.9% 5.4%
1.5% 3.4%
6.3%
FF5x5 -2.8 -0.7
1.0
-0.8 -1.2
0.5
-1.0 0.5
-0.3 -1.0 -0.5
0.5
IND12 -2.8
-1.6
1.3
0.8
-0.5
0.5 -0.4
0.7
0.9
0.2
0.6
1.6
NY/A/N -2.6
-0.5
2.0 1.4
0.9
2.3 0.9
2.2
2.0
1.0 1.6
2.6
Nasdaq Small -2.4
-0.7
2.5 1.2
-0.2
0.8 0.4
1.8
0.5
0.6 2.3
1.6
-2.0 0.3
1.2
-0.1 -0.4
1.2 0.2
1.7
-0.1 0.1 0.8
2.5
S&P500 -2.8
-1.0 1.4
0.8
0.4
1.7
0.2
1.5
1.5
0.4 1.0
1.9
Panel D: Five Year Cumulative Abnormal Returns (Excluding Firms that will be Delisted)
Firm Age 0 1
(2-3) (4-5) (6-8) (9-12) (13-18)
(19-27) (28-40)
(41-55)
(56-75) 75+
FF5x5 -37.3%
-5.8% 9.9%
14.9% 4.9%
10.7% 2.4%
0.0% -5.5% -6.6%
-9.8%
-9.4%
IND12 -31.5%
-5.1% 15.3% 24.8%
10.2%
14.0% 11.7%
7.2%
3.3% 3.3%
-2.4%
0.2%
NY/A/N -24.7%
5.2% 22.4% 35.2%
22.3% 26.6% 20.3%
15.3%
9.9% 7.7%
4.2%
4.3%
Nasdaq -28.7%
-1.5%
21.8% 30.7%
15.0%
17.9% 16.9%
15.2% 2.5% 1.7%
6.2%
-5.6%
Small -14.0%
8.6% 14.8% 23.6%
14.1% 17.9% 11.7%
9.7%
-2.4%
-0.9% -3.4%
1.6%
S&P500 -30.4%
0.4% 18.6% 32.3%
18.9% 22.9% 16.7%
11.6% 7.1% 4.1%
0.9%
1.0%
FF5x5 -3.0 -0.9 1.9 2.2
1.3
2.2 0.3
0.2
-1.0 -0.5
-1.3
-2.0
IND12 -2.8 -1.0 2.5 3.7
2.1 2.4 1.4
1.1
0.6
1.2
0.1
0.0
NY/A/N -2.4 0.1
3.6 5.0
4.2 4.7
3.2
2.7
1.9
1.9
1.3
0.7
Nasdaq Small -2.3 -0.4 3.5 4.3
2.7 2.8
2.3
1.9
0.3
1.0
1.3
-0.9
-1.8 0.6
2.6 3.7
2.9 3.3 1.9
1.9
-0.3 0.7
0.1
0.4
S&P500 -2.7 -0.5 3.0
4.5
3.5
4.0 2.4
2.0
1.4
1.2
0.6
-0.1
Pan
el E
: Thr
ee Y
ear
Cum
ulat
ive
Abn
orm
al R
etur
ns in
exc
ess
to N
YS
E, A
mex
, & N
asda
q
Fam
a Fr
ench
Por
tfol
ios
Sor
ted
bas
ed o
n In
form
atio
n A
vaila
ble
afte
r th
e Fi
rst
Day
of T
radi
ng
Low
B/M
kt
2 3 4 H
igh
B/M
kt
Low
B/M
kt
2 3 4 H
igh
B/M
kt
Low
B/M
kt
2 3 4 H
igh
B/M
kt
Sm
all
-66.
1%
-31.
3%
-56.
7%
-51.
0%
-6.2
%
Sm
all
-1.7
%
5.6%
-2
8.0%
-3
.0%
35
.5%
Sm
all
-34.
1%
6.3%
-2
7.9%
3.
3%
5.6%
Yo
un
g F
irm
s (
0 to
8 y
ears
)
2 -6
3.2%
-2
9.6%
-3
8.8%
-3
4.8%
-3
1.2%
3 -3
9.4%
-1
7.0%
-1
2.6%
-4
1.8%
-3
0.3%
4 -5
.9%
-1
6.1%
-1
7.3%
-3
4.2%
-1
9.7%
Mat
ure
Fir
ms
( 9 t
o 40
yea
rs )
2
-52.
3%
-3.8
%
0.2%
-9
.6%
-4
1.9%
3 -4
2.8%
1.
4%
1.9%
-1
5.5%
-3
1.0%
4 -4
3.7%
11
.4%
-1
3.0%
4.
3%
21.4
%
Old
Fir
ms
(41
+ye
ars
)
2 -5
.3%
0.
4%
-23.
8%
-1.8
%
-3.9
%
3 -4
8.8%
16
.0%
-1
3.6%
0.
7%
-29.
0%
4 -8
.7%
12
.7%
-6
.4%
2.
2%
-12.
4%
Big
-4
8.2%
-7
1.6%
-3
1.9%
-8
5.5%
-8
2.6%
Big
-4
0.9%
-3
3.3%
-2
.9%
-5
.6%
8.
6%
Big
-1
4.8%
0.
0%
20.0
%
-5.9
%
19.3
%
Sm
all
-1.9
2 -3
.26
-4.8
3 -4
.14
-0.2
9
Sm
all
-0.0
7 0.
51
-2.5
1 -0
.28
1.91
Sm
all
-0.2
6 0.
14
-1.4
8 0.
24
0.43
2 -3
.35
-3.5
2 -4
.55
-3.8
3 -1
.63
2 -3
.31
-0.5
0 0.
02
-1.4
8 -3
.16
2 -0
.21
0.02
-2
.84
-0.2
3 -0
.33
T-S
tatis
tics
3 -2
.09
-2.3
8 -1
.21
-4.7
4 -1
.82
T-S
tatis
tics
3 -2
.32
0.20
0.
33
-2.4
6 -2
.84
T-S
tatls
tics
3 -1
.62
1.56
-1
.73
0.11
-2
.38
4 -0
.20
-1.8
7 -1
.89
-3.0
6 -1
.22
4 -2
.83
1.11
-1
.74
0.47
1.
60
4 -0
.60
1.56
-0
.91
0.32
-0
.96
Big
-1
.70
-3.3
2 -1
.53
-2.6
9 -1
.46
Big
-1
.69
-1.8
3 -0
.24
-0.4
1 0.
48
Big
-1
.07
0.00
2.
26
-0.7
9 1.
72
Sm
all
26
342
311
214
62
Sm
all
21
140
169
123
32
Sm
all
2 9 30
44
45
1 2 45
40
6 35
1 26
0 56
I 2 39
25
0 28
4 23
2 65
I 2 9 31
95
11
4 59
slum
ber
of I
PO
' 3 46
49
5 36
6 22
2 60
dum
ber
of IP
O"
3 37
341
393
239
80
dum
ber
of
IPO
' 3 8 58
10
9 15
5 60
s 4 26
36
9 29
2 15
1 47
s 4 33
19
6 23
3 14
4 48
s 4 19
66
10
7 10
7 51
Big
10
10
0 95
33
9 Big
8 55
62
32
10
Big
15
33
51
74
23
-fc.
Pan
el F
: Thr
ee Y
ear
Cum
ulat
ive
Abn
orm
al R
etur
ns in
exc
ess
to N
YS
E, A
mex
, & N
asda
q
Fam
a Fr
ench
Por
tfol
ios
Sor
ted
bas
ed o
n In
form
atio
n A
vaila
ble
Sev
en M
onth
s af
ter
the
Firs
t D
ay o
f Tra
ding
Lo
w B
/Mkt
2 3 4
Hig
h B
/Mkt
Low
B/M
kt
2 3 4 H
igh
B/M
kt
Low
B/M
kt
2 3 4 H
igh
B/M
kt
Sm
all
-67.
4%
-61.
9%
-48.
2%
-52.
4%
-37.
5%
Sm
all
-17.
8%
-20.
0%
-23.
0%
-23.
3%
-22.
2%
Sm
all
-51.
0%
-71.
9%
-42.
3%
16.0
%
-10.
2%
Yo
un
g F
irm
s ( 0
to
8 ye
ars
) 2
-36.
8%
-38.
5%
-36.
4%
-41.
0%
-21.
8%
3 -2
2.6%
-1
6.4%
-2
1.3%
-4
9.5%
-7
4.4%
4 -3
.1%
-2
1.4%
-6
.8%
3.
7%
-31.
9%
Mat
ure
Fir
ms
( 9
to 4
0 ye
ars
) 2
-35.
7%
-14.
0%
-12.
3%
-2.1
%
-21.
6%
3 4.
0%
-5.5
%
6.5%
-2
1.7%
-2
8.8%
4 7.
7%
2.9%
15
.7%
18
.4%
1.
7%
Old
Fir
ms
(41
+ye
ars
)
2 -1
3.4%
17
.9%
-1
4.4%
-1
5.8%
-9
.1%
3 -3
0.5%
-1
2.6%
-6
.8%
2.
8%
-12.
5%
4 12
.7%
18
.1%
5.
0%
-6.6
%
-9.3
%
Big
32
.9%
-1
4.7%
-1
2.1%
-3
9.1%
-7
1.6%
Big
1.
2%
22.5
%
1.0%
-3
.5%
-7
.8%
Big
0.
5%
2.1%
12
.6%
5.
5%
18.3
%
Sm
all
-4.8
0 -6
.04
-3.1
0 -3
.33
-1.9
9
Sm
all
-1.1
8 -1
.69
-2.0
2 -1
.97
-1.3
7
Sm
all
-0.9
4 -2
.55
-1.8
2 1.
17
-0.7
7
2 -3
.43
-4.8
4 -3
.50
-3.9
4 -1
.10
2 -3
.08
-2.0
1 -1
.65
-0.2
7 -1
.38
2 -0
.60
1.15
-1
.69
-1.8
6 -0
.90
T-S
tatis
tics
3 -2
.55
-2.2
4 -2
.15
-4.4
9 -4
.42
T-S
tatis
tics
3 0.
45
-0.8
0 0.
92
-2.7
0 -2
.23
T-S
tatis
tics
3 -2
.40
-1.4
0 -0
.81
0.38
-1
.10
4 -0
.31
-2.3
0 -0
.62
0.29
-1
.18
4 0.
89
0.33
1.
90
1.81
0.
11
4 1.26
2.
00
0.71
-0
.79
-0.8
1
Big
1.
77
-0.7
2 -0
.44
-1.4
9 -1
.43
Big
0.
07
1.14
0.
07
-0.2
2 -0
.41
Big
0.
05
0.22
1.
31
0.68
1.
62
Sm
all
191
369
265
170
58
Sm
all
74
155
148
116
43
Sm
all
3 15
28
45
38
Num
ber
of
IPO
' 2 225
396
230
173
51
3 277
439
239
156
44
Num
ber
of
IPO
" 2 130
260
238
163
49
3 192
329
257
165
52
Num
ber
of
IPO
' 2 18
42
71
95
78
3 42
78
94
115
60
s 4 255
277
201
101
24
s 4 19
8 19
9 16
5 10
6 35
s 4 46
66
89
81
54
Big
10
6 79
44
22
6 Big
59
45
45
28
15
Big
37
46
42
60
30
43
o .E <v
a> . o M_ c o
Q- > , t- re
E " to £z c u. o « O a>
"1 o *
o
T - 00 CD CM C> SJ oi •*• in to CM r- °°
m <3> O) <D CO CM ~ 0 1 T-" . - ' <N CM' r O
O CO 00 O CM ( o ^ O OJ h - CD O n>
'•r co co co TT -
^ T - W O ( D P ) i co co co co CM m
, - • « - * - r - - i - CO
T - N CO CO ( O !
in ^t •* in Tt .
! !
i,
If
1 pi 111!
i
mim t - * - CM 00 TJ-
d s oi c\i d !
vP v p > o ^ ^ J ! h.
^ ** co o *- N- " i «( d id Tt CM" d ^ ! _
<y* O"* tf"* tf*» tf^ fi"» co CM co r- o co to
d c\i ^ ^ d in
•sP > p %p ^ 0 v P s p O^ tf* 0s" 0 s - 0 s - 0 s-
N T - S © IO r o d d d d d c-4
^~ x P **p Np %P %P \P W o1- ^ o1- t r ff" f » £ O CM CN| T - ; p CO
(/) O O O O O O
Illllll: flQ „ ; If) ID JQ- " 03
^ * - CM " ^ ^ I f t
t CO °i " - CD CO T . ^ d 2 3 co" co" S o
O CO -^f CD 00
O) r*- co co o co o 5 5 °- ^ "- ^ CD-
u CM CM CM < - CD
o s >- co n ^ ^ CO CO CM CO f - o
rr co ^t co co -
N N T - ( D ( D N eo t o co co TT T t CM
V - T - T - T - T - ^
-^ •<* co co •* co r : w co m in (D in Jj
•<t CM CM CD CM ~ CM CM CM CM CM ^
^ c o o N o n y
*HP **^ i*P ^^ ^^ »*<
O) 3*- 3s- 3"- 3"- 3"- 7a ~ C4 U> CD * - ^J- n . m d co •* CM" d ®
^ TJ- i n o •«* -«fr t o d o i c o V O N
^- 5 . ^s. ^ ^» js. co CM m in o co m
d ^ ^ ^ d "<t
CM •»- ^ - i n T T i - t o d d d d d ^
N M O <OCO I O)CM r*- in o) co ; L
5 -* o *-. «i <H ::-CO CM CM CM CM ^
' • t CO h - O O ^ t CO O) CO N CO t o
^ -* " i - -<r -<t -
oo CM co in o r^ r- co co co co *-T - T - r - * - i - 00
<° (D (D ^- «. S
n " •». t ». In ™ <° to •* - ' t^ V>
- s _, (~- ^t CM ,,. in
CM gj O CO CO g K - a, co co c" J CN :
= T - CM O) LO g f C M ^ - r^ i - - t o 9 £ CD CM 00 CO T - " .
w co" co" co" T-" ^
E o r • C/J O O .
is
o ^ d d
^asiss
' ^ CM
CO O r - CD CM o CO CO 00 CD CM t D
^ (M O ) O O ) P ) , . • * - CD CM CD CM »> * - CD CO CO * - TT
- r - CO CO _*. ^>
S N in m S a CZ *— CM r*- St ^ " * -O- •? CM <» M
CM g S jo r ; S N -CM r 2 2 CM £2
I 5 s s 8 S 5
™ t n o n c n S E CO CO - * CO CM ^
<0
I - CM i n ^ .
CM T - CO CO CM CM CO CO r^ CM i
^O ^D sW ^U ^U X " 5^ 5*- ^ - 5*- tf~- J -
^ w w o w o t ^ d T ^ CM CM T-^ CO
CO t - r - O CM Tt; ^
d d *- ^- d co
>? • ? v ^ • < x <3^ t5^ o^ c > tf* 0s"
CM t - CM TT -<r * - ; CM
d d d d d ^
~-~ *J& v o s y v v vy ^^ tO a-- c> c>- 5^ o^ tf^ g O O *-; T - T - CM
( / ) O O O O O O
- , CM r <o co CM -= ! [ i - o co i o o ™ ™ I O t - t o CO CM
-* in 5 g to co 5> g 5 S CM f° t
co co ^ £ co g>
N CO a CO t o r CO » - N CD CO i n
^ t en • * co i f co -CM CM O a > CO f O •« - • * r^ CD CM CM
CM
, « CO CO CM CD fc g CO CO O ) CM J g
Zi CO CO CO O CO " T - CM CO r - (B
CM « 2 S " S «
•sooojegsg E t 5 2 8 2 5
n CD CM co -^ o* -t;
- I X
ID N n i t "1 -
I f^ - 1 X
BJ CM to • * <0 -S
1 S--1 X
44
Table 1.7: Cumulative Excess Buy and Hold Returns by Age Group
The sample is composed of 9,400 identified IPOs from 1935 through 2002, grouped in 12 age categories. Panel A presents the three-year cumulative excess buy and hold return (annualized) relative to six value weighted benchmarks: 1) Fama and French size and book-to-market quintiles portfolio; 2) Fama and French 12 industry portfolios; 3) NYSE, Amex and Nasdaq stocks; 4) Nasdaq stocks; 5) NYSE small stocks; and, 6) S&P500 stocks. The left columns present the cumulative abnormal returns and the right columns the corresponding t-statistics. Panel B presents the five-year cumulative excess buy and hold returns. Panel C and D present three and five yer cumulative excess buy and hold returns excluding all firms that will be delisted.
Panel A: Three Year Annualized Cumulative Excess Buy and Hold Returns
Firm Age
0 1 (2-3) (4-5) (6-8) (9-12) (13-18) (19-27)
(28-40)
(41-55) (56-75) 75+ All IPO's
FF5x5 -13.9%
-13.1% -7.6% -8.5% -6.0%
-3.1% -2.0% -2.1% -2.1%
-1.2% -1.2% -0.6% -4.9%
IND12 -14.5%
-15.5%
-9.6% -7.5% -5.9%
-4.3% -3.2% -1.4%
-1.0% -0.6% -0.6% 0.3%
-5.2%
NY/A/N
-12.9%
-13.0% -7.1% -6.7%
-4.2% -1.7% -0.7% -0.2%
0.1% 0.4%
1.1% 1.8%
-3.4%
Nasdaq^ -12.8% -13.8%
-6.9% -6.9%
-6.0% -3.6% -2.1%
-0.3%
-1.9% 0.5% 3.6% 1.5%
-4.5%
Small -11.0% -10.4%
-5.3% -7.0%
-5.2% -2.4% -0.6% -0.7%
-1.7%
-0.1% 0.0% 1.3%
-3.5%
S&P500
-14.0% -14.2%
-8.6% -7.6%
-4.9%
-2.6% -1.7% -1.0% -0.4% -0.4%
0.5% 1.1%
-4.3%
FF5x5
-4.97
-7.95 -4.69
-6.99
-4.78
-2.55 -0.98 -1.67
-2.07
-1.03
-1.08
-0.60
-10.85
IND12
-6.16
-11.16
-6.91 -6.05
-4.85
-3.78 -2.06
-1.16
-1.01
-0.54
-0.55 0.27
-12.88
NY/A/N
-4.89
-8.00 -4.12
-5.26
-3.22
-1.42
-0.38 -0.17
0.13 0.31 0.89
1.69
-7.48
Nasdaq
-4.29 -8.47
-3.96 -5.41
-4.82
-2.79 -1.21 -0.20
-1.44
0.30
1.76 1.04
-9.21
Small
-3.95
-5.55 -2.54 -5.46
-4.08
-1.96 -0.27
-0.57
-1.51
-0.11 0.00
1.30
-6.98
S&P500
-5.49 -9.04 -5.35
-6.04
-3.75
-2.17
-0.92
-0.82
-0.38
-0.30 0.37
1.05
-9.71
Panel B: Five Year Annualized Cumulative Excess Buy and Hold Returns
Firm Age
0 1
(2-3) (4-5) (6-8) (9-12)
(13-18) (19-27) (28-40) (41-55) (56-75) 75+
FF5x5 -12.7%
-11.6% -7.1%
-5.8% -4.4% -1.8%
-3.1% -0.8% -1.9% -2.1% -2.7% -1.8%
IND12 -12.9%
-12.1%
-7.5% -5.1%
-3.8% -2.4%
-2.8% -0.9% -0.5% -1.1% -1.9% -0.7%
NY/A/N -11.5% -10.3%
-6.1% -3.4%
-2.2% -0.2%
-1.0% 0.9% 0.6%
-0.1% -0.6% 0.5%
Nasdaq -11.3% -12.0%
-5.8% -4.2%
-3.9% -2.0% -2.3% 1.0%
-1.4% -1.9%
-0.1% -0.9%
Small -8.6% -8.9% -6.0% -4.1% -3.3%
-1.6% -2.4% 1.2%
-2.0% -1.5% -2.2% -0.5%
S&P500 -12.6% -11.0%
-7.0% -4.1%
-2.8%
-0.8%
-1.6%
0.0% 0.3%
-0.6%
-1.0% 0.1%
FF5x5
-7.56 -12.14
-6.84
-6.61
-4.21
-1.76
-2.79
-0.33 -2.09
-2.38
-3.43
-2.59
IND12
-7.88
-13.35
-7.75
-6.13
-3.69
-2.49 -3.03
-0.57
-0.51
-1.30 -2.54
-0.96
NY/A/N
-6.67
-9.90 -5.61
-3.69
-2.14
-0.17
-0.97
0.41
0.56 -0.12 -0.64
0.67
Nasdaq
-5.05 -12.32
-5.08
-4.49
-3.79
-2.01
-2.12
0.42 -1.09
-1.59
-0.10 -0.87
Small
-3.49 -7.79
-4.78 -4.21
-2.96
-1.61 -2.08
0.41
-2.29
-1.78 -2.97
-0.71
S&P500
-7.74 -10.72
-6.81
-4.48
-2.68
-0.82
-1.55
0.00 0.26
-0.60
-1.06
0.13 All IPO's -4.2% -3.9% -2.3% -3.5% -3.1% -2.9% -11.52 -12.20 -6.41 -9.05 -7.50 -8.43
45 Panel C: Three Year Annualized Cumulative Excess Buy and Hold Returns (Excluding Delisting Firms)
Firm Age
0 1
(2-3)
(4-5)
(6-8) (9-12)
(13-18) (19-27)
(28-40) (41-55) (56-75)
75+
FF5x5 -9.6% -4.4%
0.9%
-3.8% -2.2%
-0.1% 0.5%
-0.1%
-0.9% 0.3%
-0.1% 0.8%
IND12 -10.7%
-7.1%
-1.3% -2.5%
-2.1%
-1.3% -0.7%
0.5% 0.2% 1.0% 0.5%
1.6%
NY/A/N
-9.0% -4.2%
1.5% -1.8%
-0.3%
1.4% 1.8%
1.8%
1.3% 2.0% 2.3% 3.2%
Nasdaq
-8.2%
-4.4% 2.1%
-1.6%
-2.1%
-0.3% 0.7%
2.2% -0.2% 3.0% 5.5% 3.4%
Small -6.9%
-1.3% 3.4%
-2.3%
-1.4%
0.5% 1.9%
1.3%
-0.6% 1.3% 1.1% 2.7%
S&P500 -10.3%
-5.5% -0.1% -2.8%
-0.9%
0.5% 0.9%
1.0% 0.8% 1.2% 1.6%
2.5%
FF5x5 -2.9
-2.1 0.5
-2.9
-1.7
-0.1 0.2
-0.1
-0.9 0.3
-0.1 0.8
IND12
-3.9
-4.1 -0.8 -1.8
-1.6
-1.1 -0.4
0.4
0.1 0.8 0.5 1.7
NY/A/N -3.0
-2.1 0.7
-1.3
-0.2
1.0 0.9
1.4
1.2 1.6 1.8
3.1
Nasdaq -2.3
-2.1
1.0 -1.1
-1.6 -0.2 0.4
1.5 -0.2 1.7 2.6 2.4
Small ! -2.1
-0.6 1.3
-1.7
-1.0 0.4
0.8
0.9
-0.5 1.1 0.9 2.7
S&P500
-3.5
-2.8 0.0
-2.0
-0.7
0.4
0.5
0.8
0.7 1.0
1.3 2.4
Panel D: Five Year Annualized Cumulative Excess Buy and Hold Returns (Excluding Delisting Firms)
Firm Age 0 1
(2-3) (4-5)
(6-8) (9-12) (13-18) (19-27) (28-40) (41-55) (56-75) 75+
FF5x5 -8.9% -5.8% -1.5% -2.2%
-1.6%
0.3% -1.5% 0.7%
-1.1% -0.9% -1.9%
-0.9%
IND12
-9.1% -6.3% -2.0% -1.5% -1.0%
-0.3% -1.1% 0.4% 0.3% 0.1%
-1.1% 0.2%
NY/A/N -7.6% -4.2% -0.5% 0.2%
0.6%
2.0% 0.7%
2.3% 1.4%
1.1% 0.3% 1.4%
Nasdaq -7.0%
-5.8% 0.1%
-0.5%
-1.0% 0.3%
-0.4%
2.7% -0.2% 0.0% 1.1% 0.3%
Small -4.4% -2.8% -0.4% -0.5%
-0.5% 0.5%
-0.8%
2.6% -1.2% -0.4% -1.4%
0.3%
S&P500 -8.7%
-5.0% -1.5% -0.5%
0.0% 1.3% 0.1% 1.4%
1.1% 0.6%
-0.2%
1.0%
FF5x5 -4.5 -4.9 -1.2 -2.3 -1.4
0.3 -1.2
0.3 -1.2 -1.0 -2.4
-1.3
IND12 -4.8 -5.7 -1.7 -1.7
-0.9 -0.3 -1.2
0.3 0.3 0.1
-1.5 0.3
NY/A/N -3.7
-3.3 -0.4 0.2
0.5 1.8 0.6
1.0 1.3 1.2 0.3 1.9
Nasdaq -2.6
-4.8 0.1
-0.5
-0.9 0.2
-0.3
1.1 -0.2 0.0
0.8 0.3
Small : -1.5 -2.0 -0.3 -0.5 -0.4
0.5 -0.6
0.9 -1.4 -0.4 -1.9 0.4
S&P500 -4.6 -4.0 -1.2
-0.5 0.0 1.2 0.1 0.7
0.9 0.6
-0.2
1.3
T
i
i
i
•
i -* 00 <J1 W
M
IH i
N
-» M -
—
~-
(II t« O
N
O
w
pp
pp
pp
pp
pp
pp
boboboboto^bo^'-ginLn^
C0US>IJk(OO1>l*(Jl!ll(!
pp
pp
pp
pp
pp
pp
bb
bb
io
bb
oN
Nf
fl
bi
ui
OC
D-
fc
-W
OO
WC
O-
vl
Cn
OJ
-^
O
poop^ooppooo
bb
bi
oi
pi
ob
ix
co
^u
iu
i c
oc
o^
io
o-
So
co
cn
So
oo
co
*
o 00
O
O
co o
o
-* C
n
O oo
oo
o CO
o C
O
en
o o
CD
0
0
CO
-^
1
o 00 o
o CO
o -J
o
o 8 o
C
J1
co
o o
o
o
-»
• o
o
co
co
co
O
l O
->
CD
C
O
CO
A
O
U
CO
O
CO
C
O
O
CO
C
O
O
p ^
p
p
CD
C
D
00
C
O
Ol
O
Ul
CD
o o
o
o
o
oo
^
co
b> e
n
M
ID
O
W
CO
O
O
O
p
p
bo
bo
c
o
In
cn
M
O
N
<
«
O
"0
Co
3 oo
in a.
u n
< o -<
<D
CO s
K>
4.
<J>
-
* 0
0 C
O
CO
•
A,
,•„
,\
T
i •
• i
• -
» 0
0 U
l C
O
vi
ui
o
M
a
—
o o
o
co
co
o
i en
po
p co
co
co
u
u
-»
o o
CD
C
D
CO
-»
J
o CO
C
O
o co
o CD
O
o o
o
o
o
co
co
N
s
bi
oo
o
co
co e
n
0 o
o
o
o
bo
bo
-J
. ^
en
01
O
00
-J
M
o o
o o
CO
C
O
CO
CO
-s
i -t
k
o o
co
co
co
-u
o co
co
o co
CO
o CO
o
o co
o
CD
C
D
CO
0
0
CD
C
O
00
N
CD
C
O
CD
co
oo
e
n
po
p io
co
co
S
J>
->
-J
.-
1.
00
00
00
00
0 ID
C
O
CO
C
O
O)
-»
U
^
-»
N
p o
o
o
p
bo
bo
^
bo is
en
-»
co
o
*.
ooooooooooo
tO
CO
N
C
O
N
O
U
ID
-•
-"
o o
o
o
o
o
oo
oo
-J
^
i en
C
t M
C
O
CO
I
8
•n
—•
-i 3 >
CO
<t> I z D
Z -< l z co
in
a CO
a £?
3 all v>
po
T)
en
o o
Panel >
H
3" s <D -<
O co s CD
CJ_ #-*
31
73
CD
CO
«•
* <
ID
in
P S3
£ ffl
T3
|-i A
S3
O.
T;
P ci.
P
,£>
2 en
CX
en
rt
-
•
< w
ft
^
<<
p
CD
e
n
P
D.
HI
p
O
tn
»-l
tn
ft
- •
i 8
ft
—
' **
H
.J
D
2-
C
p
£2-
&
CI
ft
r*
CO
O
""O
en
55 >
x »
d,
a <3"
o.
a
o
p H
rt
f{
P
p
<glL
g
ct
ft
en
en
O
TJ
O
S-&
<:
ft
5
r: •
• er
ft
il
O
ft
i-h
o o o
s s
^ I
ft
Bl
p
ft
-I
a.
a S
£L
O
en
Si
—
B" P
B
3
P I
ft °
14
Ul
GO
O o
13 O a
Ti
O
ft
y; a
*
ft
j-r*
^ rg
ft
qo
en
Jx
IT
^ C
5 M
O
&
. K
) fl
»
ft
O
5' g
.
H
00
fl)
as
JO
re
S"
<•
* 5'
n a- era n O
o
e T3
4^
ON
47
10% -
0% H
-10%
-20% -
^ -30% -
< ° -40% -
-50%
-60%
-70% -
-80% -
>/fit--~~ ,'yy\/>
J£sl/'^
0 1 (2-3) (4-5)
-*— FF5x5 -«— IND12 -
^&.-~ •±~-^^L
^ E ^ ^ ^ ^ ^ ^
(6-8) (9-12) (13-18) (19-27) (28-40) (41-55) (56-75) 75+
Age Group
-*— NY/Am/Nas - Nasdaq - * — Small —•—S&P500
Figure 1.2: 3-Year Cumulative Abnormal Returns by Age Group
4% -,
0%
-4%
< X m
-8%
-12%
-16% •
SE= 0
^^=^=se^
—* /// /A
1 (2-3) (4-5) (6-8) (9-12) (13-18) (19-27) (28-40) (41-55) (56-75) 75+
Age Group
—»-FF5x5 ~*-IND12 -*~NY/A/N * Nasdaq - * -Sma l l -•—S&P500
Figure 1.3: 3-Year Buy and Hold Abnormal Returns by Age Group
1.1
1.0
r K 0.8
L 0.6
0.5
-
/w^"
•S^~Y'
^ZZ&^tS--&s^\^*'
---4HC —^,JI—^** * s =*
0 1 (2-3) (4-5) (6-8) (9-12) (13-18) (19-27) (28-40) (41-55) (56-75) 75+
Age Group
—•— FF5x5 * - I N D 1 2 - * - - NY/AM Nasdaq - * — Small - •—S&P500
Figure 1.4: 3-Year Wealth Relatives by Age Group
48
Table 1.9: 3 Year Holding Period Return Distribution
The sample is composed of 9,400 identified IPOs from 1935 through 2002, grouped in 12 age categories. Panel A presents three-year holding period return percentile breakpoints, from bottom 2nd and 5th percentile points (the biggest losers) to top 95th and 98th % the star performers. The returns are measured from the first aftermarket closing price to the earlier of the three year anniversary or its CRSP delisting. Panel B presents the monthly geometric mean return breakpoints.
Pane A: 3 Year Holding Period Return Distribution Percentile Breakpoints
Firm Age 2% Pctp 5% Pctp 10% Pctp 20% Pctp 40% Pctp 60% Pctp 80% Pctp 90% Pctp 95% Pctp 98% Pctp 0 1 (2-3) (4-5) (6-8) (9-12) (13-18) (19-27) (28-40) (41-55) (56-75) 75+
-92% -96% -98% -97%
-95% -94% -93% -88% -85% -91% -81% -76%
-89% -90% -94% -93%
-91% -86% -86% -80% -74% -76% -76% -63%
-86% -82% -87% -86%
-83% -79% -76% -71% -63% -62% -61% -52%
-77% -70% -77% -75% -69% -64% -60% -55% -42% -34% -39% -25%
-46% -40% -49% -47%
-41% -28% -32% -22% -5% 2% -3% 9%
-6% - 1 % -5% 4%
7% 27% 6%
24% 38% 42% 42% 48%
44% 94%
69% 92%
97% 120% 86% 107% 88% 102% 97% 104%
88% 203% 211% 203%
207% 263% 192% 198% 160% 188% 161%
166%
218% 322% 401% 350%
369% 407% 330% 341% 246% 253% 215% 242%
871% 557% 737% 587%
584% 621% 694% 530% 389% 395% 377% 309%
Pane B: 3 Year Holding Period Return Distribution (monthly geometric mean)
Firm Age 2% Pctp 5% Pctp 10% Pctp 20% Pctp 40% Pctp 60% Pctp 80% Pctp 90% Pctp 95% Pctp 98% Pctp 0 1 (2-3) (4-5) (6-8) (9-12) (13-18) (19-27) (28-40) (41-55) (56-75) 75+
-6.74% -8.48%
-10.27% -9.11% -8.05% -7.41% -7.24% -5.74% -5.18% -6.35% -4.54% -3.85%
-5.84% -6.22% -7.43% -7.02% -6.48% -5.31% -5.31% -4.31% -3.65% -3.85% -3.90%
-2.72%
-5.26% -4.68% -5.57% -5.29% -4.73% -4.20% -3.90% -3.35% -2.75%
-2.63% -2.58% -2.01%
-3.97%
-3.29% -3.98% -3.80% -3.19%
-2.81% -2.49% -2.20% -1.53% -1.14% -1.35% -0.79%
-1.70% -1.39% -1.85% -1.75% -1.43% -0.91% -1.06% -0.67% -0.14% 0.04% -0.08% 0.25%
-0.18% -0.04% -0.15% 0.12% 0.18% 0.67% 0.15% 0.60% 0.90% 0.97% 0.97% 1.10%
1.01% 1.85% 1.47% 1.82% 1.90% 2.21% 1.74% 2.04% 1.77% 1.97% 1.90% 2.00%
1.76% 3.12% 3.20% 3.13% 3.17% 3.64% 3.02% 3.08% 2.69% 2.99% 2.70% 2.76%
3.26% 4.08% 4.58% 4.26% 4.38% 4.61% 4.14% 4.21% 3.51% 3.56% 3.24% 3.48%
6.52% 5.37%
6.08% 5.50% 5.49% 5.64% 5.92% 5.24% 4.51% 4.54%
4.43% 3.99%
Monthly Return Distribution per Age Group
2% Pctp
5% Pctp
10% Pctp
20% Pctp
40% Pctp
60% Pctp
80% Pctp
90% Pctp
95% Pctp
98% Pctp
Age Group
Figure 1.5: Monthly Return Distribution by Age Group
Tab
le 1
.10:
Dis
trib
utio
n of
Ini
tial P
ublic
Off
erin
gs p
er Y
ear
Coh
orts
and
Lis
ting
Exc
hang
e M
arke
t
The
sam
ple
is 9
,400
iden
tifie
d IP
Os
from
193
5 th
roug
h 20
02. P
anel
A, s
umm
ariz
es th
e nu
mbe
r of
IPO
s re
aliz
ed d
urin
g fi
ve y
ear
perio
ds a
nd
clas
sifie
d by
the
listin
g st
ock
exch
ange
. Pan
el B
, pre
sent
s th
e di
strib
utio
n as
a p
erce
ntag
e of
tota
l IPO
s re
aliz
ed d
urin
g th
e co
ncur
rent
per
iod.
Age
Gro
up \
Yea
r
Yo
un
g
Mat
ure
Old
NY
SE
Am
ex
Nas
daq
NY
SE
Am
ex
Nas
daq
NY
SE
Am
ex
Nas
daq
Oth
er e
xch
ang
es
To
tal
•35
- "4
0
10 - -
28 - -
17 - - .
55
Pan
el A
•41
- '4
5 6 - -11
- -17
- - . 34
,: N
umbe
r of
Initi
al P
ublic
Off
erin
g b
y A
ge
Gro
up
, Sto
ck E
xcha
nge,
and
Fiv
e-ye
ar P
erio
ds
•46
- '5
0 3 - . 59
- . 53
- . . 11
5
"51
- '55
10 - -
39 - -
44 - . -
93
•56
-'6
0
14 - -
69 - -
54 - . .
137
•61
- '6
5
19
44 -
63
52 -
68
14 - .
260
•66
- 7
0
19
55 -
51
101 .
90
60 . .
376
71
- 7
5 2 32
72
11
45
58
25
16
23 8
292
76
-'8
0 3 2
159 - 2
122 1 1
28 7
325
•81
- '8
5 4 12
720 15
11
482 10
4
160 1
1,41
9
•86
- "9
0
18
38
700 40
36
349 30
16
203 -
1,43
0
"91
- '95
61
21
1,12
8
115 17
754 95
6
170 .
2,36
7
•96
- "0
0
84
37
1,23
9
112 21
660 84
2 94 .
2,33
3
•01
- '02
11 3 52
21 3
44
25 - 5 -
164
To
tal
264
244
4,07
0
634
288
2,46
9
613
119
683 16
9,40
0
Pan
el B
: In
itial
Pub
lic O
ffer
ing
Dis
trib
utio
n b
y A
ge
Gro
up
, Sto
ck E
xch
ang
e, a
nd
Fiv
e-ye
ar P
erio
ds (
Per
cen
tag
e o
f T
ota
l Off
erin
gs p
er P
erio
d)
Ag
e G
roup
\ Y
ear
Yo
un
g
Mat
ure
Old
NY
SE
Am
ex
Nas
daq
NY
SE
Am
ex
Nas
daq
NY
SE
Am
ex
Nas
daq
Oth
er e
xch
ang
es
"35
- "4
0
18%
0%
0%
51%
0%
0%
31%
0%
0%
0%
•41
- '4
5
18%
0%
0%
32%
0%
0%
50%
0%
0%
0%
"46
- '5
0
3%
0%
0%
51%
0%
0%
46%
0%
0%
0%
•51
- '5
5
11%
0%
0%
42%
0%
0%
47%
0%
0%
0%
•56
- *6
0
10%
0%
0%
50%
0%
0%
39%
0%
0%
0%
'61
- '6
5
7%
17%
0%
24%
20%
0%
26%
5%
0%
0%
"66
- 7
0
5%
15%
0%
14%
27%
0%
24%
16%
0%
0%
71
- 7
5
1%
11%
25%
4%
15%
20%
9%
5%
8%
3%
76
- '8
0
1%
1%
49%
0%
1%
38%
0%
0%
9%
2%
•81
- *8
5
0%
1%
51%
1%
1%
34%
1%
0%
11%
0%
•86
- '9
0
1%
3%
49%
3%
3%
24%
2%
1%
14%
0%
"91
-'9
5
3%
1%
48%
5%
1%
32%
4%
0%
7%
0%
•96
- '0
0
4%
2%
53%
5%
1%
28%
4%
0%
4%
0%
"01
- '02
7%
2%
32%
13%
2%
27%
15%
0%
3%
0%
To
tal
3%
3%
43%
7%
3%
26%
7%
1%
7%
0%
Tot
al
100%
10
0%
100%
10
0%
100%
10
0%
100%
10
0%
100%
10
0%
100%
10
0%
100%
10
0%
100%
Tab
le 1
.11:
Cal
enda
r Po
rtfo
lio R
etur
ns f
or I
PO
's b
y A
ge G
roup
and
Lis
ting
Exc
hang
e
The
sam
ple
is 9
,400
iden
tifie
d IP
Os
from
193
5 th
roug
h 20
02. I
POs
that
hav
e go
ne p
ublic
ove
r th
e la
st y
ear
are
grou
ped
in p
ortfo
lios
base
d on
th
eir
age
(you
ng,
mat
ure,
old
) an
d/or
lis
ting
stoc
k ex
chan
ge,
thes
e po
rtfol
ios
are
held
for
12,
36
and
60 m
onth
per
iods
. T
he m
eans
are
co
mpu
ted
with
ove
rlapp
ing
obse
rvat
ions
, the
refo
re t
-sta
tistic
s ar
e co
mpu
ted
with
New
ey-W
est
(198
7) s
tand
ard
erro
rs w
ith a
lag
leng
th o
f one
le
ss th
an th
e ho
ldin
g pe
riod
horiz
on in
mon
ths.
All
the
retu
rns
are
repo
rted
in a
nnua
lized
term
s.
Pan
el A
: H
old
ing
Per
iod
Ret
urn
s (1
93
5 -
2002
)
Yo
un
g
Mat
ure
O
ld
All
Yo
un
g
Mat
ure
O
ld
&
NY
SE
| N
asd
aq &
Am
ex
?! A
ll
13.1
8%
13.5
8%
16.5
3%
15.3
5%
4.20
12
.12%
14
.56%
16
.28%
13
.32%
2.
82
14.2
0%
15.7
7%
19.0
1%
16.3
4%
4.26
All
5.12
5.
57
5.58
4.
20
4.67
3.
65
5.86
6.
58
5.53
Exc
ess
Ret
urn
s (N
YS
E/N
asd
aq/A
mex
) (1
93
5-2
00
2 )
Yo
un
g
Mat
ure
O
ld
All
Yo
un
g
Mat
ure
O
ld
0.12
%
1.82
%
0.08
%
2.53
%
1.77
%
3.48
%
4.31
%
4.26
%
6.21
%
3.30
%
1.28
%
3.74
%
0.04
0.02
0.68
0.93
1.03
1.83
All
1.94
1.
61
1.59
0.
49
2.91
1.
74
° N
asd
aq &
Am
ex
3 A
ll
11.8
9%
12.3
9%
13.6
4%
13.3
0%
6.02
9.
90
10.0
5 12
.03
0.39
%
1.24
%
2.41
%
1.97
%
0.20
1.
15
2.04
2.
36
14.5
6%
15.3
6%
14.9
6%
14.4
5%
4.17
7.
03
7.35
5.
97
3.72
%
4.68
%
4.19
%
3.58
%
0.97
2.
07
2.09
1.
40
13.6
5%
14.6
0%
15.8
2%
14.1
3%
5.65
10
.31
11.5
1 10
.11
2.53
%
3.41
%
4.54
%
2.61
%
1.00
2.
54
3.77
2.
02
| N
YS
E
° N
asd
aq &
Am
ex
S A
ll
11.8
9%
12.5
5%
13.0
2%
13.1
1%
7.27
12
.60
11.6
8 15
.05
13.9
7%
15.4
9%
13.8
2%
14.3
3%
4.51
6.
50
6.58
5.
81
13.5
1%
14.8
9%
14.6
9%
14.3
1%
6.20
10
.25
11.0
9 10
.09
0.85
%
1.77
%
3.87
%
6.03
%
2.81
%
4.37
%
2.24
%
3.64
%
3.99
%
2.26
%
4.39
%
3.50
%
0.43
1.
03
1.05
1.73
2.
25
2.73
1.89
2.
59
1.53
1.
51
2.78
2.
16
| N
YS
E
o
5 N iths 1 8 ths Mon
o
<0
Am
ex
All
NY
SE
Am
ex
All
NY
SE
Am
ex
All
Pan
el B
: H
old
ing
Per
iod
Ret
urn
s (1
93
5 -1
97
0)
Yo
un
g
Mat
ure
O
ld
All
Yo
un
g
Mat
ure
O
ld
All
Exc
ess
Ret
urn
s (N
YS
E/N
asd
aq/A
mex
) (1
93
5 -1
97
0 )
Yo
un
g
Mat
ure
O
ld
All
Yo
un
g
Mat
ure
O
ld
19.2
3%
18.2
6%
19.2
3%
20.8
2%
30.1
7%
22.1
9%
24.1
7%
25.6
9%
22.8
5%
18.9
1%
19.9
4%
21.9
1%
13.2
8%
14.5
9%
14.7
3%
14.9
2%
31.4
6%
23.6
4%
15.5
2%
25.2
1%
18.4
9%
15.9
2%
14.8
8%
16.3
3%
13.4
0%
14.2
5%
14.2
2%
14.4
8%
21.0
8%
19.0
2%
13.5
4%
19.5
3%
15.3
8%
15.2
3%
14.4
3%
15.4
8%
All
4.17
2.72
4.56
4.81
2.96
4.53
6.09
3.01
5.74
4.92
2.75
4.90
9.02
4.71
9.24
10.7
3
3.08
9.41
4.36
2.93
4.48
9.28
3.85
9.44
9.77
3.13
9.80
4.85
3.09
4.98
10.1
4
3.84
9.50
12.2
6
3.17
10.5
7
6.92
%
20.1
4%
10.5
6%
3.16
%
26.0
9%
9.35
%
4.13
%
17.6
4%
6.88
%
6.20
%
12.1
6%
6.86
%
3.98
%
17.5
1%
5.60
%
4.12
%
15.3
1%
5.53
%
6.17
%
14.1
3%
6.88
%
3.43
%
8.42
%
3.62
%
3.93
%
8.98
%
4.23
%
8.15
%
15.6
5%
9.25
%
3.95
%
19.2
5%
5.66
%
4.41
%
15.8
8%
5.83
%
1.88
2.02
2.59
1.24
2.39
2.24
1.61
2.55
2.19
2.31
2.00
2.47
2.92
3.66
3.72
3.34
2.67
3.41
1.92
2.44
2.12
3.48
2.17
3.67
3.45
2.36
3.58
2.54
2.43
2.80
4.54
2.95
4.13
4.75
2.72
4.09
60 Months
> > Z = 3 -<
ro co x m
u u u
» u t>
36 Months
> > Z
= 3 -< CD CO
x m
O) U Ul to co bo saw
12 Months
> > Z = 3 -< CD CO
x m
CO 30%
CO
61%
^ 34%
co ^
IO
O 00
CO O IO
CO Ul 4k. ^9
CD
CO
CO co 00
to 00 --1 ^9
O
4*. ^9
to O
to
•vl to ~9
oo Ol 4>-
CO oo
^9
00 -J
CO
4* cn
0s
to CO co 4k. ^9
o> Ol to ^9
to Ol to
^ -^1 o
CO Ol
co to
5^
CO to to
to
00 CO
5?
co o 00 ^9
^1 ^1
^9
to o 4k. CO
CO o
-8
to to _k CO ^9
to 4k. _k
-9 ©^
to Ol cn CO
4k.
CO Ol 4k.
o
CO o 00
CO CO
CO CO CO
o CO
4k. CO CO
CO CD
4*. --4
to 4k.
4k. to CO
CD cn
to Ol CD
to
IO
Ol
to to to
CO tO CO
to co o CO CO Ol
CO CO CO
60 Months
> Z = fii sd
a
JO
B» > 3 CD X
z -< CO m
36 Months
> z z 8 -< a. ffi
.Q
B» > 3 ro X
2
CO CO 4k>
CO CO O
CO IO -k _k _k _i
b) co Ik 4k. co co
CO CO CO
12 Months
> Z Z = B) -<
2. JS B> ni
> 3 ro x
cn —4 cn 85%
*» 54%
-fc. CD
84%
*• 52%
CO CD O
94%
CO 85%
—* co co
27%
CO
18%
CD
co
35%
CO
04%
A 00
16%
co 00
%
to to o
98%
to 49%
oo o Ol
51%
to 61%
4*. to
83%
O) 74%
CO
o Ol
Ol
Ol CD 4k.
cn
Ol
Ol
^i
Ol
Ol 00 CO
Ol 00
Ol o o
co
-^1 Ol 4k.
O)
CO
CO 4k. Ol
O) Ol
Ol 00 4»-
->l 4k. Ol
Ol CO
Ol co
Ol •^1
co CO 4».
Ol o O)
CO co
4k. CO CO
Ol to CD
cn CO 00
O) cn
CO CO CD
4k. CO
to
Ol
-0. to
CO CO CO
CO
cn
to Ol 00
•<
o c 3 to
3 ft c ^ ro
O E
> =
•<
o c 3
CO
2 0) !••
c -1 ro
Q Q.
>
TJ D) ne
l
p X o Q.
3 CQ
TJ CD
5' a 70 CD
c T
3 (A
. <o ~J
( IO o o IO
"*—'
Ol S 00
*• CO CD U A O
io cn ^i co co co co -v co
O) CO 4v
JO IO [O N Ul U co oi co
CO 00 Ol N) ^1 -»
co to co Ik CO 1-k CO O) O
CO IO CO
o -g
CO 4k.
©•* d*
O) CO Ul
^ *• *.
Ol IO CD >£ >8 *£
o^ d^ d*
^ Ol 00 ^
-J 00 CO Ol 00 A
CD Ol
00 IO 4k. O Ol O) -*8 %9 SP o^ e* o**
IO IO -k
CO CO to IO CD CO
CO CO -k
-k b> b) 4> CD -0.
co to co >l J M
CO CO -vl O) Ol CD
S O Ol
o Ik co 00 *k OO
oo to oi
CO 4k Ol
o oi cn -vl OO CO Ol --9 --8
to to -k b
CO IO
o o
to to Ol *•
IO IO -k
CO jk. CD
1
15%
CO
10%
CO
74%
o
71%
o to
_k
o CO
_ 4k. O
O to CO
i
16%
CO
04%
IO
09%
o 45%
o to -»i
_Jfc o oo
o
4k.
O
4k-
CO
81%
i to 12
%
• o 1
2%
i o 81
%
CO
~A
33)
3 05)
o
(es
CO
95%
—i
03%
Ol 49%
• o 91%
_k
4k. CO
O 4k CD
to 4*. CO
3 42)
• CO 85
%
o 85%
CO
06%
1
27%
•k
4*. to
o CO ->l
_ CO CO
3 57)
i CO 19
%
i CO 07
%
_» 06%
o 71%
•^
$
N3
(00
o 4k. 4>.
3 48)
i Ol 56
%
i o 05
%
Ol
51%
to 23%
_k
00 4»-
3
(30
IO o o
3 91)
• Ol 03
%
o 07%
_* 73%
to
38%
Ol ->i
o o CO
o Ol CD
3
(06 i
00 14%
i 4k. 61
%
_» 91%
to
97%
IO
o Ol
I.)
92)
o O) Ol
*Z1
58)
19
Tab
le 1
.12:
Reg
ress
ion
Coe
ffic
ient
s fo
r In
itial
Pub
lic O
ffer
ings
Thr
ee Y
ear
Ret
urns
The
sam
ple
is 9
,400
iden
tifie
d IP
Os
from
193
5 th
roug
h 20
02. T
he t
able
pre
sent
s th
e co
effi
cien
ts o
f th
e re
gres
sion
: R
etun
ii =
/3 +
Log
(l+
Age
) +
M
arke
tj +
FF5
x5; +
Ind
l2j
+ L
og(M
E)
+ B
M_R
atio
+ E
xcha
nge-
Dum
mie
s +
Ind
ustr
y-D
umm
ies
+ T
ime-
Peri
od-D
umm
ies
+ e
j. W
here
, R
etun
ij is
the
thr
ee-y
ear
raw
ret
urn,
mea
sure
d fr
om t
he f
irst
aft
erm
arke
t cl
osin
g pr
ice
to t
he e
arlie
r of
the
thre
e-ye
ar a
nniv
ersa
ry o
r its
CR
SP d
elis
ting
date
. M
arke
tj is
the
CR
SP v
alue
-wei
ghte
d m
arke
t re
turn
, F
F5x
5 ; a
nd I
ndl2
; are
the
val
ue-w
eigh
ted
mat
chin
g Fa
ma
and
Fren
ch s
ize
and
book
-to
-mar
ket
quin
tile,
and
, tw
elve
ind
ustr
y cl
assi
fica
tion
port
folio
ret
urns
res
pect
ivel
y fo
r th
e sa
me
retu
rn i
nter
val
as t
he d
epen
dant
var
iabl
e.
Log
(ME
) is
the
log
of
the
IPO
's m
arke
t va
lue
afte
r th
e fi
rst
day
of t
radi
ng,
BM
_Rat
io i
s th
e av
erag
e of
the
thr
ee S
IC d
igit
indu
stry
boo
k-to
-m
arke
t ra
tios
assi
gned
to
each
IP
O. E
xcha
nge,
Ind
ustr
y an
d T
ime
Peri
od d
umm
ies
iden
tify
the
listin
g ex
chan
ge,
indu
stry
gro
up a
nd f
ive
year
tim
e pe
riod
res
pect
ivel
y.
Reg
ress
ion
Co
effi
cien
ts f
or
3 ye
ar H
old
ing
Per
iod
IP
O R
etu
rns
Ro
-0.4
346
(8.4
4)
•0.6
077
(7.0
2)
-0.0
158
(0.0
3)
-0.1
508
(0.3
3)
-0.2
665
(0.5
0)
-0.4
746
(9.1
8)
-0.7
029
(8.1
3)
-0.3
056
(0.6
7)
-0.4
077
(0.8
9)
-0.5
425
(1.0
3)
Lo
g A
ge
0.12
47
(7.4
2)
0.10
70
(6.0
7)
0.10
26
(5.5
9)
0.11
52
(6.1
3)
0.10
68
(5.6
6)
0.12
00
(7.2
1)
0.10
20
(5.8
8)
0.09
89
(5.4
7)
0.10
70
(5.7
8)
0.09
57
(5.1
4)
NY
/A/N
0.92
26
(15.
66)
0.95
38
(15.
63)
0.98
04
(15.
88)
0.97
63
(15.
82)
1.01
43
(14.
53)
-0.4
781
(4.8
8)
-0.4
358
(4.4
2)
-0.4
074
(4.1
1)
-0.3
883
(3.8
9)
-0.4
139
(3.8
0)
FF
5x5
0.45
03
(7.5
7)
0.45
11
(7.4
4)
0.44
63
(7.3
4)
0.45
65
(7.4
7)
0.35
94
(5.5
8)
IND
12
0.86
60
(15.
50)
0.86
87
(15.
55)
0.86
66
(15.
51)
0.84
28
(14.
78)
0.91
39
(15.
63)
Lo
g M
E
0.02
55
(1.7
8)
0.02
86
(1.8
3)
0.01
46
(0.9
2)
0.02
55
(1.5
0)
0.04
03
(2.8
4)
0.04
31
(2.7
9)
0.03
24
(2.0
6)
0.05
95
(3.5
2)
BV
ME
0.11
24
(3.2
9)
0.11
07
(3.2
2)
0.14
21
(3.9
6)
0.09
02
(2.4
4)
0.08
49
(2.5
0)
0.08
44
(2.4
7)
0.10
38
(2.9
2)
0.05
05
(1.3
8)
Lis
tin
g E
xch
ang
e
NY
SE
-0.6
070
(1.3
2)
-0.4
701
(1.0
2)
-0.2
514
(0.5
4)
-0.4
169
(0.9
2)
-0.3
258
(0.7
2)
-0.2
116
(0.4
6)
Am
ex
-0.4
211
(0.9
1)
-0.2
822
(0.6
1)
-0.1
223
(0.2
6)
-0.2
525
(0.5
6)
-0.1
560
(0.3
4)
-0.0
935
(0.2
0)
Nas
daq
-0.6
235
(1.3
7)
-0.5
589
(1.2
3)
-0.2
231
(0.4
9)
-0.4
254
(0.9
5)
-0.3
814
(0.8
5)
-0.1
643
(0.3
6)
No
n-D
ur
-0.1
149
(1.3
1)
-0.1
447
(1.6
5)
-0.1
313
(1.5
2)
-0.1
727
(2.0
0)
Ind
ust
ry
B-E
qu
ip
0.22
19
(3.4
7)
0.19
91
(3.1
1)
0.17
80
(2.8
1)
0.13
91
(2.2
0)
Hea
lth
0.17
45
(2.2
5)
0.14
51
(1.8
7)
0.09
87
(1.2
9)
0.04
37
(0.5
7)
Fin
ance
0.15
35
(2.0
1)
0.15
81
(2.0
6)
0.11
64
(1.5
4)
0.08
71
(1.1
5)
'61
- "6
5
0.38
81
(1.4
0)
0.39
40
(1.4
3)
Tim
e P
erio
d
76
- '8
0
0.58
11
(2.0
9)
0.63
22
(2.3
0)
•96
- '0
0
-0.2
894
(1.1
0)
-0.2
706
(1.0
4)
'01
- '0
2
0.22
50
(0.7
7)
0.30
21
(1.0
5)
R2
0.03
02
0.03
14
0.03
23
0.0
36
6
0.04
66
0.06
19
0.06
32
0.06
39
0.06
63
0.07
54
1.6 References
53
1. Ang, Andrew, Li Gu and Yael V. Hochberg, working paper 2005, "Is IPO Underperformance a Peso Problem?"
2. Barberis, Nicholas and Ming Huang, working paper 2007, "Stocks as Lotteries: The Implications of Probability Weighting for Security Prices"
3. Brav, Alon, 2000, "Inference in Long-Horizon Event Studies: A Bayesian Approach with Application to Initial Public Offerings," Journal of Finance, 55, 5, 1979-2016.
4. Brav, Alon, Geczy, C. and Gompers, P. A., 2000, "Is the Abnormal Return Following Equity Issuances Anomalous?" Journal of Finance, 56, 2, 209-249.
5. Brav, Alon and Paul A. Gompers, 1997, "Myth or Reality? The Long-Run Underperformance of Initial Public Offerings: Evidence from Venture and Nonventure Capital-Backed Companies," Journal of Finance, 52, 5, 1791-1821.
6. Daniel, Kent, Hirshleifer D., and Subrahmanyam A., 1998, "Investor Psychology and Security Market Under- and Overreactions", Journal of Finance, 53, 6, 1839-1885.
7. Daniel, Kent, Hirshleifer D., and Subrahmanyam A., 2001, "Overconfidence, Arbitrage, and Equilibrium Asset Pricing", Journal of Finance, 56, 3, 921- 965.
8. Davis, James L., Eugene F. Fama and Kenneth R. French, 2000, "Characteristics, Covariances and Average Returns: 1929-1997," Journal of Finance, 55, 389-406.
9. Dealers' Digest Publishing Company, 1961, "Corporate Financing, 1950-1960" (Dealers' Digest Publishing Company, New York).
10. Dean, Arthur H., William Piel Jr., and Row H. Steyer, 1951, "Issuer Summaries: Securities Issues in the United States - July 26, 1933 to December31, 1949" (privately printed, New York).
11. Eckbo, B. E. and Norli, O., 2005, "Leverage, Liquidity and Long-Run IPO Returns," Journal of Corporate Finance, 11, 1-35.
12. Evans, David S. , 1987, "The relationship between firm growth, size, and age: Estimates for 100 Manufacturing Industries," Journal of Industrial Economics, 35, 4, 567-581.
13. Fama, E. and French K., 1992, "The Cross-Section of Expected Stock Returns," Journal of Finance, 47, 2, 427-466.
54
14. Fama, E. and French K., 1993, "Common Risk Factors in the Returns on Stock and Bonds," Journal of Financial Economics, 33, 1, 3-56.
15. Fama, E. and French K., 1996, "Multifactor Explanations of Asset Pricing Anomalies," Journal of Finance, 51, 1,55-84.
16. Fama, E. and French K., 2004, "New Lists: Fundamentals and Survival Rates," Journal of Financial Economics, 73, 229-269.
17. Field, Laura C. and Jonathan Karpoff, 2002, "Takeover Defenses of IPO Firms," Journal of Finance, 57,5,1857-1889.
18. Fink, Jason, Fink, K. E., Grullon, G. and Weston, J., working paper 2005, "IPO Vintage and the Rise of Idiosyncratic Risk."
19. Gompers, Paul A. and Josh Lerner, 2003, "The Really Long-Run Performance of Initial Public Offerings: The Pre-Nasdaq Evidence," Journal of Finance, 58, 4, 1355-1392.
20. Hillstrom, Roger, and Robert King, 1970, "A Decade of Corporate and International Finance: 1960-1969" (Investment Dealers Digest, New York).
21. "International Directory of Company Histories, " St. James Press, Vols. 1 to 82.
22. Jain, B. A. and Kini, O., 1994, "The Post-Issue Operating Performance of IPO Firms," Journal of Finance, 49, 5, 1699-1726.
23. Jovanovic, Boyan and Rousseau, P. L., 2001, "Why Wait? A Century of Life Before IPO," AEA Papers and Proceedigs, 91,2, 336-341.
24. Kelley, M. Etna, 1954, "The Business Founding Date Directory", Morgan & Morgan Publishers, New York.
25. Konthari, S. P. and Warner, J. B., 1997, "Measuring Long-Horizon Security Price Performance," Journal of Financial Economics, 43, 3, 301-339.
26. Lerner, Joshua, 1994, "Venture Capitalists and the Decision to Go Public," Journal of Financial Economics, 35, 293-316.
27. Loughran, Tim, and Jay R. Ritter, 1995, "The New Issues Puzzle," Journal of Finance, 50, 1, 23-51.
55
28. Loughran, Tim, and Jay R. Ritter, 2000, "Uniformly Least Powerful Tests of Market Efficiency," Journal of Financial Economics, 55, 3, 361-389.
29. Loughran, Tim, and Jay R. Ritter, 2004, "Why has IPO Underpricing Changed Over Time," Financial Management, 33, 3, 5-37.
30. Lowry, Michelle, and Schwert, W., 2002, "IPO Market Cycles, Bubbles or Sequential Learning?" Journal of Finance, 57, 3, 2002.
31. Lyon, J. D., Barber, B. M. and Tsai, C-L, 1999, "Improved Methods for Tests of Long-Run Abnormal Stock Returns," Journal of Finance, 54, 1, 165-201.
32. Moody's Industrial Manuals, various dates (Moody's Investor Services, New York).
33. Newey, Whitney K. and Kenneth D. West, 1987, "A Simple Positive Semi-definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix," Econometrica, 55, 3, 703-708.
34. Pagano, M., Fabio Panetta, and Luigi Zingales, 1998, "Why do Companies Go Public? An Empirical Analysis," Journal of Finance, 53, 1, 27-64.
35. Rajan, Raghuram and Servaes, H., 1997, "Analyst Following of Initial Public Offerings," Journal of Finance, 52, 2, 507-529.
36. Ritter, Jay R., 1991, "The Long-Run Performance of Initial Public Offerings," Journal of Finance, 46, 1, 3-27.
37. Ritter, Jay R. and Ivo Welch, 2002, "A Review of IPO Activity, Pricing, and Allocations", The Journal of Finance, 57, 4, 1795-1828.
38. Shiller, Robert J., 1990, "Speculative Prices and Popular Models," Journal of Economic Perspectives, 4, 55-65.
39. Shumway, T, 1997, "The Delisting Bias in CRSP's Data," Journal of Finance, 52, 1, 327-340.
40. Shumway, T., and Warmer, V. A., 1999, "The Delisting Bias in CRSP's Nasdaq and its Implications for the Size Effect," Journal of Finance, 54, 6, 2361-2379.
41. Teoh, S., Welch, I. and Wong, T., 1998, "Earnings Management and the Long-Run Market Performance of Initial Public Offerings," Journal of Finance, 53, 6, 1935-1974.
56
Chapter 2: Life Cycle of Public Firms: Firm Maturity and
Post IPO Performance in Fundamentals and Returns
2.1 Introduction
This paper links industrial economics to finance by exploring the effects of a
firm's age on realized returns and firm fundamentals for the first time. Using a unique
dataset with the founding and/or incorporation dates for 14,665 firms, we describe the
evolution through time (1965 -2006) of public firms as they mature in terms of their
growth potential, innovative edge, process efficiency, liquidity issues, cash flow risks,
default risk, and, profitability. We show that an equal weighted portfolio composed of
mature firms earns between 20 and 30 monthly basis points in excess to industry
portfolios and Fama & French size and value portfolios. Analyzing firms' age cycle can
help us understand the challenges that the firm overcame, but more importantly, we can
asses what its' future might be in a systematic form.
We divide the time path of a company in three stages: 1) youth, 2) maturity, and
3) old age.8 We find that in their youth firms are in a highly uncertain stage, from which
only the strongest ones survive and become mature. Innovation drives young firms; they
develop a limited number of products that are often highly original, allowing them to
differentiate themselves from well established firms. Successful young firms grow their
assets and sales quickly, about 50% to 100% faster than their industry peers. Initially,
We define mature firms as firms in their mid teen years, twenties, and up to their mid thirties.
We consider firms below 12 years as young, firms in their late teens and twenties as mature and aging firms as those older than 55 years. We define specific age groups in detail later.
57
young firms have very high Tobin's Q's on average 2.8 and as high as 7.7, which
declines to an average 1.7 as they mature and enter their early-twenties.
The desire for young firms to innovate and establish brand name along with a
tendency towards having a small market share causes these firms to incur in high
expenses resulting in low and even negative profitability, ROA is on average 6% below
comparable industry peers. At the same time, we find that on average young firms
finance themselves with a higher proportion of short term debt relative to total debt than
older firms, 34% vs. 24%. Short term obligations, along with higher illiquidity and low
return on invested capital contribute to a high delisting and default probabilities. On
average Ohlson's default is 2.8% vs. 1.3% for old firms (more than twice) and actual
annual delisting probability is 7% vs. 2%. Average credit ratings increase from BBB- to
BBB+ as firms become old, while bond yields decrease from an average 8.65 for young
firms to 8% for old ones. In addition, on average young firms require investors to sign for
a long-term investment horizon. This is to say, measured in terms of equity duration, cash
flows will be further in time than for younger firms.
Nevertheless, the story becomes sunny if young firms9 survive to maturity and
develop well established product lines, brand, and market share. They consolidate growth
opportunities and become process efficient. Their gross margins double from an average
14% to 32% as they reach their twentieth anniversary. Meanwhile, mature firms are able
to maintain an innovative edge and, having proven themselves as reliable, significantly
decrease default and delisting probabilities, while increasing dividend distribution. In
maturity, firms can be seen as being "star performers," and their investors will certainly
9 As shown in the companion paper "Firm Maturity and IPO Underperformance", about 30 percent of young IPO's never reach their fifth year anniversary before being delisted.
58
bear fruit in return space. Mature firms yield average monthly alphas between 40 and 50
basis points in excess of an standard four factor model capturing market returns, size,
value, momentum.
Alas, outperformance is not sustainable for an infinite time period. At some point
these firms will begin an aging process. Innovation and growth opportunities will drop
significantly below industry averages. The firms will start to resemble cash cows as their
cash distributions substantially increase, dividend yield for old firms is on average 3.3%
and up to 6.3% in the top 95th percentile, while mature firms have a yield of 1.1% and
young firms yield 0.7%. Overall, old firms are highly liquid and stable with very low
default probabilities and low return uncertainty (low variance), Sharpe ratios for these
firms are close to 1. After a while, as these firms age and become too old they will most
likely face the following choices: 1) re-invent themselves into a different company; 2)
partition themselves into several entities; 3) perish.
The chapter is divided as follows: Section 2.2 establishes the relevance of age and
summarizes key models developed to capture firm age and/or firms' product life-cycles.
Section 2.3 describes the dataset. Section 2.4 links firm age with expected returns.
Section 2.5 describes key fundamental characteristics and their relation with firm age.
Section 2.6 describes investment opportunities. Finally, Section 2.7 concludes.
2.2 Firm Age Models and the Relevance of Maturity
Boyan Jovanovic 1982 proposes a theory of "noisy" selection in which firms learn
about their efficiency as they operate in their industry. The efficient firms grow and
survive, while the inefficient ones fail and cease to exist. The theory partially captures
59
firm's evolution in time, along with firm size and industry concentration effects and
derives the following: 1) Small firms will grow faster and will be less likely to survive. 2)
Firm size and concentration are positively related to rates of return. 3) The correlation
over time of rates of return will be higher for big firm and those in a concentrated
industry. 4) Concentrated industries will have higher variance in their rates of return. 5)
Average profits will rise as the industry matures and firms become larger, this is a result
of the unprofitable firms leaving while profitable ones are able to stay and grow;
therefore, overall profitability increases as long as product prices do not fall.
David Evans presented the first industrial economics paper to directly study firm
dynamics linked with age in 1987. In his paper, looking at a sample of firms in
manufacturing industries, Evans finds that firm growth, growth variability, and the
probability of failure decrease with firm age. In addition to the age effect, growth
decreases with firm size as survival rate increases. Evans concludes that "firm age is an
important determinant of firm dynamics."
Steven Klepper develops a model to capture innovation over the product life-
cycle of technologically progressive industries from birth through maturity. The model
emphasizes the differences in innovation capabilities among firms as well as the
importance of firm size. Klepper's model predicts that over time product R&D of firms
will decrease while process R&D will increase. The life-cycle is described in the
following form: 1) when industries are young, firm entry will be high and firms will offer
many versions of their industries' product, innovation is very high. 2) Subsequently,
entry slows, exit overtakes entry and there is a "shakeout" in the number of producers;
product innovation and diversity of competing versions decline as "the depletion of
60
opportunities to improve the product coupled with locked-in of the dominant design leads
to a decrease in product innovation." 3) Finally, firms switch gears to improve the
production process; firms that are inefficient in the production of the dominant design
perish and efficient ones increase benefits form less competition. Agarwal and Gort
present a similar model relating firm survival probabilities and product life-cycle. They
conclude that firm survival will be dependent on both product and firm life-cycles.
Pastor and Veronesi develop a valuation model that captures investors learning
about firm's profitability over time. Their model predicts that market-to-book ratio will
increase with uncertainty about firm's average profitability, particularly for firms that do
not pay dividends. The ratio is predicted to decline over a firm's lifetime as uncertainty
declines due to investors learning. Investors attempting to value newly listed firms will be
challenged with substantial uncertainty of the firm's future profitability and growth rates.
In their model market-to-book ratio will increase with the uncertainty about book equity
growth rates due to the convex relation between the growth rate and terminal value. They
present empirical results to their model, however when estimating firm age "a crucial
variable in our empirical investigation" they follow Fama and French 2004 in which they
consider each firm as "born" in the year of its first appearance in the CRSP database.
Fama and French's proxy firm age by the company's listing date. They do not
find that firm age is reliably related to financial characteristics. However, when we
measure age by incorporation and/or founding date, we find that firm age is related to
fundamental measures and expected returns. In addition by using our firm age measure
we are able to differentiate age effects, listing year effects, and year effects.
61
Figure 2.1 shows that young firms only yield an annual expected return of 15%
equal weighted and 6.2% value weighted. As the firm matures (in its' late-teens and
twenties) expected returns climb to 20.5% equal weighed, 14.5% value weighted. After
maturity, as firms age, returns gradually decline back to 15% equal weighted returns or
12% value weighted. The return premium between mature firms and young/old firms is
on average 5.5% annualized.
When we look at Figure 2.2, the portfolios annualized standard deviation along
with Figure 2.1 causes the curvature to become puzzling. Contrary to predictions of a
classical asset pricing model following Harrison and Kreps, a portfolio of young firms
has very low returns and very high volatility. The classical model would predict that for
higher volatility we should observe proper compensation and have higher returns, which
is not the case. Young firms offer low returns although they have very high variance.
After firms reach maturity between their mid-teens and early-twenties, returns climb
while variance decreases. From this point on, returns behave consistently with a rational
pricing model. From mature to old firms, returns and volatility experience a gradual
decline with firm ageing. Potential reasons for the puzzling failure of the rational model
in which young firms do not offer compensation for their risk is explored in the following
sections.
Consistent with Pastor and Veronesi's model Figure 2.3 shows that young firms
have lower book-to-market ratios, which increase monotonically as firms mature from an
average of 0.68 for young firms to 0.94 for aging firms, the median rises 0.52 to 0.83.
The average and median ratios slightly decline from the peak when we reach the oldest
age group, it could be that after firms in these age groups we capture some firms that
62
have reinvented themselves and hence become more "growth" like, thus decreasing the
average and median ratios. It is important to note that within each age group there is a
very wide spread between growth and value stocks that ranges from 0.15 for the bottom
5th percentile up to about 2.0 for the top 95th percentile.
2.3 Data
2.3. A Measuring Firm Age
This paper is the first to relate returns and fundamental measures to an accurate
measure of firm age; previous papers such as Fama French 2004, computed firm age as
the difference between year t and the year when the firm becomes listed on CRSP for the
first time. In this paper firm age is computed as the difference between year / and the
earliest of either its founding or incorporation date. The difference can be substantial: for
example, under the Fama French 2004 age measure in June of 2005 a firm such as
Goldman Sachs that went public and was listed in CRSP in 1999 would appear to be 6
years old, and Yahoo, which went public in 1996 would appear to be 3 years older than
Goldman. In reality, Goldman was founded in 1869 making it 136 years old and Yahoo
founded in 1994, was 11 years old by 2005. In addition, under the Fama and French age
measure it is impossible to differentiate age effects from listing year effects.
Founding dates and/or incorporation dates used to compute firm age were
obtained from the compilation of mainly two datasets and several additions/modifications
done by hand: (1) The extended Jovanovich and Rousseau (2001) dataset. This dataset
was extended by Fink, Fink, Grullon and Weston (2005); the dataset contains the date of
first incorporation and/or founding date for a sample of publicly traded firms between
63
1925 and 2005. (2) The Field-Ritter dataset of company founding dates, used in Field and
Karpoff (2002) and Loughran and Ritter (2004). This dataset contains the founding dates
for 8,309 firms that went public in the U.S. during 1975-2005. All founding dates that
were not in agreement between both datasets were verified by hand. The final Founding
Dates dataset received more than 1,700 corrections and additions10 and contains
incorporation and/or founding dates for 14,665 firms listed in CRSP from 1925 to 2005.
2.3.B Sample Selection
The sample includes all NYSE-, AMEX-, and NASDAQ-listed securities with
share codes 10 or 11 (exclude REITs, ADRs, LP's, and Close-end Funds) that are
contained in the intersection of the CRSP monthly returns file, the COMPUSTAT
industrial annual file, and the Founding Dates Dataset, between July 1963 and December
2006. To avoid delisting bias and extreme negative returns from young firm IPO's as
shown in the companion paper "Firm Maturity and IPO Underperformance" in which
mainly young firms are unable to survive beyond their 3rd year post-IPO anniversary, we
restrict the sample firms to those that will be listed at least 3 years after their first listing
year on CRSP. Furthermore, firms less than 1.5 years of age are also excluded from the
sample. The final sample is composed of 141,865 firm-years; table 2.1 lists the number
of firms per year and the percentage of the total CRSP firms and market value that these
firms represent. The number of firms per year ranges from 1,360 firms in 1965 to 5,598
in 1998, and as can be observed in Figure 2.4, the sample covers at least 90% of the total
We had 846 additions and over 900 corrections/verifications. A detailed description of the sources used for all additions and corrections can be found in the companion paper "Firm Maturity and IPO Underperformance"
64
market value for every year (except during 1973 when the NASDAQ listed several
thousands of young/small firms).
To ensure that accounting information is already impounded into stock prices, we
match CRSP stock return data from July of year t to June of year t+1 with accounting
information for fiscal year ending in year t-1, as in Fama and French (1992). To be
included in the return tests, a firm must have CRSP stock price, shares outstanding and
age data for June of year t.
2.3.C Fundamental Characteristics
This paper focuses on thirteen fundamental measures which are classified in six
categories: 1) Growth; 2) Innovation; 3) Efficiency; 4) Liquidity and Cash Flow Risk; 5)
Default and Debt Structure; and 6) Profitability. Appendix A has a full description of
fundamental variables used. This paper will show the evolution of these fundamental
characteristics as firms mature, tying them together and suggesting how such
characteristics impact expected returns and can be used for investment purposes. In order
to avoid outliers and data errors influencing results, measures will be winsorized for each
year at the top/bottom 1%.
Throughout our analysis of returns and/or fundamental characteristics, we will use
ten Age portfolios in which firms are classified every year based on their current age: 1)
young firms (four or less years old); 2) five to seven years (young*); 3) eight to twelve
years; 4) thirteen to eighteen years (mature); 5) nineteen to twenty five years (mature*);
6) twenty six to thirty five years (mature**); 7) thirty six to fifty five; 8) fifty six to
65
seventy five years; 9) seventy six to one hundred years; 10) above one hundred years
(old.) Firms older than 35 years are considered to be in an aging stage.
When expected returns are analyzed spreads between mature vs. young (M-Y)
firms and mature vs. old (M-O) firms will be computed. In general, throughout the
analysis spreads between the extreme young firms (4 or less years) and the early mature
firms (13 to 18 years) will be examined. In some instances, particularly if performing a
double sort, the youngest 2 age groups may be assembled together to add to the number
of firms and add power to our results. Additionally, in some very rare instances, maturity
groups might be coupled together or spreads computed using an older maturity group,
which will be indicated with and asterisk.
2.4 Firm Maturity and the Cross-Section of Returns
2.4.A The Firm Maturity Return Spread
As shown earlier, the average expected return for these portfolios present a
humped curve, where returns are low for young firms, peak for mature firms and then
decrease as firms age. Table 2.2 presents average returns from 1965 to 2006 for the ten
age groups and the spreads between mature vs. young (M-Y) firms and mature vs. old
(M-O) firms. Panel A presents equal weighted returns, the average monthly spreads for
mature minus young firms is 45 basis points (bps) and 47 for mature minus old firms both
statistically significant, this represents an annualized return of about 5.5% on these zero
cost strategies. Furthermore, consistent with the predictions in the model by Pastor and
Veronesi, as firms mature, overall risk (return variance) decreases; the Sharpe ratio
increases almost strictly monotonically from .525 for young firms to .954 for old firms.
66
Young firms have higher maximum monthly returns but also have a higher failure rate
and a higher default probability.11 Not surprisingly, firm size increases as firms mature,
however there is presence of "sma//" firms all across age groups.
Panel B presents value weighted portfolio returns, which are lower than equal
weighted returns for all firms, however the difference is greater for young firms. As the
contribution to the portfolio's return from small firms decreases, the spread between
mature and young firms increases to 69 basis points or an annualized return of 8.24%;
Implying that big young firms tend to big underperformers. The size effect that small
firms carry over portfolio returns decreases as firms mature, the difference between equal
and value weighting is low for aging firms.
Panel C presents skewness and kurtosis for the return distribution of the ten age
portfolios. Young firms have a positive skew, which implies that some returns in the right
tail would not be expected under normality which might point to periods where investors
are overly optimistic of young firms or that some small firms experience sharp increases
in fundamental value; as firms age, skewness changes from positive to negative.
Observing the excess kurtosis, on average 2.9, all age portfolios have very high
excess kurtosis, which in fact indicates that there are several extreme months in the return
distribution that would be considered 'extremely infrequent' under normality. Kurtosis
for equal weighted portfolios increases as firms mature, implying that extreme returns are
more likely for older firms. With the intention of identifying how much the return
distribution would change if we did not have 'infrequent'' months, we eliminated the top
and bottom 1% returns for each portfolio's return distribution and report skeweness,
It is important to consider that this sample already mitigates the delisting bias for young firms by excluding all firms that will delist before their third year IPO anniversary from the sample.
67
kurtosis and Sharpe ratio for the adjusted distribution. The Sharpe ratio increases slightly
as portfolio variance decreases slightly. Overall skewness remains very similar and
excess kurtosis drops below 1.
Given that firm age is positively correlated with size and book-to-market (B-M)
ratios, an important question is whether the return curvature based on firm age (so far
described) can be explained by standard size and value (B-M) portfolio returns or if this
age effect is independent of size and value. Panel D presents average raw returns and
excess returns to size-value portfolios. Following Daniel et al. 1997, we form five size
portfolios and then subsequently within each size portfolio, divide each portfolio into
portfolios based on the firm's book to market ratio, an then subtract the corresponding
size-value portfolio's return from each firm pertaining to the size-value group. The results
show that on average older firms have negative (statistically insignificant) excess returns
to the size-value factors. Young firms underperform matched size-value benchmarks by
33 basis points (statistically insignificant); however, mature firms show a strong
outperformance between 19 and 28 monthly basis points all statistically significant.
When we take a look at the spreads for mature vs. young/old firms we find impressive
annualized excess returns of 7.33% and 3.71% respectively.
A check for robustness of the results presented so far is to group firms in age
deciles. Instead of keeping the age buckets fixed, every year we group firms in ten age
deciles each containing the same number of firms. Therefore in years prior to the creation
of the Nasdaq, when the number of young firms is low, there are fewer young firms in the
young firm decile. Panel E presents returns and excess returns for each age decile and the
resulting average age for each of the deciles. Figures 2.5 and 2.6 show previously
68
presented results hold to be true for the curvature in both returns and excess returns for
size-value portfolios. The spreads between mature firms (third bucket) and young/old
firms are still between 30 and 40 basis points. The returns for mature firms and their
spreads are a bit lower because by deciling puts some firms in their early teen years in the
youngest firm bucket (increasing that bucket's return) while at the same time the mature
buckets include firms that otherwise we would consider as being in their early aging
stages.
Next we see if the results differ by industry. Panel F presents excess returns to
industry portfolios using Fama and French 2000 industry/sector classification based on
12 industrial sectors and 48 industries. Each year we group firms based on sector/industry
and compute excess returns to sector/industry portfolio average. The return curvature
show on Figure 2.7 is consistent with our previous results for both industry and sector:
mature firms earn between 16 and 23 monthly basis points in excess to its sector/industry
portfolio. The spread between young - mature - old firms is between 7.08% and 3.63%
annualized excess to sector/industry returns.
To further examine the return curvature by industry, as well as industry age
distribution, in Panel E presents returns, Sharpe ratios, and average number of firms for
each of the twelve industrial-sectors and classifies firms in five age categories.12
Consistent with previous results the return curvature is present in all 12 sectors and the
Sharpe ratio increases as firms mature. However, some sectors have a flatter curve
(Health and Energy), while others (Business Equipment, Shops, and Manufacturing)
The number of categories was shrunk; otherwise, very few firms would remain in each bin. The categories are: I)young, 9 or less years old; 2) mature, between 10 to 20 years; 3) mature*, between 21 and 35 years; 4) between 36 and 55 years; 5) old, above 55 years. We consider two maturity ranges because some industries seem to peak later than others.
69
show a more pronounced curve. Nevertheless, the average monthly spreads are always
positive and range from 18bps to 80bps for mature minus young firms, and between
12bps and 74bps for mature minus old firms.
We look at the average number of firms per age group within each sector and
classify industrial sectors based on their tilt towards being dominated by young or old
firms. The Business Equipment sector (computers, software, etc.) is heavily dominated
by young firms. Quintiles are formed13 and it is found that average age for the youngest
and oldest quintile firms is 8 and 61 years. In contrast, Utilities' youngest quintile
average age is 30 years and oldest is 129 years: this sector is heavily tilted towards older
firms. The average sector firm age is 45 years, the youngest sectors in ascending order
are Business Equipment, Health, Telecom and Other with an average firms age of 34
years; the oldest sectors are Manufacturing, Chemicals, Non-Durables, and Utilities with
and average age of 62 years, all other sectors have average age.
So far we have showed that age return curvature persists and cannot be attributed
to either an industry effect or a size-value effect. Panel H analyses the age portfolios in 4
time periods,14 in all cases the curvature is present, mature firms between 13 and 35 years
earn a premium with respect to younger and older firms. The mature firms' excess returns
to size & value portfolios yield on average between 23 and 32 monthly bps, and excess
returns to 48-industry portfolios between 15 and 23 monthly bps, all periods have
statistical significance for mature firms.
In addition to the reported Age-Industry breakout, we also performed a sort where every year firms were quintiled according to age within their industrial sector, thus having a uniform distribution of firms. The results were very similar to reported ones. We used the reported table because it provides clear age cut points.
14 1) 1965 - 1975; 2) 1976 - 1985; 3) 1986 - 1995; 4) 1996 - 2006.
70
A simple zero cost trading strategy that buys mature and sells young/old firms
yields consistent excess returns to size & value between 30 and 53bps, and between 15
and 53bps in excess to the industry peers. However, the spreads are not statistically
significant for all periods. The zero-cost strategy would have failed significantly in some
years when young or old firms seem to get the upper hand with respect to mature ones. In
the past last two decades (1986 - 2006), as the number of listing young firms grew, the
zero cost strategy between mature and young firms became more consistent, yielding raw
and excess returns between 37 and 58 monthly bps all with statistical significance to the
99th percentile. Further investing implications of our results are discussed in Section 2.6.
2.4.B Time Series of Returns
To further examine the relation between firm maturity and stock returns, we
construct monthly time series returns for each age group portfolio and regress excess
returns (Rp - Rf) with the excess return to the market (RMKT - Rf), Fama and French size
and value factor returns (RSMB and RHML), and Carhart's momentum factor (RUMD)-
Table 3, presents alphas, factor betas, and R2 for each age group portfolio for the
following specified regressions.
Rp - Rf - a + / W • (RMKT - Rf) + eP [1]
Rp - Rf = a + /3MKT ' (RMKT - Rf) + /?SMB ' RSMB + 0HML * RHML + £p [2]
Rp - Rf = a + /3MKT ' (RMKT - Rf) + 0SMB • RSMB + 0HML • RHML + 0UMD ' RUMD + £P [3]
71
The first panel in Table 2.3 presents the resulting alphas and betas to the market
and reveals a curvature in the portfolio alphas. Alphas for young firms are statistically
insignificant from zero. Mature firms have an alpha of 58 bps, while the average old firm
alpha decreases to about 33 bps, both statistically significant. The long/short portfolio of
mature minus young firms yields an impressive alpha of 68 bps with a beta and R equal
to zero. This implies that portfolio returns are market neutral i.e. unexplained by returns
in excess to the market. Mature vs. old firms alpha is 25 bps and statistically
insignificant. At the same time, we introduce a new portfolio that longs mature firms and
shorts 50% of young and 50% of old firms; this portfolio gives an alpha of 46 bps, which
is statistically significant. As expected the portfolio betas with respect to the market
decrease monotonically from 1.36 for the young firms to 0.90 for the oldest firms.
The second panel includes in the regressions Fama and French size (SMB) and
value (HML) factors. By including these factors, the alphas of young and old firms'
portfolios become statistically insignificant; however, mature firms' alpha remains large
and statistically significant, between 24 and 29 bps. Market beta for each portfolio
approaches unity while the loading on size decreases monotonically and the loading on
value increases monotonically as firms mature. As expected, young firms load as small-
growth firms and old ones load as big-value firms. Mature firms load as mid-sized
slightly value firms and remain unexplained by either of the two factors. Furthermore, all
three zero cost strategies yield statistical significant alphas ranging from 28 to 56bps.
The third panel adds Carhart's momentum factor (UMD) in the regressions, which
increases the magnitude of mature portfolios alphas to a range between 37 and 50 bps.
The alpha of the three zero cost strategies becomes 40 bps for all three strategies. An
72
interesting result is that all portfolios load negatively on momentum (behave as losers)
and loading decreases monotonically as firms mature. As a robustness check for previous
results, in the forth panel firms are grouped in age deciles instead of fixed age groups as
previously done, and regress the age decile portfolios with the four factor model. The
results hold and yield significant alphas in the range of 42 to 50 bps for portfolios with an
average age between 11.7 and 29.9 years. The zero cost strategies have alphas between
27 bps and 39 bps, and the mature vs. young portfolio remains the hardest to explain by
the four factors, R2 and very low loadings.
Finally, we take a look into the persistence of mature firms' alpha over time. Each
year we estimate the alpha of a portfolio of firms between 13 and 25 years old {mature)
using the four factor model. Figure 2.8 shows the estimated monthly alphas for each year.
The average monthly alpha is 33bps. Alphas are positive in 30 out of 42 years (71.5%)
with increasing significance and consistency over the past two decades.
2.5 What is Age?
We have shown that over a four decade time span, mature firms present
continuous excess returns to size & value, or industry portfolio returns that range between
20 and 50 monthly bps. This section aims to characterize the effects of aging and relate
them to several firm fundamental measures which we will then relate to the return
curvature. The section is subdivided in six parts: a) Firm Growth Rates; b) Firm's
Innovation Edge; c) Firm's Process Efficiency; d) Firm's Liquidity and Cash Flow Risk;
e) Default, Delisting and Debt Structure; and f) Firm's Profitability. In most of these
73
subsections we will characterize the fundamental measure under analysis in the following
manner:
1) A descriptive panel that presents the average, standard deviation, bottom 5l
percentile, median, and top 95th percentile from 1965 to 2006 of each age-group's yearly
average of the measure. In addition, to show that results presented are not dominated by a
particular industry, every year we compute the industries' average of the fundamental
measure and subtract it from each firm within the industry; then we group the firms by
age and compute the average deviation of the fundamental measure by age group. The
seventh and eighth columns of the panel have these averages across time by industry.
Finally, the ninth column presents the average number of firms per year in each age
group.16
1 7
2) We decompose the measure's excess to its industry average into firm age ,
listing decade cohorts,18 and year effects as described in Deaton 1997. Additionally, in
our regressions as a robustness check we include the log of market value, log of book-to-
market ratio and interaction variables of these two characteristics with age. The resulting
coefficients of the decomposition19 are used to model the effects of age on the
We classify each firm in one of 48 industries based on Fama and French's 2000 industry classification, as described in Kenneth French's website.
1 The number of firms may vary slightly depending on data availability from Compustat.
To estimate the age effect we use the logarithm of age and age squared as our parameters.
18 Listing cohorts are defined based on the year in which the firm is first listed on CRSP, following Deaton 1997 the first cohort (firms listed prior to 1940) is dropped, then after we have a cohort for every decade, 1940 to 1949, 1950 to 1959,...., and 2000 to 2006. Listing cohort will capture the common effects only to the firms that are listed in that cohort and will be independent of the year effects which can affect all firms in the regression.
The decomposition used for the model is the one that does not include size or book-to-market variables.
74
fundamental measure, which are summarized in three graphs which show firm age, listing
cohort, and year effects.
3) In addition we present Fama MacBeth average estimates of yearly cross-
sectional regressions. For these regressions we also include the log of market value, log
of book-to-market ratio, interactions, and cohort effects as robustness checks. In all our
regressions the fundamental measures being analyzed carries no industry effects since our
dependant variable is the excess to its yearly industry mean.
2.5.A Firm Age and Growth
We begin with a description of a firm's evolution in terms of its growth. For this
we use three fundamental measures: a) 1 year Asset Growth; b) 1 year Sales Growth; c)
Tobin's Q proxy (Market value of Assets / Book Value of Assets). In the early years of a
firm (less than seven years old) assets grow on average at a rate of up to 30%, or 10% in
excess to the industry mean, the top 95th percentile (top growth firms) grow at more than
145% per year. As firms mature growth rates decrease monotonically to a growth rate of
10% for the oldest group or 3.2% under the industry average. The spread between top
growers and asset destructors (bottom 5th percentile) also narrows significantly as firms
mature. In youth the spread is 180%, bottom 5th percentile firms shrink at rates of up to -
34%o, while outperform growers in the top 95th percentile grow at rates of 145%. As firms
become old the spread is reduced to 45.5%; the bottom only shrinks at rates of 8.7% and
the top growers are only able to grow only at rates of 36.8%. The dispersion between
growth rates is reduced to a third as firms mature from young to old, or from a standard
deviation of 52.4% to 17.9%. All this is consistent with predictions of Pastor and
75
Veronessi's model and Evan's 1987 findings; uncertainty in growth rates shrink as firms
mature.
Similarly, when firms are young they are able to grow sales at an average rate of
33.9% or 15.6% in excess to the industry average, and as they age, sales growth drops to
9.5% or 3.6% below industry average. The spread between the bottom 5th percentiles and
top 95th percentiles also shrinks as firms mature from 158.3% to 46.2%. It is also worth
noting that asset and sales growth is stronger (above industry average) until the firm's
twentieth anniversary; afterwards it flattens and then drops to -3.5% from the industry
average. The yearly dispersion (standard deviation) of the fundamental measures also
drops significantly after the firm's twentieth anniversary. We believe results are very
robust as the sample size for each age bin and year is sufficient, each bin usually has 300
or more firm, except for the very young age groups.
The model derived form the regression estimates of asset growth in excess to the
industry suggests that extreme young firms (1 year old) grow at an average rate of 50%
above the industry mean, however this extremely fast paced growth rate quickly declines
to zero by the firms 20th anniversary and then after remains slightly below industry mean.
The cohort decomposition shows, not surprisingly, that firms listed during the 1990's
experienced an additional 8% average growth rate, which corresponds to the dot-com era
when many technology oriented firms were indeed able to increase their assets and grow
sales at very fast paces. However, in contrast to the newly listed firms of the 1990's that
grew at very fast pace, in general, all other firms during that decade grew 2% below
industry average. After the bubble burst in 2001, newly listed firms' asset growth
decreased to 6% above industry peers.
76
An interesting result is that once we take into consideration the age effects, size
has a positive effect on growth. In other words larger firms are able to grow faster than
smaller ones of the same age. Furthermore, interaction coefficients show that the size
effect dissipates in older age groups. Book-to-market as expected has a negative
coefficient implying that low book-to-market firms {growth) are indeed growing at faster
paces than value firms.
The sales growth measure exhibits very similar behavior to the asset growth
measure: young firms are able to double sales in their early years, but by the late teen
years these firms drop below the industry average growth rate and remain beneath it as
they age. Size and book-to-market contribute in the same form to sales growth, bigger
firms relative to their age group grow faster and high B-M firms experience less growth.
Another way to proxy expected growth opportunities is Tobin's Q; in equilibrium,
Tobin's Q measure should equal 1, a Q higher than one implies that profitable growth
opportunities underlie and hence the firm could invest in new projects and grow its
assets. Consistent with previous results, the Q drops as firms mature, the ratio of market
value to book value of assets decreases from 2.8 for young firms to 1.3 for older firms.
Hence, as firms mature their Q approaches unity. If we look at the top 95th Q percentiles
young firms have Q ratios of up to 7.7 market-to-book value of assets the ratio decreases
to 2.7 for old firms. The median of firms older than 55 years shows a Q that is very close
to unity, implying that growth opportunities are limited for old firms.
Taking a look at the regression estimates of the Q-measure in excess to industry
average, it is important to note the significance of the interaction between age and book-
to-market ratio. It is not surprising that low book-to-market firms will have a higher Q-
77
ratio, however it is a bit surprising that the interaction between age and B-to-M is
positive, as it implies that while keeping book-to-market constant older firms will have a
higher Q-ratio. Firm size seems to have no relevance for potential growth opportunities.
Firms listed after the 1980's, and particularly during the last decades had significant
above industry growth opportunities, while their competitors had relatively low
opportunities. However, this could be a result of the high valuations given by the market
to newly listed firms during the 1990's which fed into the construction of the Q-measure.
2.5.B Firm Age and Innovation Perspectives
It turns out that not only growth rates decrease as firms mature, but alongside the
innovation capabilities of firms decrease as well. Obviously a decline in the creative edge
of a firm would result in fewer new projects to undertake and shrinkage of growth
opportunities. It is possible that once firms establish their primary product lines
investment in new products decreases, turning their R&D focus into incremental
innovations of current products and resources towards being process efficient. The
average expense in research and development as a percentage of total assets declines
from 10% for young firms to 2.5% for the oldest age group or a decrease from 2.2%
above industry average to 1.6% below average. During the early years, firms less than
twenty years old, young and mature firms spend a median 6% of total asset value in R&D
and a maximum of 33.3%. Old firms top 95th percentile R&D spenders only go as high as
8.3%, and on the median spend 2% on R&D of total assets.
The regressions yield interesting results, as estimates predict a decline of R&D
expenditure with firm age, although this is not statistically significant. However, once we
78
interact age with size and book-to-market variables, age provides additional information
indicating that aging firms decrease R&D expenditure. As expected, smaller and 'growth'
(low book-to-market) firms tend invest more heavily on R&D. Additionally, firms listed
during and after the 1990's have a higher expense in innovation than earlier listing firms,
this effect is counter-balanced by overall under-investment in R&D during the 1990's.
To further explain what happens to firms' innovative edge as they age, we
combine our sample with the U.S. granted patents dataset compiled by Hall, Jaffe, and
Tratjenberg 2001. Their data is comprised of detailed information on almost 3 million
U.S. patents granted between January 1963 and December 1999, all citations made to
these patents between 1975 and 1999 (over 16 million), and a reasonable match of
patents to Compustat. Even though the data yields interesting results, it is important to
take into consideration that there are three biases in the dataset20 which will tend to give
stronger results for old established firms. 1) CUSIP match is based on the 1989 universe
of companies, imposing a big limit on the sample of young firms that we are able to
match. 2) The measure of Patent Originality uses the number of citations made to
previous granted patents in the sample. Therefore, in the early years originality will be
overestimated. 3) The measure of Patent Generality uses the number of citations received
by other patents. Hence, generality for patents granted during the 1990's will be biased
downwards. 4) Additionally, the sample only covers granted patents, therefore we do not
have any information on overall patent fillings that young firms may have done and either
patents were not granted (abandoned the claims) due to the firm going under.
To further understand patent variables construction and the inherited biases refer to Hall, B. H., A. B. Jaffe, and M. Tratjenberg (2001).
79
However, even after considering all the biases and the relatively small portion of
the full sample that we are able to match to the patent dataset (on average only 500 firms
are matched) and limiting our sample from 1980 to 1999,21 the results obtained shed light
on what may be occurring during the innovation process. Our results propose that young
firms tend to be granted fewer patents than older firms; however, these patents tend to be
more original and will be later referenced more frequently (higher generality). Hence, it
seems that in youth, firms invest in an original idea that will provide the base and
structure for later incremental improvements. For example, a young firm may develop a
new high tech cell phone and then as the product becomes successful file several patents
on the same core item, which protect for multiple configurations. Older firms, even
though they file and are granted a larger number of patents, often undergo more
incremental changes to already granted patents in their youth or of acquired young firms.
2.5.C Firm Age and Process Efficiency
The story described so far indicates that as firms age their innovative edge and
growth opportunities decrease. However, while growth potential decreases, firms work
hard on becoming more efficient. Efficiency dramatically increases from youth to
maturity, peaking when firms are between twenty and thirty five years old, then after
efficiency slightly decreases as aging firms become old. The average gross margin
increases from 15.3% in youth to 32% when the firms are in their twenties, then remains
steady at about 30%. However, if we consider gross margin in excess to the industry, it
increases from being 12.7% below industry average to 5.6% above average in maturity
21 We limited the sample to begin in 1980, because prior to 1975 patents do not have links to citations received which underestimates originality. Additionally, we wanted to reasonably capture young firms effects, which is limited to being matched to Compustat universe in 1989.
80
and decreases to 1.7% above industry. The regressions indicate that gross margins
increase not only with age, but also with size and B-M ratio. As firms become bigger, for
example Walmart, they gain purchasing power which allows them to have higher gross
margins. We also note that over the last two decades margins have increased for all firms,
while for firms that listed during the 21st century margins have been lower.
Our second efficiency measure proxies for firms having an advantage over the
market in terms of the margins they charge to consumers and their efficient process.
Firms with high net operating profit after tax (NOPAT) margin are not only able to
charge higher margins to their products as indicated by gross margin but also have under
control their administration process. NOP AT margins as well as gross margins increase
with maturity from 8.6% to 13.9%. Margins flatten slightly above industry average when
the firm turns 18 years. In contrast to young firms that can have negative margins of up to
11.9% for the bottom 5% underperformers, old firms' lowest margin is only 4.8%.
Margins at the top 95% performers are very similar across all age groups around 27%.
Our third measure for efficiency is invested capital turnover, fueled by the belief
that firms with high turnover can be thought of as having a production advantage, which
allows for higher sales in comparison to peers with similar capital investments. Figures
2.9 and 2.10 show curvatures of invested capital turnover which increase from 1.73 for
young firms to 2.09 for firms in their teens and twenties, which later decreases to 1.45 for
old firms. Turnover in excess to industry average is 16.1% below average for young
firms, peaks at 11.7% above average and decreases back to 16.4% below industry
average as firms become old. The efficiency curve seems to be consistent with the return
curvature, the low returns associated with young and old firms is consistent with below
81
industry average efficiency. Contrary to our previous two efficiency measures, the
regression results show that small and growth firms tend to have higher turnover ratios.
This points towards a shift in efficiency along with size and value; as firms grow in size
and become value firms, efficiency shifts from a production advantage (higher invested
capital turnover) to purchasing power advantage (gross margins) aligned with a more
competitive administrative process.
2.5.D Liquidity and Cash Flow Risk
When thinking of the return curvature, two questions arise: 1) whether the curve
is able to be explained in terms of a liquidity premium, and 2) whether it is related to cash
flow risk. To answer the first question, Amihud's illiquidity measure is used. Every year
at the end of June, the illiquidity of stocks is computed based on the prior year's trading
activity. As firm uncertainty decreases liquidity will most likely increase as more
investors trade on the stock. The first panel of Table 7 shows that, indeed as firms
mature, illiquidity decreases substantially, particularly after hitting the 35 year old mark.
In youth average illiquidity measures about 1 with a median of 0.78, our measure for
illiquidity then drops strictly monotonically on the median to 0.15 for the eldest group.
Regressions show that liquidity is extremely positively correlated with firm size
and age, consistent with previous studies smaller firms tend to be more illiquid than
larger ones. Surprisingly, when we combine age and firm size we find that the interaction
coefficient is positive; implying that among similarly sized firms, older firms will tend to
be more illiquid. At the present time, we cannot conjecture an explanation as to why
older firms would become more illiquid once we account for firm size. At the same time,
82
value firms, which tend to be older and better established, tend to be more illiquid. The
interaction between value and age shows that liquidity increases for older firms once
value is taken into account. This suggests that "young" value firms are illiquid when in a
state of distress, while older value firms are well established cash cows with good
liquidity. Another unexpected result is that since 1996 stocks in general have become
more illiquid, particularly due to the fact that during the late 1990's and post-2003 the
stock market was highly liquid; at the same time, overall new listing firms tend to have
higher levels of illiquidity.
The relation between firm age and liquidity is explored further by sorting stocks
into three liquidity groups and age groups. As described in the literature, low liquidity
firms command higher returns and have significantly lower market value than their peers.
High liquidity firms' average market value is 1 billion dollars for young firms, 2 billion
for mature, and 6.2 billion for old firms. In contrast, low liquidity stocks average market
value ranges from 30 to 40 million dollars among young and old firms.
Mature firms command a premium on top of their liquidity group for all three
liquidity groups. Figure 2.11 shows the return curvature is present in all liquidity groups
for mature firms between 13 and 35 years old. The age effect is more predominant for
low and medium liquidity groups where the percentage of young and mature firms is
higher compared to their age peers with high liquidity. Across all liquidity groups mature
firms are the only group that has significantly positive excess returns to Fama and French
size and book-to-market portfolios; excess returns for mature firms range between 20 and
51 basis points. Consistent with our previous results, Sharpe ratios increase
monotonically with firm age regardless of liquidity constraints. Older firms have lower
83
uncertainty and higher return-volatility payout; however, within age groups Shape ratios
are higher for low liquidity stocks.
Unexplainably, young firms that belong to high and medium liquidity groups are
significantly punished by the market, earning raw returns between -30 and 65 bps per
month or between 49 and 120 bps below comparable size & value stocks. It may be that
overall these medium-high liquidity young firms create a great amount of hype causing
significant trade volumes and high market valuations; however when these firms fail to
establish product lines their returns plummet. As liquidity decreases, excess returns to
size and value portfolios significantly increase for mature firms. In the most illiquid
tercile, mature firms yield between 28 and 51 monthly bps in excess to size and value.
Low liquidity young and old firms have high returns relative to other liquidity returns but
returns are below their liquidity group average. Even though the age effect is clearly
more pronounced for low liquidity firms, returns are consistently higher for mature firms
across liquidity groups, which indicate that age has relevance on top of liquidity.
In addition to liquidity, we look into trading frequency: the percentage of non-
traded days is on average 19.7% for young firms and decreases monotonically to only
5.5% for the oldest firms. The median of non-trading days decreases from 15.2% to zero,
and even for the most non-traded stocks (top 95 th percentile) young firms have twice the
percentage 39.6% vs. 21.1% for old non-traded stocks. Regression results show that non-
traded days decrease as firms age and as they become larger. Consistent with liquidity
results, it seems that in the last decade overall firms non-traded days have increased by
6%, this is a counterintuitive result since we would expect that as markets became more
integrated and moved to electronic platforms, overall trading would have increased. We
84
observe post-Nasdaq introduction in 1972, there was a sharp decline in the percentage of
non-traded days to 5% below industry average; afterwards, the percentage of average
non-traded increased gradually to 6% above industry average.
Finally, we use equity duration as proposed by Dechow, Sloan, and Soliman, to
proxy for the time that it would take cash flows to pay back investors. A higher duration
implies that it will take more years to achieve a pay back and therefore, investors should
have a longer investment horizon in mind. If high duration is accompanied with high
default probabilities there would be a very high risk that future cash flows would not be
realized, hence we can also think of equity duration as a proxy for cash flow risk. Not
surprisingly, duration is higher for young firms on average 17 and decreases as firms
mature to 14.5. For the top 95th percentile, a young firm's duration is up to 23 years,
compared to 17 for old firms. Therefore, firm maturity lowers the risk of cash flows not
being received and reduces the investment horizon. A firm's duration drops below
industry average after the firm's twenty-fifth anniversary. As we would expect,
regressions show that duration is significantly negatively correlated to both size and
value: bigger firms will offer a shorter investment horizon and growth stocks will have
higher duration or longer investment horizons. We also observe that firms listed during
and after the 1990's have higher durations, most likely because young technology
oriented firms require longer investment horizons. However, we did not find a direct link
of equity duration to the return curvature.
85
2.5.E Default and Debt Structure
Up to this point, we have found that young firms offer great growth opportunities
and are innovation driven, while mature firms retain some growth opportunities as they
become more efficient and old firms seem to offer security in terms of higher liquidity
and lower equity duration. In order to identify a source of observed underperformance of
young firms, we examine default probability estimated using Ohlson's 1980 default
measure. Table 2.8 shows the probability of default for young companies is on average
2.8% and climbs up to 7.7% for the top 95th percentile of riskier firms. As firms become
older the probability of default shrinks in half, or on average 1.3% and the probability of
risky firms decreases to only 2.7%. After the firm's eighteenth anniversary, average
default probabilities fall below the industry mean.
Regression results show that once size and book-to-market are controlled, default
probability will decrease based on both variables but will increase with firm age. This
implies that younger, big value firms will have lower default probability than older firms
of the same size and similar book-to-market ratio, which is a bit counterintuitive.
Additionally, it is found that default probability has been on the rise for firms listing
during the 1980's and peaked for firms listed in the 1990's. Default could be one of the
underlying factors that explain low returns of young firms, investors could underestimate
the probability of such firms perishing. Particularly the failure of highly liquid young
firms may be underestimated by investors, while the returns of young "star performers"
are not high enough to compensate for the vast number of failing youths that default and
cease to exist.
86
Actual delisting is further inspected by forming a table that tracks firms delisting
in time. Firms are grouped according to initial age group when the firm is first listed,
surviving firms are classified in the subsequent age group into three categories using
CRSP delisting codes, survivals, delisted because of a merger or acquisition (codes 200),
and delisted because of death or dropped by unexplained reasons (codes 400 and 500).
The delisting probability is scaled within each age group by year number in which the
delisting occurs; thus, if the delisting is in the first year the annual probability is 1, if it
occurs in year 5 the annual probability is going to be 0.2, presenting annualized delisting
probability. By scaling we avoid overestimating the probabilities for older age groups
that have several more years within their age bucket than younger age buckets.
Delisting probabilities show that young firms have a higher probability of being
delisted in their earlier initial years ranging between 7.1% and 10.5% annual probability
for firms under 18 years. This probability substantially decreases to range between 2.0%
to 3.8% for firms over 55 years old. The table shows that overall annualized probability is
indifferent to whether a firm has been listed longer or not in the stock market. In sum,
firm's age matters the most. While delisting probability decreases with age, merging
probability is very similar across age groups, the lowest probability is for young firms
and ranges from 5.2% to 8.5%, as firms mature merging probability increases to range
between 7.5% and 10%, or on average about 9%; these results are consistent with
previous results regarding IPO survival obtained in a companion paper.
A firm's debt structure is further examined by computing the percentage of long
term debt to total debt. It is found that on average young firms have lower percentage of
long term debt compared to older firms, 66% vs. 75%. Young firms' long term debt
87
access might be relatively limited and hence have a lower percentage of long term debt in
relation to overall debt structure. At the bottom 5 th percentile, young firms have only 10%
of long term debt, which implies that these firms are financing themselves almost solely
on short term debt which expires in less than one year. In comparison, the bottom 5th
percentile of old firms have between 25% and 33% of long term debt. The higher
percentage of short term debt in young firms explains at least in part the higher estimated
default probability. However, unreported regression results show that age is secondary to
firm size and book-to-market ratios and age coefficients become statistically significant
only when age is interacted with these characteristics.
FISD bond data was used to obtain ratings and bond yields which was matched to
our firms using the following procedure: 1) According to industry and debt-to-capital
ratio we matched our firms to similar firms rated bonds; 2) Converted the ratings into
numeric values (See Appendix B); 3) Used the weighed average of bond yields with
similar ratings. Results only cover the time period from 1985 to 2006 because prior to
1985 bond data is very limited. Corporate yields decrease as firms mature from an
average of 8.63% to 8.02% and from a median of 8.7% to 7.7%. The top 95th and bottom
5 percentiles show that high and low yields are very similar for all age groups.
Conversely, as yields decrease with maturity, debt ratings improve with maturity.
Old firms, on average, have better credit ratings than younger firms, which is consistent
with older firms having lower default rates. The median rating for young firms is 6.86
which is equivalent to a BBB- rating and increase to 7.4 for old firms, equivalent to a
BBB+ rating. At the top 95th percentile all age groups seem to have a rating of 8.2, or
equivalent to an A+ rating, likewise at the bottom 5th percentile all age groups ratings are
88
about 5.4 which is equivalent to B+ rating. Additionally, every year we estimate the cost
of equity using a three factor model (Mkt-Rf, SMB, and HML) with sixty months of prior
returns, and compute the weighted average cost of capital (WACC). The cost of capital
decreases with age from 13.7% to 12%, the decrease is slightly less significant when the
median is observed, which drops from 11.5% to 10.5%. The bottom and top 5th
percentiles show that in all age groups there are firms with very low and very high cost of
capital: the spread is as low as 5.6% for old firms and as large as 25.9% for young firms.
2.5.F Profitability and Cash Flow Distribution
Finally, the analysis is rounded out by examining profitability and cash flow
distribution. Results in Table 2.9 show that young firms have on average - 1 % ROA, or
6% below industry average, which increases to an average ROA of 8.1% or 1.3% above
the industry mean. During youth, a firm is likely to present losses, as market share is
small and expenses to develop an innovative product and create a brand are high. Young
firms ROA in the bottom 5th percentile can be as low as -48%, compared with only - 1 %
for old underperformers. However, as firms mature, and conditional on surviving their
youth stage, returns on assets increase as prior investments bear fruit. As firms mature
ROA will be above industry average as long as firms remain competitive. The upside for
the top 95' percentile performers is significantly higher for young firms which have an
ROA of 24% while top old performers only yield a return of 17%. When regression
estimates are examined further we find that ROA increases not only with age but also as
firms become bigger and as book-to-market ratios increase. The decomposition shows
89
that ROA for firms listed during and post the 1990's declined significantly below
industry peers.
In addition to analyzing profitability in terms of ROA, cash flow distributions to
shareholders are measured using firm's dividend yield. The yield differential among
young and old firms increases monotonically both on the average and the median from
0.7% to 3.3% and 0.2% to 3.1% respectively. Old firms big dividend distributors (top 95th
percentile) give as much as 6.4% of market value in dividends per year, event old firms in
the bottom 5l percentile will give at least 0.6%. When firms reach their thirty-sixth
birthday, average dividend distribution is above the industry mean. Regressions show that
dividend yield increases with size and book-to-market ratio as well. Dividend yield has
been declining through the decades for newly listed firms, hitting a low for firms listed
during the 1980's. This is consistent with general observations of newly listed stocks that
tend to offer low dividend yields in order to grow by reinvesting cash flows in new
projects, an example of this is Microsoft.
2.6 Firm Age and Investment Opportunities
Up to this point mature firms have been shown to posses excess returns to both
industry portfolios and Fama and French size and book-to-market portfolios. Monthly
excess returns for firms between 13 and 35 years old range from 17 to 28 basis points.
Furthermore, when the intersection between liquidity groups and firm age is examined
we find that, consistent with the literature, illiquid stocks carry a significant premium
over liquid stocks, and this premium is accentuated by high returns (an average 2% per
month) pertaining to mature illiquid firms.
90
Table 2.10 categorizes firms according to the level of concentration in their
particular industry and firm age. We follow Hou and Robinson 2006 and compute the
Herfindahl index of a firm's assets and sales as a percentage of total industry assets and
sales, a three-digit SIC number is used to classify the industries. A higher Herfindahl
indicates a more concentrated industry and similarly a low index indicates a highly
fragmented industry. The industry Herfindahl is averaged over three years to avoid one
year errors influencing results. Hou and Robinson describe concentrated industries as
"innovation-poor, profit-rich industries with high barriers to entry" and find that on
average low concentration industries have higher returns that highly concentrated
industries. Within our sample a significant spread between high concentrated and low
concentrated industries is not found, however excess returns to Fama and French
portfolios is 10 bps statistically significant for low concentrated industries.
Industry sales concentration and firm age portfolios show the maturity effect is
persistent for all levels of concentration, the spread between mature and young/old is
between 33 to 58 monthly basis points, or between 24 and 58 bps in excess to Fama and
French portfolios. The spread for concentrated industries when measured using assets is
slightly higher, between 40 and 64 bps on raw returns. Interestingly, returns are higher
for young firms in low concentration industries: after firms reach maturity and begin the
aging process (after their thirty-fifth birthday) returns increase for firms in concentrated
industries.
Furthermore, young and mature firms in highly competitive industries (low
concentration) are on average bigger than age peers in highly concentrated industries. It is
probable that young firms competing in a very concentrated industry will incur a very
91
high costs to establish themselves as competitors and will have a high probability of
failure because the barriers to entry are very high, hence we observe low returns for these
firms. In comparison young firms in a fragmented industry will have a higher probability
of success and this is reflected in overall returns and market value. As firms mature and
establish product lines as well as improve efficiency, the level of industry concentration
is no longer relevant since these firms reap the fruits of competitive products, an efficient
process, and continue to exploit growth opportunities.
When a firm begins the aging process it is better off in a concentrated industry
where the cost of entry is higher, allowing it to maintain substantial market share even if
products are no longer as competitive. Therefore, aging/old firms in medium to highly
concentrated industries perceive slightly higher returns than fellow aging firms in low
concentration industries. Consistent with the predictions of Jovanovic's model, firms in
highly concentrated industries have higher variance in their returns, however returns are
not higher for bigger firms in a concentrated industry as suggested by Jovanovic 1982,
but rather for mature firms enjoying the fruits of establishing successful products, while
being efficient and retaining some growth opportunities.
In Table 2.11 intersection of Invested Capital Turnover (efficiency) and firm age
is pinpointed. Every June stocks are sorted into Turnover terciles and firm age. On
average, efficient firms (high turnover) earn higher returns 1.59% vs. 1.18% compared to
low efficiency firms. The raw return difference between mature vs. young/old firms'
portfolios is between 26 and 49 basis points, or between 18 and 38 basis points in excess
to Fama and French. Furthermore, mature efficient firms have significantly higher returns
than firms with low turnover, excess returns for Fama and French portfolios are between
92
21 and 41 bps. Consistent with previous regression results, average firm size of high
turnover firms is smaller than that of low turnover firms. The fact that inefficient mature
firms have higher rates of return than comparable inefficient firms that are either
younger/older, may be because mature firms are less prone to delisting while being able
to profit on established product lines and growth opportunities; however, not necessarily
in the most efficient way.
Firms are sorted even further using the difference between return on invested
capital (ROIC) and the weighted average cost of capital (WACC), and firm age.
Shareholder value is created if firms are able to generate returns on investment in excess
to their cost of capital. Hence, a positive spread between ROIC and WACC is a form of
"Value Creation" while a negative spread depicts "Value Destruction." ROIC by is
estimated by multiplying Net Operating Profit After Tax (NOPAT) and Invested Capital
Turnover.
Table 2.12 sorts stocks into three "Value Creation" buckets and firm age every
June and examines portfolios average monthly returns. Surprisingly, overall returns
within "Value Creation" remain fairly constant across terciles. However, only the high
"Value Creation" tercile has positive significant excess returns to Fama and French
portfolios, 22 basis points. The average spread between ROIC and WACC for low Low
value creators (actually "value destructors") is -20%, young firms destruct at rates up to -
29.3%. As firms mature, average destruction declines monotonically to -10.69% for old
firms. Medium value creators earn an average spread of 1% and the average spread of
high value creators is 17.60% which is as high as 20.45% for young and mature firms and
declines to 15% for old firms.
93
Results become even more interesting when looking at young firms (seven or less
years old). In such an instance, high value creators earn high returns, 1.63% monthly raw
return or 44 basis points in excess to Fama and French portfolios. In contrast, mid-value
creators and value destructors earn a low raw return of .94% or -.39 basis points in excess
of Fama and French portfolios. The spread between young value creators and value
destructors is a raw 70 monthly bps of 83 bps in excess of Fama and French portfolios. It
appears that value destruction affects young firms the most because such firms are not
strong enough to survive an inability to cover their cost of capital; hence, they will likely
default and/or be delisted. It is important to take into consideration that these firms have a
very high value destruction rate of-26.8%.
Once firms move towards maturity and progress in age further, returns for value
creators vs. value destructors are not that different. Moreover, in some cases, it would
even seem that value destructors earn higher returns. However upon taking a closer look
three key aspects can be identified: 1) Value creators are the only firms to earn significant
excess returns to Fama and French portfolios, in contrast young and old value destructors
have significant negative excess returns. 2) Sharpe ratios are consistently higher for value
creators than value destructors across all age groups. 3) Value creator firms have
consistently higher average market values. Mature and aging firms that are value
destructors may be experiencing one very bad year where they experience big losses and
write downs. However, they are strong enough to withstand such a year and
simultaneously rebound the following year showing good returns for the holding period.
Combining all results to this point can lead to multiple investing strategies. In the
simplest form a portfolio manager would overweight mature firms vs. young/old ones
94
when forming a portfolio. Furthermore, the relation of firm age portfolio returns to
industry, size, book-to-market, efficiency, default/delisting probabilities, industry
concentration, and liquidity constraints, would allow investors to form long-short
strategies enhancing traditional factors that carry premiums such as size and value.
A more ambitious project would be to incorporate all previously defined
fundamental variables into a systematic valuation approach: for example combining age,
size and value to estimate a firm's sales growth rate, gross margins, default probability,
etc. Table 2.13 simply illustrates the benefits of incorporating sales growth estimates in a
systematic form to re-estimate equity duration. Instead of merely assuming sales growth
to mean reversal over the next ten years as suggested by Dechow et al. when computing
equity duration, sales growth rate estimates are used (having been computed based on
firm's age and industry group).
Results yield a larger spread on average equity duration, which decreases
monotonically from 35.85 for young firms to 13.87 for old ones, compared to the
previous equity duration spread that was between 17 and 14.6. The dramatic increase in
equity duration for young firms is not surprising as this is linked with high growth rates,
and implies that for young firms duration will tend to be higher since the horizon in
which an investor receives cash flows is farther apart. As expected our decomposition
shows equity duration decreases with size (bigger firms also tend to pay higher dividend
yields which would shorten the duration). Finally, "new " equity duration increases with
book-to-market ratio, this could be due to "new" duration capturing distresses effects
from value firms.
95
2.7 Conclusions
Throughout this paper we showed that firm age is relevant to understand firm's
evolution in time and changes in its fundamentals as well as expected returns. As firms
mature returns have a curvature that classical finance models are not be able to explain.
Age portfolios yield lower returns when firms are in their youth, returns rise as firms
mature and then gradually decline as they begin their aging process, while volatility is
high for young firms and gradually declines with aging. Over four decades mature firms
have persistent excess returns to size & value, or industry portfolio returns that range
between 20 and 30 monthly basis points.
In youth firm's present very high uncertainty, their efforts are geared towards
product innovation, developing few highly original products, and growth. Successful
young firms can grow sales and assets at rates of 50% or more. Tobin's Q is on average
2.8 and as high as 7.7 for young firms. Growth comes with a high cost, young firms incur
in high expenses to innovate resulting in low and even negative profitability an average
ROA of 6% below comparable industry peers. Young firm's higher illiquidity, low return
on invested capital, low debt rating, and longer investment horizon, contribute to high
probabilities of delisting and default.
In Maturity firm's consolidate growth opportunities and become process efficient
doubling their gross margins and significantly decreasing their default and delisting
probabilities. As firms begin an aging process, innovation and growth opportunities will
drop significantly below industry averages. Firms will start to resemble cash cows as
their cash distributions substantially increase, dividend yield for old firms is on average
3.3% and up to 6.3%, while their default probability drops significantly.
96
We take a look at the puzzling low returns for young firms, and find that young
firms competing in highly concentrated industries seem to experience lower returns than
young ones competing in fragmented industries where they are able to have bigger
market share. Returns of young firms that are "value destructors" (ROIC - WACC) are
significantly low -39 basis points below comparable size and value portfolios, compared
to 44 basis points in excess for successful young "value creators," an impressive monthly
spread of 84 basis points. Furthermore, medium and high liquidity young firms are the
biggest underperformers by far, between -49 and -120 monthly basis points in excess to
size and value portfolios. It seems that the low performance of young firms can be
attributed to hyped firms (highly liquid) that fail to meet market expectations (unable to
create value above their cost of capital) and that are unable to establish market share
when competing in concentrated industries. This will result in high levels of delisting and
default by these young firms hampering the overall performance of the group.
Finally, we conclude that firm age has deep relevance to produce better forms of
modeling firms in terms of expected sales growth, margins, etc. Traditional valuation
mechanisms try to forecast firms' future cash flows, analysts look at past growth
realizations and extrapolate from all their information future growth rates. However,
analysts usually don't take into consideration the firm's growth potential based on the
firm's maturity, and do not follow a systematic way to estimate growth. Modeling firms
based on their maturity and industry could provide a systematic approach for better
valuations as exemplified in our re-estimated equity duration measure.
97
Graph 1. Annualized Firm Age Portfolios Returns
(ST) (8-E) (B-fl) (&2S) (2635) (36-55) (56-75) (76-UO) «©D
Avg BW Return * Avg VW Return
Graph 2. Annualized Firm Age Portfolios Standard Deviation
(5-7) (8-E) (Q-B) (B-25) (2&35) (36-55) (56-75) (76-00) *C0
— B / V Pbrt Vol — * — V W Fort Vol
Figures 2.1 and 2.2, Annualized Firm Age Portfolio's Returns and Standard Deviations
Graph 3. Firms Book-to-Market Ratio by Firm Age
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100) +100
Firm Age
—•—Average x Median
Figure 2 3 Firms Book-to-Market Ratio by Firm Age
Graph 4. Firm Sample as a Percentage of Total CRSP Firms
90%
1965 1970 1975 1
• % of Total CRSP Firms - » - % of Total Market Value
Figure 2.4 Firm Sample as a Percentage of Total CRSP Firms
Table 2.1: Firm Sample and Summary Statistics
The sample includes all NYSE-, AMEX-, and NASDAQ-listed securities with share codes 10 or 11 that are contained in the intersection of the CRSP monthly returns file, the COMPUSTAT industrial annual file, and the Founding Dates Dataset, between July 1965 and December 2006. We restrict the sample firms to those that will be listed at least 3 years after their first listing year on CRSP, and exclude firms under 1.5 years old. First column identifies year t which starts in July t and finishes in June of t+1; the second computes the number of firms in the sample; the third computes the total market value of the firms in the sample; finally, the fourth and fifth columns compute the percentage of CRSP firms and percentage of market value covered by our sample.
Year
1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Firms in
Sample 1,360 1,375 1,411 1,415 1,439 1,535 1,581 1,634 1,690 2,891 2,870 2,900 2,903 2,803 2,735 2,678 2,671 2,817 2,832 3,192 3,353 3,410 3,710 3,893 3,884 3,863 3,882 4,006 4,260 4,661 4,971 5.195 5,549 5,598 5,417 5,246 5,033 4,763 4,482 4,221 4,004 3,762
Sample ME
447,275 465,767 518,988 588,447 585,360 445,346 641,453 714,663 680,054 614,751 711,893 787,934 796,921 792,818 869,242 975,777
1,166,045 955,280
1,557,687 1,357,149 1,723,369 2,240,514 2,627,923 2,347,225 2,625,506 2,874,141 2,997,809 3,462,893 4,061,595 4,119,323 5,125,803 6,467,899 8,413,336 11,033,217 13,158,496 14,902,041 13,022,912 10,749,133 10,664,529 12,733,354 13,355,879 13,982,362
% of Total
CRSP Firms 66.21% 66.17% 67.51% 67.80% 66.90% 67.92% 68.32% 66.86% 32.41% 60.27% 62.61% 62.74% 62.14% 61.85% 60.85% 59.38% 55.11% 56.26% 55.49% 55.61% 59.19% 59.42% 60.31% 63.37% 66.07% 67.02% 68.95% 69.26% 70.96% 70.71% 74.29% 73.04% 75.20% 77.17% 80.79% 80.62% 84.83% 88.12% 89.96% 87.65% 84.15% 80.14%
% of Total
Market Value 93.79% 93.19% 93.03% 92.21% 91.58% 93.61% 93.56% 92.90% 83.92% 92.92% 92.68% 92.88% 92.34% 91.69% 91.70% 91.00% 88.92% 89.20% 88.59% 86.50% 89.84% 89.88% 91.05% 91.86% 93.36% 94.03% 93.99% 93.72% 93.24% 92.53% 93.41% 92.24% 93.82% 95.01% 95.61% 94.18% 97.00% 97.12% 96.60% 95.84% 94.71% 93.38%
99
Table 2.2: Expected monthly returns by age group and industry
The sample covers on average 3000 firms per year from 1965 to 2006. Firms are grouped in ten age categories based on the firm's age (current year - founding/incorporation date). Panels A and B present average monthly returns for equal and value weighted portfolios, ^-statistics, standard deviations, maximum and minimum returns, annualized returns, Sharpe ratio (excess return to unit of risk), average number of firms within each group and average firm size. The panel also presents returns to a simple long/short strategy which longs mature firms (13 to 25 years old) and shorts young or old firms. Panel C presents higher moments for both equal and value weighted portfolios, skewness, kurtosis, and an adjusted measure (which excludes the top/bottom 1% of the distribution) for skewness, kurtosis and Sharpe ratio computed with the adjusted average returns and standard deviations. Panel D shows excess returns to size and book-to-market quintile portfolios computed following Daniel and et al 1997. As a robustness check Panel E presents raw and excess returns for age deciled portfolios, instead of using the standard age breakpoints defined in this paper, every year we group firms in deciles based on their age. The first column of Panel E presents the average firm age per decile, all other columns follow previous definitions. Panel F presents industry excess returns to 12 and 48 industry portfolios, firms are classified using the sic classification found in Kenneth French's website. Panel G explores further industry returns by grouping firms in 12 industrial sectors and subsequently classifies each firm in five age groups (we shrinked the number of age groups to allow sufficient firms in young and early maturity firms). The Panel presents average return, ^-statistic, Sharpe ratio, and average number of firms. Finally, Panel H breaks the sample in four time periods and age groups, and presents raw returns, Excess return to Fama and French portfolios (FF Excess), Excess returns to 48 Industry portfolios (Ind Excess), all with their corresponding t-statistics.
Panel A: Equal Weighted Portfolios
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100 Mature - Young
Mature - Old All Firms
Avg Return
1.26% 1.45% 1.52% 1.71% 1.62% 1.55% 1.42% 1.27% 1.24% 1.24% 0.45% 0.47% 1.44%
t
3.36 4.00 4.40 5.31 5.50 5.81 6.06 5.97 6.13 6.14 2.24 2.45 5.56
SD
0.083 0.081 0.077 0.072 0.066 0.059 0.052 0.048 0.045 0.045 0.044 0.043 0.058
Max
44.27% 38.73% 33.27% 35.23% 30.51% 29.41% 27.84% 25.69% 25.64% 21.60% 22.26% 25.90% 29.10%
Min
-30.47% -31.99% -30.94% -30.71% -29.60% -28.64% -27.40% -24.67% -24.25% -23.16% -42.86% -13.36% -28.20%
Annulized
Return 15.13% 17.41% 18.22% 20.49% 19.45% 18.58% 17.08% 15.28% 14.89% 14.89% 5.38% 5.60% 17.22%
Sharpe
Ratio 0.525 0.620 0.683 0.824 0.854 0.902 0.941 0.927 0.952 0.954 0.350 0.380 0.863
Avg # of
Firms 95 189 388 428 365 352 477 376 326 283 521 711
3,278
Avg Size
232 215 227 318 576 711 711
1,388 1,822 3,960 272
2,139 1,005
Panel B: Value Weighted Portfolios
Firm Age
4<= (5-7) (8-12)
(13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100 Mature - Young
Mature - Old All Firms
Avg Return
0.51% 0.99% 0.90% 1.06% 1.20% 1.08% 1.02% 0.93% 0.92% 1.03% 0.69% 0.17% 097%
t
1.39 2.70 269 345 4.06 4.13 4.62 4.65 5.01 5.48 2.92 0.90 4.78
SD
0.082 0.082 0.075 0.068 0.066 0.058 0.049 0.045 0.041 0.042 0.052 0.043 0.045
Max
28.62% 29.70% 25.83% 21.52% 24.11% 21.57% 19.77% 18.52% 16.39% 19.94% 22.54% 24.65% 17.20%
Min
-36.89% -31.18% -31.32% -29.86% -26.19% -25.87% -22.99% -21.94% -21.69% -19.49% -25.11% -19.33% -22.00%
Annulized Return 6.17% 11.94% 10.85% 12.70% 14.44% 12.99% 12.18% 11.19% 11.00% 12.37% 8.24% 2.07% 11.61%
Sharpe
Ratio 0.217 0.419 0.417 0.535 0.630 0.641 0.717 0.722 0.778 0.850 0.457 0.140 0.742
Avg # of
Firms 95 189 388 428 365 352 477 376 326 283 455 648
3,278
Avg Size
232 215 227 318 576 711 711
1.388 1.822 3,960 398
2,268 1,005
Panel C: Higher Moments of Equal & Value Weighted Portfolios
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mature - Young Mature - Old
All Firms
Skew
0.42
0.28 0.13 -0.06
-0.06 -0.23 -0.36
-0.33 -0.34 -0.34 -2.14 0.89 -0.20
Equal Weighted Port
Kurt
2.52
2.13 2.14 2.17 2.37 2.70 3.55
3.66 4.05 2.74
20.78 4.01 2.90
Adj-Skew
0.23
0.13 0.05 -0.07 -0.12 -0.17 -0.35 -0.39 -0.38 -0.24 -0.75 0.51 -0.21
Adj-Kurt
0.82 0.39 0.99 0.41
0.68 0.65 0.84 1.00 0.71
0.50 3.24
1.12 0.49
Adj-SR
0.54
0.64 0.70 0.86 0.88 0.95 0.99 0.97 1.01 1.01 0.43 0.38 0.91
Skew
-0.29 -0.15 -0.32 -0.41 -0.24
-0.36 -0.21 -0.41 -0.31 -0.15 -0.17 0.12 -0.35
Value Weighted
Kurt
2.20 2.06 2.00 1.20 1.31
1.61 1.56 2.15 2.55 2.03 2.84 2.57 1.91
Adj-Skew
-0.05 -0.03 -0.32 -0.26 -0.20 -0.24 -0.08
-0.29 -0.03 -0.09 0.16 0.15 -0.20
Port
Adj-Kurt
0.61 1.17 1.34 0.29 0.62 0.46 0.08
0.79 0.54 0.43 0.57
0.43 0.44
Adj-SR
0.25 0.44
0.42 0.56 0.65 0.68 0.76 0.75 0.83 0.89 0.41
0.02 0.78
Panel D: Equal Weighted Portfolios - Excess Returns to Size and B/M
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mature - Young Mature • Old
All Firms
Avg Return
1.01% 1.29% 1.46% 1.67% 1.61% 1.58% 1.42% 1.24% 1.23% 1.22% 0.66% 0.44% 1.39%
t
2.57 3.43 4.14 5.15 5.45 5.89 6.03 5.84 6.09 6.16 3.01 2.25 5.32
SD
0.088 0.084 0.079 0.072 0.066 0.060 0.052 0.048 0.045 0.044 0.049 0.044 0.058
Excess
Return
-0.33% -0.04% 0 . 1 1 % 0.28% 0.25% 0.19% 0.04% - 0 . 1 1 % -0.08% -0.03% 0 . 6 1 % 0 . 3 1 % 0.06%
* r. „ r Annulized t - Excess SD Excess Excess -1.64 -0.31 1.23 4.44 5.06 4.10 0.82 -1.76 -1.20 -0.49 2.84 2.82 5.61
0.046 0.028 0.019 0.014 0.011 0.010 0.011 0.014 0.016 0.015 0.048 0.024 0.002
-4.02% -0.48% 1.29% 3 .31% 2.96% 2.23% 0.48% -1.28% -1.00% -0.40% 7.33% 3 .71% 0.67%
Avg # of
Firms
100 193 384 419 353 337 443 347 305 265 519 684
3,146
Avg Size
248 218 231 316 579 724 735
1,458 1,904 4,093 282
2,204 1,022
Panel E: Equal Weighted Age Deciles - Excess Returns to Size and B/M
Average Firm Age Avg Return
6.3 11.7 17.5 23.6 29.9 38.6 48.9 63.5 82.9
122.S Mature - Young
Mature - Old All Firms
1.19% 1.51% 1.59% 1.50% 1.52% 1.41% 1.44% 1.29% 1.20% 1.23% 0.40% 0.37% 1.39%
t
3.25 4.32 4.92 5.13 5.40 5.67 6.17 5.96 5.96 6.21 3.60 1.77 5.32
SD
0.082 0.078 0.072 0.065 0.063 0.055 0.052 0.048 0.045 0.044 0.025 0.046 0.058
Sharpe Ratio 0.505 0.671 0.764 0.796 0.839 0.880 0.958 0.925 0.925 0.963 0.558 0.275 0.825
Excess Return -0.12% 0.18% 0.26% 0.17% 0.14% 0.06% 0.09% -0.07% -0.11% -0.03% 0.38% 0.30% 0.06%
Excess - Excess
Sharpe -1.10 2.07 3.91 3.45 3.09 1.47 1.94
-1.17 -1.61 -0.48 3.67 2.42 5.61
-0.171 0.322 0.607 0.536 0.479 0.228 0.301 -0.182 -0.250 -0.075 0.570 0.347 0.376
Annulized Excess -1.43% 2.12% 3.16% 2.05% 1.67% 0.71% 1.06% -0.81% -1.30% -0.39% 4.58% 3.54% 0.67%
Avg # of
Firms 310 311 312 309 318 318 314 316 319 319 622 631
3,146
Avg Size
231 227 262 410 469 925 810
1,185 1,769 3,847 246
2,054 1,022
Panel F: Equal Weighted Age Deciles - Excess Returns to Industry
Average Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mature - Young Mature - Old
; Excess
Return -0.38% -0.11% 0.01% 0.21% 0.17% 0.17% 0.04% -0.08% -0.11% -0.09% 0.59% 0.30%
12 Industrial Sectors
-1.78 -0.88 0.10 3.24 3.42 3.35 1.00
-1.55 -1.80 -1.20 2.78 2.43
Excess Annulized
Sharpe -0.276 -0.137 0.016 0.503 0.531 0.521 0.154 -0.241 -0.280 -0.186 0.431 0.378
Excess -4.53% -1.38% 0 12% 2.56% 2.06% 2.07% 0.54% -1.01% -1.35% -1.08% 7.08% 3.63%
Excess
Return -0.33% -0.12% 0.01% 0.23% 0.16% 0.16% 0.04% -0.09% -0.11% -0.11% 0.57% 0.35%
48 Industries
Excess
-1.62 -0.99 0.12 3.81 3.50 3.59 0.91 -1.75 -1.81 -1.63 2.78 3.03
Sharoe -0.252 -0.154 0.018 0.591 0.543 0.558 0.141 -0.272 -0.281 -0.253 0.431 0.471
Annulized
Excess -4.02% -1.46% 0.12% 2.80% 1.89% 1.95% 0.50% -1.09% -1.27% -1.35% 6.82% 4.15%
Avg # of
Firms 100 193 384 419 353 337 443 347 305 265 519 684
Avg Size
248 218 231 316 579 724 735
1,458 1,904 4,093 282
2,204
101
Panel G: Equal Weighted Portfolios, by Industry
Non-Dur*
Durables
Manufac
Energy
Chem*
Buss Equip
Telecom*
Utilities
Shops*
Health
Finance*
Other
Firm Age
Avg Return f
Sharpe Ratio Avg # of Firms
Avg Return f
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return f
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
Avg Return t
Sharpe Ratio Avg # of Firms
9<=
0.89% 2.25
0.3499 21
0.58% 7.76
0.1807 9
0.84% 2.26
0.3514 25
1.45% 3.31
0.5142 19
1.05% 1.89
0.2938 6
1.25% 2.49
0.3861 103
1.46% 2.70
0.4240 19
0.92% 7.69
0.3376 4
0.99% 2.49
0.3862 45
1.42% 3.22
0.5122 71
1.07% 3.66
0.5675 51
1.28% 3.27
0.4982 79
(10-20)
1.26% 4.11
0.6375 40
1.37% 3.29
0.5163 13
1.64% 4.65
0.7224 58
1.64% 4.24
0.6586 28
1.27% 3.54
0.5500 12
1.90% 4.39
0.6811 200
1.64% 4.29
0.6652 21
1.53% 3.97
0.6070 4
1.44% 4.63
0.7183 89
1.66% 4.31
0.6780 103
1.46% 4.95
0.7690 90
1.68% 5.12
0.7948 110
(21-35)
1.37% 4.89
0.7597 45
1.41% 3.20
0.4967 14
1.60% 5.37
0.8332 68
1.73% 4.71
0.7317 26
1.70% 6.14
0.9525 15
1.76% 4.74
0.7355 119
1.95% 5.76
0.8215 9
1.15% 5.04
0.7823 10
1.64% 5.57
0.8651 77
1.83% 5.53
0.8585 41
1.69% 6.08
0.9675 69
1.57% 5.45
0.8458 84
(36-55)
1.21% 5.10
0.7915 55
1.60% 5.22
0.8097 17
1.43% 5.37
0.8237 80
1.44% 4.67
0.7252 24
1.40% 5.30
0.8229 17
1.65% 4.74
0.7351 46
1.29% 3.98
0.6179 5
1.02% 6.04
0.9372 22
1.41% 5.36
0.8319 66
1.69% 5.52
0.8565 11
1.45% 6.04
0.9378 44
1.48% 5.71
0.8857 57
+56
1.25% 5.83
0.9057 157
1.19% 4.49
0.6973 32
1.23% 5.08
0.7883 188
1.32% 4.96
0.7694 32
1.19% 5.44
0.8452 39
1.28% 4.50
0.6988 36
1.21% 5.73
0.7957 11
0.99% 5.92
0.9195 97
1.22% 5.07
0.7869 83
1.36% 6.09
0.9460 22
1.37% 5.97
0.9169 143
1.25% 5.22
0.8107 75
Mature -Younq
0.48% 7.84
0.2861 66
0.72% 7.75
0.2742 23
0.80% 3.84
0.5966 83
0.18% 0.66
0.1027 48
0.65% 7.35
0.2094 20
0.65% 2.98
0.4622 303
0.54% 7.23
0.1983 29
0.79% 7.75
0.2297 9
0.66% 3.03
0.4706 123
0.33% 7.36
0.2165 176
0.48% 2.32
0.3688 122
0.39% 7.88
0.2911 189
Mature -Old
0.12% 0.86
0.1339 202
0.23% 0.87
0.1266 46
0.41% 2.73
0.3311 246
0.31% 7.32
0.2046 61
0.51% 2.87
0.4359 54
0.62% 2.62
0.4063 236
0.74% 2.49
0.3963 21
0.54% 7.50
0.2331 101
0.42% 3.77
0.4828 161
0.30% 7.07
0.1688 126
0.28% 1.67
0.2660 218
0.43% 2.50
0.3885 185
All Firms
1.23% 5.23
0.8117 318
1.28% 4.40
0.6825 86
1.35% 5.77
0.7940 419
1.49% 4.68
0.7268 129
1.30% 5.56
0.8630 88
1.68% 4.26
0.6606 504
1.58% 4.65
0.7219 64
1.03% 6.23
0.9674 136
1.36% 4.97
0.7617 362
1.58% 5.00
0.7760 242
1.40% 5.89
0.9144 394
1.45% 5.76
0.8005 404
Pan
el H
; E
qu
al W
eig
hte
d P
ort
folio
s, b
y T
ime
Per
iod
s
Firm
Age
1965
-197
5 A
vg R
etur
n
f S
harp
e R
atio
FF
Exc
ess
Ret
urn
f-
F
F E
xces
s In
d E
xces
s R
etur
n
f-
Ind
Exc
ess
Avg
# o
f Fi
rms
1976
-198
5 A
vg R
etur
n
t S
harp
e R
atio
FF
Exc
ess
Ret
urn
r-
F
F E
xces
s In
d E
xces
s R
etur
n
(-
Ind
Exc
ess
Avg
# o
f F
irms
1986
-199
5 A
vg R
etur
n
t S
harp
e R
atio
FF
Exc
ess
Ret
urn
t-
F
F E
xces
s In
d E
xces
s R
etur
n
(-
Ind
Exc
ess
Avg
# o
f Fi
rms
1996
-200
6 A
vg R
etu
rn
t S
harp
e R
atio
FF
Exc
ess
Ret
urn
r-
F
F E
xces
s In
d E
xces
s R
etur
n
f-
Ind
Exc
ess
Avg
# o
f Fi
rms
7<=
0.91
%
10
6 0.
3352
-0
.06%
-0
.26
-0.0
2%
-0.0
9 58
1.73
%
2.38
0.
7569
-0
.18%
-0
.65
-0.1
4%
-0.4
7 15
8
0.73
%
1.27
0.
4045
-0
.10%
-0
.78
-0.3
8%
-2.7
5 44
4
1.25
%
1.60
0.
4720
-0
.23%
-0
.81
-0.2
4%
-10
6 48
1
(8-1
2)
1.07
%
1.31
0.
4168
0.
14%
0.
77
0.19
%
0.89
76
1.90
%
2.95
0.
9358
-0
.02%
-0
.15
-0.1
2%
-0.9
2 24
2
1.11
%
19
4 0.
6174
0.
20%
1.
61
-0.0
8%
-0.6
0 50
4
1.64
%
2.18
0.
6431
0.
12%
0.
56
0.06
%
0.30
66
9
(13-
18)
1.24
%
15
5 0.
4929
0.
27%
1
62
0.38
%
19
3 10
3
2.19
%
3.58
1.
1364
0.
23%
1
99
0.19
%
18
4 27
7
1.31
%
2.41
0.
7655
0.
28%
2.
44
0.15
%
18
3 42
1
1.79
%
2.87
0.
8473
0.
30%
3.
01
0.20
%
2.46
81
1
(19-
25)
1.14
%
1.61
0.
5123
0.
31%
2.
42
0.23
%
16
5 10
0
2.24
%
3.84
1.
2184
0.
32%
3.
28
0.20
%
2.32
25
4
1.33
%
2.62
0.
8309
0.
24%
2.
81
0.15
%
2.44
42
5
1.63
%
2.90
0.
8546
0.
14%
1
69
0.07
%
1.10
59
6
(26-
35)
1.13
%
16
3 0.
5178
0.
29%
2.
36
0.22
%
1.91
10
4
2.07
%
3.98
1.
2631
0.
08%
1
21
0.15
%
2.49
30
3
1.28
%
2.72
0.
8640
0.
12%
1
84
0.09
%
13
0 36
6
1.72
%
3.75
1.
1052
0.
23%
2.
44
0.18
%
17
3 54
1
(36-
55)
0.80
%
13
4 0.
4261
-0
.02%
-0
.31
-0.0
5%
-1.0
1 32
0
1.96
%
4.19
1.
3301
0.
06%
1
06
0.06
%
1.10
43
8
1.30
%
3.05
0.
9688
0.
11%
1
39
0.14
%
17
5 41
9
1.52
%
3.83
1.
1306
0.
01%
0.
05
0.02
%
0.17
57
4
(56-
75)
0.64
%
12
2 0.
3880
-0
.11%
-1
.41
-0.1
4%
-18
2 26
2
1.87
%
4.27
1.
3572
0.
02%
0.
27
-0.0
1%
-0.1
2 38
4
1.15
%
2.84
0.
9006
-0
.10%
-1
25
-0.0
3%
-0.3
2 37
8
1.26
%
3.63
1.
0704
-0
.20%
-1
.10
-0.1
6%
-1.1
5 36
0
+76
0.66
%
13
6 0.
4328
-0
.03%
-0
.36
-0.1
2%
-10
0 29
2
1.73
%
4.29
1.
3630
0.
01%
0.
20
-0.1
6%
-16
3 54
2
1.21
%
3.03
0.
9627
-0
.04%
-0
.64
0.01
%
0.07
62
0
1.24
%
3.95
1.
1662
-0
.16%
-0
.73
-0.1
5%
-0.9
8 79
1
Mat
ure
-Y
oung
0.34
%
12
3 0.
3894
0.
33%
1.
22
0.41
%
16
3 16
1
0.46
%
13
2 0.
4185
0.
41%
1
32
0.33
%
11
0 43
5
0.58
%
3.53
1.
1203
0.
37%
2.
52
0.53
%
3.48
86
5
0.53
%
2.10
0.
6202
0.
53%
2.
05
0.44
%
2.16
1,
292
Mat
ure
-O
ld
0.58
%
14
2 0.
4514
0.
30%
1
41
0.50
%
1.71
39
4
0.47
%
16
1 0.
5111
0.
22%
1
49
0.34
%
1.92
81
9
0.10
%
0.40
0.
1276
0.
31%
2.
05
0.15
%
0.99
1,
041
0.55
%
11
7 0.
3450
0.
46%
1.
54
0.35
%
15
8 1,
602
All
Firm
s
0.83
%
1.38
0.
4366
1,31
5
1.92
%
3.84
1.
2182
2,59
9
1.18
%
2.49
0.
7900
3,57
6
1.51
%
2.96
0.
8734
4,82
3
103
20%
19%
• 18%
% 17%
"g 16%
• 15%
< 14%
13%
12%
Graph 5. Firm Age Decile Portfolios
6 3 11.7 17.5 23.6 29.9 38 6 48.9
Average D e d l e Age
63.5 82.9 122.5
Graph 6. Firm Age Deciles Returns in Excess to Size and B/M
4 % -
3%
E 2% -
f 1% 3 M
- i %
/ \ / V_^
6.3 11.7 17 5 23 6 29.9 38 6 48 9 63.5 82.9
Average Decile Age
122.5
Figures 2.5 and 2.6, Firm Age Decile Portfolio Returns Raw and in Excess to Size and Book-Market
0.30% -I
0.20%
0.10% -
0.00% •
-0.10% -
-0.20%
-0.30% -
Graph 7. Monthly Excess Return to Industry
/'S^.. / "«=:==J'S.
••^/- ^ s ^
gr >•**_ _., —-•
/ 4
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100) +100
• 12 Ind Sectors 48 Industries
Figure 2.7 Firm Age Portfolio Excess to Industry Returns
o
u
©
u
(J ss i-e « fa
•o
a C
5 c/5
d 3 <u u
I E
3
tON
T-
Nc
tioiN
me
oio
c\|r-e
qo
>*-;o
>r"-
:r--Tj;c
D
SJ
int-
oin
ox
Mo
oin
io "
mi
no
no
io
iN
in
oo
96
^r
ir
io
ri
ri
ri
o
SO
OO
Or
NI
DN
r s
flO'-
mo
ii-in
MO
iM
•*m<
oc
oc
or^
r'~r-r~
oo
ddddddddod
T-
r-
T-
T-
v-
T-
T-
OO
O
om
*if
fliD
fl0
iDn
on
T-
ni
i)
U)
in
i;
no
n o
od
oo
oo
oo
.834
o 00
r~-o co
o o d
•* o d •
ss 0
0 co d O) oun
>-• ature
S
.377
o m
0)
CO
T_
CO
f
-T
—
d
(N
Tf
d
ss C
O
CM
d
•o
O • C
D Matur
915
iri
CM
1
^ C
M
CO
to C
O
O
d
O)
d
S?
CO
T
f
d fN
O
+ >-iture -
3
<2 (O
s
<N
IO
00
^c
o'2
S^
&I
S n
no
cv
is
(£
!o
c!
9c
!9
•-> .
V^
^T
-^
CM
CN
CN
I
mo
-3
-m
co
co
co
o>
CM
O)
CM
rt
O^
vS
cO
MO
T-
••
tf
CM
CN
cr
id
r^
co
cd
dc
M
•r-CM
CO
CO
CO
CO
CO
CM
CM
v-
CD
to
t-
CO 0
0 t-
CM CO
If) CM T
- Tf
CD
od
cD
Tfc
oo
du
oin
^-c
dr^
i-
Nf
OW
in
oM
DI
D
fepJCM
r-CD
OT
fOj-^
99
00
rN
SN
'!
r '
•O
'-
NC
O'
IO
VO
OO
nO
JN
O^
-N
CO
CO
Nr
cor~-00O)O)O>o)O)O)O)
ddddddddod
Q2S(DO)MCMCl)C»CO
OOO-r-T-cOTj-tr-cl-. Tf
oooddddddd
OT
-o
co
oo
o5
oo
ir)
T-
r^
ol
DN
OO
IO
KO
n
T-C
MC
Or-O
OC
DtJ
-CO
CM
r
-'
T-
^T
^^
T^
od
dd
o
Of
Ol
rl
On
Ol
MI
KI
l ^
T-
0*
-0
00
)0
)0
)0
)
o o o o
<DOC0O)00'"J-t-XPjr-
°dddddd
00d
C2
:
*»A
££
£&
S*s
CO
NO
c\ ^
T
0) C
T)
o>
•*
CD C
M
CM "
-
O
CD
T-
O
CO
O)
d
•*
t^
CM
CM C
O
CO
O)
IT)
T-
O
O
O
r~ C
O
od
d
m
N
IO
O)
P"
<
C
™ to
o
CD
LO
C
O
m to
o)
o
O)
•q-
od
d
CD CO
C
M
in N
«
od
d
is
ii
<o
cp
r"-cp
cp
r^q
>0
)N.(0
r N
(1
xt
b'-'c
six
tirJT
frrT^T
-^^
©O
OC
OO
>0
)0)0
)01
0)0
) d
do
'do
'do
ci
oo
co*-co
^rcMr^co
com
o)
O'
OO
O'
do
'o
'o
'o
'o
'
999ddddddd
Tr
or
-c
No
oo
oo
-"
-• —
-
o> N
. Co co
'
d ci d
d '
*- CM
CO
T
- .
00
00
00
0)
0)
0)
0)
^T
^T
^T
^t
^r
^d
dd
d
a? £
* £
a? 5
s a! I?
§? a? X&
a? •
CO CM
©
! d
d
dd
d
*si!!Hii? If)
(O
CN
CM
N-
H
ID
d s d o 00
*""
CO
CM
o d CO
•*r d r—
C
M
CM
d CO
W
o d O)
o d &s
^ d O)
c 3 o
>-ature
S
O)
CO
i O)
(0
<?'
in in
CM
00 CO
o>
m
CO
o CM
h
-
d TT
9
O)
cS r-•
*
d
o CO
T
CM
m
O)
^ CO
in
o CO d T
t
o d 5
o 9
9
CO
C
O
o>
d CM
o T
-
d 5? o T
T
d •o
O >
0) Matur
in
O)
-* d
o 6 5? T
-•«r
d CM
a O
+ >; ture -n
E
l
era e -i
ft bo 3 «-*
B
1 re
I £ fD
w o s o s N° n c re
sa
3 a o 3 3 £"
3
Est
imat
ed M
onth
ly A
lpha
-l o l-S
-
o
—1
—
o o
—u~
.
o
.,—
I.L
.
—*
o -s
—1
—
_k
. £ —
1—
Nj
O
—1
—
rO
—1
—
w
o 5s
—i
-<
(D
At
Mature - Young Mature - Old
Mature - (Y + 0)/2
0.27% 0.39% 0.33%
-0.105 0.047 -0.029
-0.015 0.947 0.466
0.098 -0.453 -0.177
0.164 -0.112 0.026
0.142 0.683 0.487 O
) w
ro
C
O
1A
4^.
w
•&•
-«i
-3.96 1.57 -1.44
-0.44 24.95 18.09
2.45 -10.11 -5.84
635 -3.85 1 33
KM
WS
JM
IO
U^
I^
?1
! <
oln
t©a
)co
b>
ui:>
it**|
oo
oo
oo
oo
oo
o
o
o
o
o
o •
to
o o
o
-*
00
00
00
0-
'-
-*
-*
ro
•&-
i •f
r.O
C._
--
-_
. _
-
(J
1N
JO
U(
D0
0(
0r
O0
)'
oo
oo
oo
oo
06
'~~
~ '"
S N
ti?
iJ v
cf
l S
I S
_ C
Df
OO
JO
.
66
66
66
66
66
b'
j,
al
*l
»U
uf
OM
U
(o-t
-tc
ja>
&o
oo
G)0
>
O)
J£
»-
»-
'NU
lOlC
0<
O
00
0
00
00
00
0 (O
<_)
(D
(_)
(O
(O
<D
(D
00
00
W0
)0
1U
W^
-»
O0
D'
_
Ol£
-*C
)C-C
OU
lfJ
lO)
Mb
ib
bb
uib
uc
n
(D
(D
-^
rO
00
00
O>
(O
-
<» o
i b
u
b
i a
b
'*»
N w
-
i!
D-
»S
Ul
O-
»0
)0
i^S
!_
KU
„?
°*
N>
P-
' 2
_„
88
--
_;
66
66
66
6A
)6
; c
oc
nb
ob
ot
oc
jco
^p
;
8
<T
o
106
Table 2.4: Firm Age and Growth Rates
The table examines the relation of firm age with respect to three growth measures: a) 1 year Asset Growth, b) 1 year Sales Growth, c) Tobin's Q proxy (market value of Assets / Book Value of Assets). For each measure we present two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5th and top 95"1 percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to industry which is computed in the following form: 1) compute the industry average following Fama and French 2000, 48 industry classification; 2) subtract the industry average at the firm level from the raw measure; 3) group the firms by age and compute the average and median of the of the excess to the industry measure (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the measure's excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. Finally, for each measure we present graphs of the measure's decomposition into age, listing cohort, and year effect. As a robustness check in all our regressions we add the firms size (log(ME)) and book-to-market ratio (log(BVME)) and interact those variables with the age (log(Age)) variable. T statistics are reported under the coefficient estimates in parenthesis.
Asset Growth
Firm Age
4<= (5-7) (8-12)
(13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mean
0.295 0.276 0.215 0.197 0.154 0.142 0.119 0.109 0.101 0.098
SD
0.524 0.506 0.428 0.377 0.311 0.283 0.241 0.216 0.196 0.179
Bottom 5%
-0.230 -0.294 -0.271 -0.233 -0.192 -0.164 -0.135 -0.116 -0.099 -0.087
Med
0.155 0.147 0.111 0.119 0.096 0.091 0.081 0.075 0.071 0.070
Top 95%
1.452 1.454 1.095 0.933 0.704 0.619 0.481 0.441 0.385 0.368
Mean Excess 0.105 0.099 0.044 0.030 -0.006 -0.010 -0.022 -0.026 -0.032 -0.032
Med Excess -0.019 -0.017 -0.041 -0.036 -0.053 -0.050 -0.051 -0.048 -0.052 -0.051
Avg # of Firms
45 146 345 400 349 334 445 346 304 265
Asset Growth Age, Listing Cohorts, & Year Effects Decomposition
Intercept
-0.0672 (3.78)
-0.1497 (7-12)
-0.6617 (26.40)
Log (Age)
08540 (16.58; 0.6431 (12.52) 1.3729 (24.47)
Log (Age2)
-0.4228 ( (7 .22; -0.3293 (13.46) -0.6216 (23.71)
Log (ME)
0.0334 (49.54) 0.1211 (48.15)
Log (BVME)
-0.0312 (22.07) -0.0327 (6.46)
Log (Age) Loo (MEI
-0.0250 (35.65)
Log (Age) Loo. (BVME)
0.0003 (0.17)
ft2
0.0142
0.0517
0.0635
Obs
125.082
121.207
121,207
Cohort
Effects Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
0.0099 (0.69)
-0.0440 (2.89)
-0.1566 (6.84;
-0.5864 (9.64)
Log (Age)
0.6151 (5.98; 0.5401 (5.56; 03367 (3.67; 0.9154 (7.11)
Log (Age2)
-0.3130 (6.20;
-O.2709 (5.70;
-0.1789 (3.96)
-0.4026 (6.83;
Log (ME)
0.0267 (9.22) 0.1076 (9.89)
Log (BVME)
-0.0369 ( (0 .9 ) ; -0.0616 (3.85)
Log (Age) Log (ME)
-0.0228 (9.06)
Log(Age) Log (BVME)
0.0069 (1.72)
R2
0.0159
0.0210
0.0750
0.0904
Avg # of Firms 2975
2975
2887
2887
Cohort Effects
No
Yes
Yes
Yes
Gfowth in Exc«ts to Industry by Firm A Listing D*cad« Cohort Effects
- i o n .
/
ISM
- \
1971 1
Y w EfUcta
/V / \
76 1W1 19H
A
1981
\ V V '•» "
1 I 01 3006
Sales Growth
Firm Age
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
Mean
0.339 0.317 0.248 0.205 0.154 0.135 0.118 0.105 0.097 0.095
SD
0.529 0.538 0.455 0.400 0.311 0.279 0.249 0.225 0.200 0.173
Bottom 5%
-0.212 -0.340 -0.304 -0.272 -0.219 -0.208 -0.171 -0.161 -0.130 -0.108
Med
0.194 0.183 0.156 0.137 0.112 0.104 0.091 0.081 0.077 0.077
Top 95%
1.371 1.584 1.128 0.904 0.620 0.548 0.466 0.423 0.372 0.354
Mean Excess 0.156 0.137 0.070 0.031 -0.010 -0.019 -0.024 -0.030 -0.035 -0.036
Med Excess 0.024 0.016 -0.009 -0.026 -0.042 -0.040 -0.043 -0.043 -0.047 -0.045
Avg # of Firms
21 109 301 370 333 321 432 336 296 259
Sales Growth Age, Listing Cohorts, & Year Effects Decomposition
Intercept
-0.1445 (7.65;
-0.2242 (10.17) -0.4560 (17.07)
Log (Age)
1.8503 (25.81) 1.5567 (24.28) 1.7720 (25.49;
Log (Age2)
-0.8136 (26.57; -0.7685 (25.03) -0.8437 (25.77)
Log (ME)
0.0137 (20. (2 ; 0.0485 08.34)
Log (BVME)
-0.0538 (36.82) -0.1559 (28.22)
Log (Age) Loo (ME)
-0.0093 02.78)
Log (Age) Log (BVME)
0.0316 0 9 . 0 6 ;
R2
0.0244
0.0494
0.0566
Obs
116,587
112,974
112,974
uonor t Effects
Yes
Yes
Yes
F a m a M a c B e t h A v e r a g e E s t i m a t e s
Intercept
-0.0718 (2.96)
-0.1151 (4.45)
-0.1981 (6.96;
-0.4612 (9.96)
Log (Age)
1.2963 (7.23; 1.2057 (7.28; 1.0800 (7.00) 1.3482 (7.93)
Log (Age2)
-0.6458 (7.38;
-0.5971 (7.40)
-0.5367
(7.U) -0.6326 (7.83)
Log (ME)
0.0124 (10.52) 0.0567 (10.21)
Log (BVME)
-0.0529 08.42) -0.1494 02.86)
Log (Age) Log (ME)
-0.0119 (8.52;
Log (Age) Log (BVME)
0.0292 (9.75;
R2
0.0230
0.0263
0.0631
0.0731
Avgf f of F i rms 2773
2773
2691
2691
Cohort Effects
No
Yes
Yes
Yes
Sal«s Growth In E X C M I to Industry by Firm Ag« Listing Docacte Cohort Effects
108 Q (Market Value of Assets / Book Value of Assets)
Firm Age
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
Mean
2.801 2.465 2.187 1.925 1.737 1.548 1.422 1.345 1.302 1.320
SD
3.834 2.721 2.308 1.891 1.529 1.229 0.975 0.998 0.806 0.736
Bottom 5%
0.849 0.798 0.802 0.794 0.773 0.763 0.737 0.750 0.756 0.784
Med
1.785 1.658 1.477 1.371 1.275 1.197 1.153 1.101 1.087 1.096
Top 95%
7.714 6.944 5.732 4.761 4.190 3.434 2.999 2.702 2.564 2.700
Mean Excess 0.922 0.561 0.299 0.067 -0.046 -0.151 -0.161 -0.164 -0.192 -0.145
Med Excess 0.131 -0.037 -0.156 -0.215 -0.246 -0.271 -0.250 -0.213 -0.204 -0.155
Avg # of Firms
81 178 381 425 363 343 452 351 310 270
Q - Market Value of Assets / Book Value of Assets Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.1358 (1.09)
-1.2611 (11.86) -1.6744 (13.80)
Log (Age)
3.4730 (15.23; 3.1922 (16.27) 12699 (6.05)
Log (Age1)
-1.7683 (16.39; -1.5269 (16.46; -0.5044 (5.20;
Log (ME)
0.0004 (0.11) 0.0156 (»24 ;
Log (BVME)
-1.1287 (152.59; -2.7821
(111.96)
Log (Age) Log (ME)
0.0048 (1.38)
Log (Age) Log (BVME)
0.5347 (69.47)
R2
0.0132
0.1880
02217
Obs
132.414
129.661
129.661
Cohort Effects
Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
0.2178 (1.8S) 0.2077 (1-79)
-1 1776 (7.28)
-1.4850 (6.05;
Log (Age)
25836 (3.5); 21136 (2.98; 2.3577 (3.89; 0.1764 (0.37)
Log (Age2)
-1.3434 (3.75J
-1 0991 (3.18;
-1.1215 (3.87; 00203 (0.09;
Log (ME)
0.0026 (0.38J 0.0149 (0.53)
Log (BVME)
-1.0234 (17.80) -2.2886 (9.21)
Log (Age) Log (ME)
0.0023 (0.34)
Log (Age) Log (BVME)
0.4135 (5.99)
R2
0.0239
0.0291
0.3292
0.3611
Avg # of
Firms 3150
3150
3088
3088
Cohort Effects
No
Yes
Yes
Yes
Table 2.5: Firm Age and Innovative Edge
The table examines the relation of firm age with respect to three innovation measures: a) R&D over Assets, b) Granted Patent's originality, c) Granted Patent's Generality. For R&D/Assets we present two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5th and top 95"1 percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to industry (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the measure's excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. The Granted Patent's originality and generality measures are summarized by age groups in the first panel, which presents average from 1980 to 1999 of originality, generality, citations received and made citations. For each of the two Patent measures we present results of the regression decomposing the measure by firm age, listing cohorts, and year effects, however this decomposition has biases explained in detail on the paper that should be taken into consideration. T statistics are reported under the coefficient estimates in parenthesis.
R&D / Assets
Firm Age
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
Mean
0.091 0.100 0.104 0.087 0.068 0.047 0.034 0.029 0.026 0.025
SD
0.154 0.146 0.136 0.118 0.084 0.060 0.046 0.031 0.026 0.025
Bottom 5%
0.009 0.005 0.002 0.004 0.003 0.000 0.000 0.001 0.000 0.001
Med
0.044 0.062 0.071 0.058 0.046 0.030 0.021 0.020 0.020 0.017
Top 95%
0.322 0.333 0.327 0.257 0.202 0.156 0.108 0.089 0.075 0.083
Mean Excess 0.020 0.022 0.023 0.009 -0.003 -0.013 -0.014 -0.015 -0.013 -0.016
Med Excess -0.007 -0.004 0.000 -0.005 -0.008 -0.011 -0.010 -0.011 -0.009 -0.009
Avg # of Firms
39 98 218 241 191 167 192 137 123 105
R&D / Assets Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.0805 (5.73) 0.0530 (3.29) 0.1581 (9.48)
Log (Age)
-0.0310 (1.52) 0.0167 (0.99)
-0.2386 (12.98)
Log (Age2)
0.0060 (0.63)
-0.0091 (1.14) 0.1022 (12.03)
Log (ME)
-0.0114 (39.10) -0.0348 (34.21)
Log (BVME)
-0 0328 (58. (fi) -0.1028 (52.40
Log (Age) LOG (ME)
0.0075 (25.18)
Log (Age) Loo (BVME)
0.0235 (37.33)
R2
0.0198
00764
0.0979
Obs
63,330
61.715
61.715
Cohor t Effects
Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
0.0441 (5.17) 0.0518 (4.89) 0.0303 (4.22) 0.1096 (7.20)
Log (Age)
0.0265 (0.93)
-0.0057 (0.17) 0.0658 (2.83)
-0.1620 (5.97)
Log (Age2)
-0.0203 (1.51)
-0.0041 (0.27)
-0 0332 (2.98) 0.0664 (5.13)
Log (ME)
-0.0069 (6.17)
-0.0276 (8.09)
Log (BVM6)
-0.0243 (9.64)
-0.0822 (10.50)
Log (Age) Log (ME)
0.0064 (8.66)
Log (Age) Log (BVME)
0.0192 (9.63)
R2
0.0319
0.0384
0.0928
0.1209
Avg # of Firms 1505
1505
1470
1470
Cohor t Effects
No
Yes
Yes
Yes
R&D I Assets in Exc*ss to Industry by Firm A g * Listing D*cad« Cohort Effect*
00*
«s*
, t t* t s n we
Y w Effects
.». .** .» , « » 20 C MOS
110 Average US Patents Granted by Firm Age (1980 -1999)
Firm Age Average Age Average Patent
Originality
Average Patent
Generality
Citations Made
Citations Received
Average Patents Granted
Avg # of Firms
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
3.6 6.2 10.4 15.5 21.9 30.3 45.1 65.5 86.8 128.3
44.54% 38.38% 40.07% 40.48% 40.15% 40.29% 39.37% 40.71% 40.86% 39.27%
37.77% 38.39% 36.68% 34.93% 32.44% 31.37% 29.21% 29.24% 29.10% 28.31%
7.74 8.17 11.52 10.85 10.37 11.08 11.23 12.10 11.32 10.95
12.89 13.85 10.34 7.74 6.25 5.80 4.91 4.48 4.71 4.88
8.37 9.70 8.10 12.69 11.03 10.97 14.66 28.46 53.04 53.39
5 12 30 48 55 63 94 97 107 101
Patent Originality (1980 - 1999)
Intercept
0.3069 (12.55) 0.3207 (12.52) 0.4039 (9.78)
Log (Age)
0.1190 (1.04)
-0.0002 (0.00)
-0.1986 (1.46)
A g e , L i s t i n g C o h o r t s , & Y e a r E f f e c t s D e c o m p o s i t i o n
Log (Ager(
-0.0567 (1.02) 0.0020 (0.03) 0.0696 (1.39)
Log (ME)
-0.0006 (0.49)
-0.0156 (2.92)
Log (BVME)
-0.0038 (1.43)
-0 0366 (332)
Log (Age) Log (Age)
Log (ME) Log (BVME)
0.0040 0.0093 (2.95) (3.13)
R2
0.0354
0.0364
0.0375
Obs
12,156
11.798
11.798
Cohort
Effects Yes
Yes
Yes
Patant Originality by Firm Ag« Listing D»cad« Cohort Effect*
Patent Generality (1980 -1999) Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.4972 (19.46) 0.4885 (18.28) 0.4744 (11.05)
Log (Age)
-0.0818 (0.69;
-0.1113 (0.87)
-0.1043 f0 .7 f l
Log (Age2)
0.0273 (0.48J 0.0417 (0.67J 0.0401 (o.6o;
Log (ME)
0.0020 (1.60) 0.0041 (0.74)
Log (BVME)
-0.0106 (3.78;
-0.0136 (1.17)
Log (Age)
Log (ME)
-0.0005 (0.37;
Log (Age)
Log (BVME)
0.0008 (0.25;
R2
0.2069
0.2083
0.2084
Obs
11.039
10,721
10,721
Cohor t
Effects Yes
Yes
Yes
Patent 0*n*ratrty by Firm Aga
I l l
Table 2.6: Firm Age and Process Efficiency
The table examines the relation of firm age with respect to three efficiency measures: a) Gross Margin; b) Net Operating Profitability After Tax; c) Invested Capital Turnover. For each measure we present two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5lh and top 95th percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to 48 industries (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the measure's excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. Finally, for each measure we present graphs of the measure's decomposition into age, listing cohort, and year effect. As a robustness check in all our regressions we add the firms size (log(ME)) and book-to-market ratio (log(BVME)) and interact those variables with the age (log(Age)) variable. T statistics are reported under the coefficient estimates in parenthesis.
Firm Age
4<= (5-7) (8-12)
(13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mean
0.153 0.138 0.209 0.229 0.319 0.321 0.302 0.307 0.313 0.327
SD
0.841 0.951 0.837 0.772 0.448 0.306 0.302 0.161 0.190 0.170
Gross Marg Bottom
5% -1.493 -2.168 -0.978 -0.830 0.062 0.085 0.090 0.097 0.103 0.107
Med
0.333 0.350 0.366 0.356 0.344 0.315 0.294 0.289 0.297 0.310
in
Top 95%
0.716 0.757 0.754 0.722 0.692 0.649 0.623 0.584 0.589 0.620
Mean Excess -0.127 -0.106 -0.032 -0.010 0.056 0.043 0.017 0.017 0.005 0.027
Med Excess 0.021 0.040 0.057 0.050 0.041 0.019 -0.001 -0.006 -0.005 0.001
Avg # of Firms
78 174 374 421 362 345 460 357 312 272
-OM
::
Intercept
-0.0120 (0.31)
-0.0326 (0.70)
-0.3480 {6.40)
Intercept
0.0794 (5.60; 0.0056 (0.31)
-0.0508 (2.11)
-0.3536 (5.31)
Log (Age)
-0.8481
(9.72; -1.1209 (12.83) -0.4118 (4.33)
Log (Age)
-0.8653 (B-13)
-0.6032 (5.65;
-0.9044 (5.97;
-0.28O0 (2.70)
Gross Margin A g e , L i s t i n g C o h o r t s , & Y e a r E f f e c t s
Log (Age2) Log (ME)
0.4265 (10.33; 0.5401 0.0436 (f3.07J (27.57; 0.2315 0.1128 (5.26; (20.02)
L-<BVME> E E 2
0.0434 ((3.25J 0.2977 -0.0212 (26.40; (13.48;
D e c o m p o s i t i o n
Log (Age) Loa IBVMEI
-0.0820 (23.58)
F a m a M a c B e t h A v e r a g e E s t i m a t e s
Log (Age2) Log (ME|
0.4280 (6.03; 0.3024 (5.68; 0.4354 0.0320 (5.91) (7.73) 0.1725 0.1041 (3.23; (6.74;
Gross Margin in Exc«ss to Industry by Firm Ag«
r f
* • " « •
„ ,„. ,., ::
Log (BVME) £ £ »
0.0188 (2.20) 0.2172 -0.0216 (6.29; (6.39)
Log (Age) Log (BVME)
-0.0635 (7.49)
Listing D*cad« Cohort Effvcts
,-.,-,-..»,».«..-.,--,--..-.
R2
0.0070
0.0128
0.0172
R2
0.0108
0.0140
0.0454
0.0557
ooso
owo
oo»
4020
oa»
Obs
132.413
128.385
128.385
l!{
3150
3058
3058
Cohort
Effects Yes
Yes
Yes
Cohort
Effects No
Yes
Yes
Yes
Y M T Effects
, n ,., ,« - '- " ""
Net Operating Profit After Tax Margin (NOPAT)
112
Firm Age Mean SD Bottom 5%
Med Top 95% Mean Excess
Med Excess
Avg # of Firms
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
0.110 0.086 0.089 0.100 0.110 0.119 0.120 0.131 0.135 0.139
0.177 0.189 0.184 0.165 0.136 0.125 0.118 0.124 0.107 0.097
-0.069 -0.119 -0.098 -0.039 0.003 0.022 0.029 0.034 0.041 0.048
0.101 0.092 0.094 0.095 0.095 0.095 0.096 0.102 0.110 0.118
0.312 0.276 0.268 0.260 0.242 0.250 0.257 0.295 0.285 0.263
-0.010 -0.026 -0.019 -0.005 0.005 0.009 0.006 0.005 0.004 0.005
-0.011 -0.010 -0.007 -0.003 -0.002 0.000 -0.001 0.000 0.001 0.002
31 100 268 340 316 313 426 339 301 263
Intercept -0.0035 (0.46)
-0.0367 (4.21)
-0.1522 (14.20)
00136 (1.75) 0.0147 (2.47)
-0.0347 (5.18)
-0.1578 (9.06)
Net Operating Profit After Tax Margin (NOPAT) Age, Listing Cohorts, & Year Effects Decomposition
Log (Age)
-0.2567 (10.74) -0.4443 (19.23) -0.2626 (10.69)
Log (Age*) Log (ME) Log (BVME)
01290 (11.29) 0.2136 (19.34) 0.1386 (12.01)
Log (Age) Log (Age) Log (MEI Loo (BVME) Obs
00158
Cohort Effects
113.163 Yes
0.0208 0.0022 (70.78J (3.45; 0.0432 0.0482 (36.80) «9 .72)
-0.0063 (19.91)
-0.0138 (19.38)
109.977
109.977
Fama MacBeth Average Estimates Intercept Log (Age) Log (Age2)
-0.2322
(2.92;
-0.1932
(3.15) -0.3737
(5.52;
-0.1837
P-07;
0.1157
(2.95;
0.0958
(3.14) 0.1783
(5.35;
0.1014
(3.33)
Log(ME) Log,BVME, £ * £ > L o g ^ E ,
0.0173 -O.0008 ()5.02; (0.43) 0.0423 0.0348
(12.05) (4.13) -0.0070 (9.95)
-0.0106
0.0078
0 0112
0.0766
0.0845
Avg # of Firms 2691
2691
2620
2620
Yes
Cohort Effects
No
Yes
Yes
Yes
NOPAT in Excess to Industry by Firm Ag« Listing Dseads Cohort Effects
113
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
Mean
1.731 1.939 1.952 2.098 2.086 2.086 2.035 1.803 1.653 1.451
SD
1.416 1.781 1.756 1.819 1.755 1.681 1.563 1.467 1.404 1.233
Invested Capital Bottom
5% 0.498 0.330 0.337 0.407 0.485 0.515 0.477 0.299 0.295 0.197
Med
1.439 1.536 1.531 1.663 1.677 1.720 1.720 1.578 1.470 1.327
Turnover
Top 95%
3.303 3.991 4.000 4.090 3.995 3.804 3.795 3.279 2.879 2.498
Mean Excess -0.161 -0.051 -0.017 0.117 0.081 0.085 0.061 -0.041 -0.097 -0.164
Med Excess -0.299 -0.283 -0.291 -0.177 -0.175 -0.177 -0.147 -0.193 -0.194 -0.208
Avg # of Firms
31 100 268 340 316 313 426 339 301 263
Invested Capital Turnover Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.8199 (11.16) 0.7640 (9.09) 1.6166 (15.63)
Log (Age)
-3.4758 (15.22) -1.8305 (8.22)
-3.0740 (12.96)
Log (Age2)
1.6422 (15.04) 0.9197 (8.64) 1.4245 (12.79)
Log (ME)
-0.1875 (66.10) -0 3475 (30.70)
Log (BVME)
-0.4110 (66.59) -0.6537 (27.70)
Log (Age) Log IMEI
0.0448 (14.69)
Log (Age) Lou (BVME)
0.0723 (10.55)
R'
0.0039
0.0557
0.0577
Obs
113.163
109.977
109.977
Cohor t
Effects Yes
Yes
Yes
F a m a M a c B e t h A v e r a g e E s t i m a t e s
Intercept
0.9966 (17.70) 0.8451 (13.95) 0.9528 (13.08) 2.0461 (17.07)
Log (Age)
-4.3666 (10.39) -4.1351 (9.56)
-2.1418 (5.09;
-3.7196 (7.70)
Log (Age1)
2.0679 (10.06) 1.9657 (9.31) 1.0642 (5.(7) 1.7003 (7.32)
Log (ME)
-0.1663 (25.38) -0.3852 (22.65)
Log (BVME)
-0.3640 (18.91) -0.5072 (9.35)
Log (Age) Log (ME)
0.0588 03.57)
Log (Age) Log (BVME)
0.0409 (3.42)
0.0063
0.0090
0.0552
00593
A v g # of Firms 2691
2691
2620
2620
Cohort Effects
No
Yes
Yes
Yes
Inv. Capital Turnover in Excess to Industry by Firm Age
L i s t i ng Decade Cohor t Effects
Graph 9. Invested Capital Turnover (1965 -2006)
(5-7) (6-12) (13-18) (19-25) (26-35) (36-55) (56-75) (?6-100> +100
F i rm Age
—•—A\erage —w— Median
Graph 9. Excess to Industry Invested Capital Turnover
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) [76-100) +100
F i rm Age
—•— Mean Excess to the Industry
Figures 2.9 and 2.10, Firm Age Raw and Excess to Industry Invested Capital Turnover
114
Table 2.7: Liquidity and Cash Flow Risk
The table examines the relation of firm age with respect to: a) Amihud's Illiquidity measure, b) Percentage of non-traded days over the year, c) Equity Duration as defined by Dechow, Sloan, and Soliman. For each measure we present two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5th and top 95th percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to 48 industries (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the measure's excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. Finally, for each measure we present graphs of the measure's decomposition into age, listing cohort, and year effect. As a robustness check in all our regressions we add the firms size (log(ME)) and book-to-market ratio (log(BVME)) and interact those variables with the age (log(Age)) variable. T statistics are reported under the coefficient estimates in parenthesis. With our illiquidity measure we also include a panel that presents the equal weighted portfolio returns and excess returns to Fama & French portfolios for firms sorted into three Liquidity categories (low, medium, and high) and Firm Age. Additionally, this panel computes the average market value of the stocks in the portfolio and the average number of firms.
Illiquidity (Amihudl2)
Firm Age Mean SD Bottom 5% Med Top 95% Mean Excess
Med Excess
Avg # of Firms
4<=
(5-7)
(8-12)
(13-18)
(19-25)
(26-35)
(36-55)
(56-75)
(76-100)
+100
0.995
1.009
0.980
1.018
0.969
0.927
0.739
0.552
0.448
0.341
0.856
1.049
1.085
1.255
1.204
1.186
0.997
0.799
0.669
0.576
0.274
0.194
0.150
0.139
0.134
0.117
0.091
0.074
0.056
0.053
0.786
0.696
0.641
0.588
0.553
0.503
0.378
0.280
0.219
0.146
1.938
2.127
2.211
2.430
2.307
2.234
1.835
1.346
1.079
0.817
0.235
0.228
0.184
0.214
0.181
0.151
-0.001
-0.158
-0.222
-0.308
0.064
-0.033
-0.098
-0.167
-0.172
-0.187
-0.264
-0.324
-0.341
-0.426
94
173
359
410
355
334
454
357
320
282
Illiquidity (Amihudl2) Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.2152 (7.10) 1.1439 (33.08) 2.1244 (54.63)
Log (Age)
-0.0980 (26.52; 0.0475 (14.68) -0.2492 (38.25)
Log (ME)
-0.3001 (226.61) -0.5111 (111.78)
Log (BVME)
0.0318 (11.58) 0.1105 (12.19)
Log (Age) Log (ME)
0.0595 (47.34)
Log (Age) Log (BVME)
-0.0256 (9.19)
R2
0.0418
0.3753
0.3899
Obs
131,350
120.701
120,701
Cohort Effects
Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
1.2446 (10.84) 0.0823 (1.96) 1.4242 (14.16) 2 7835 (12.36)
Log (Age)
-2.9375 (9.26)
-0.1052 (7.58;
0.0467 (9.33)
-0.3605 (8.18)
Log (ME)
-0.2994 (1.27)
-0.6386 (4.57;
Log (BVME)
0.0128 0.00
0.1117 (8.27)
Log (Age) Log (ME)
0.0905 (4.80)
Log (Age) Log (BVME)
-0.0295 (63.93;
R2
0.0426
0.0749
0.4474
0.4722
Avg # of
Firms 3124
3124
2876
2876
Cohort
Effects No
Yes
Yes
Yes
Listing Decade Cohort Effects
Equ
al W
eigh
ted
Por
tfol
ios
Sor
ted
by
Liqu
idity
and
Age
grou
p
Jity f •5 •s, i ~ 5 •o £ | 3
N Liq 3
Fir
m A
ge
Avg
Ret
urn
r S
har
pe
Rat
io
FF E
xces
s R
etur
n
t-
FF
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f F
irm
s
Avg
Ret
urn
t
Sh
arp
e R
atio
FF
Exc
ess
Ret
urn
t-
F
F E
xces
s A
vg M
arke
t V
alue
A
vg #
of
Firm
s
Avg
Ret
urn
t
Sh
arp
e R
atio
FF
Exc
ess
Ret
urn
f-
F
F E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Firm
s
4<=
-0.3
0%
-0.5
6 -0
.099
-1
.20%
-3
.07
1296
20
0.33
%
0.66
0.
114
-0.6
1%
-2.1
1 14
4 44
1.68
%
3.66
0.
596
0.37
%
1.24
34
45
(5-7
)
0.30
%
0.63
0.
106
-0.5
4%
-1.9
4 89
4 39
0.65
%
15
0 0.
232
-0.4
9%
-2.1
3 13
5 68
1.86
%
4.18
0.
657
0.27
%
1.14
28
78
(8-1
2)
0.85
%
2.25
0.
350
-0.0
8%
-0.4
4 91
0 72
1.09
%
3.00
0.
466
-0.0
9%
-0.7
1 12
5 13
0
1.83
%
4.75
0.
737
0.28
%
2.15
27
14
5
(13-
18)
0.99
%
2.93
0.
454
-0.0
3%
-0.2
1 12
00
88
1.31
%
3.84
0.
595
0.06
%
0.61
13
4 14
3
2.08
%
6.04
0.
938
0.51
%
5.72
28
15
7
(19-
25)
1.03
%
3.34
0.
519
-0.0
1%
-0.0
7 20
77
81
1.47
%
4.82
0.
748
0.25
%
3.05
14
9 11
4
1.94
%
5.99
0.
929
0.36
%
4.29
28
13
5
(26-
35)
1.21
%
4.44
0.
689
0.21
%
2.19
21
24
89
1.41
%
5.05
0.
783
0.14
%
18
7 16
7 10
3
1.94
%
6.40
0.
994
0.28
%
3.38
28
11
9
(36-
55)
1.11
%
4.65
0.
721
0.06
%
1.10
20
32
136
1.33
%
5.35
0.
831
0.04
%
0.55
19
2 13
8
1.77
%
6.59
1.
022
0.08
%
1.01
34
13
8
(56-
75)
1.07
%
5.19
0.
806
0.02
%
0.36
31
86
142
1.18
%
5.12
0.
795
-0.1
7%
-2.1
5 20
1 10
7
1.57
%
6.37
0.
989
-0.1
6%
-1.5
0 43
75
(76-
100)
1.05
%
5.31
0.
825
0.01
%
0.17
36
76
153
1.23
%
5.53
0.
859
-0.1
5%
-1.6
9 22
5 83
1.53
%
6.31
0.
979
-0.2
0%
-1.7
0 43
58
+100
1.14
%
5.74
0.
890
0.09
%
14
4 62
74
169
1.22
%
5.59
0.
867
-0.1
7%
-17
3 21
7 56
1.40
%
5.00
0.
776
-0.3
7%
-2.2
9 42
33
Mat
ure
-Y
ou
ng
1.36
%
3.44
0.
612
1.25
%
3.23
19
41
119
0.81
%
2.80
0.
484
0.79
%
2.81
15
0 17
4
0.10
%
0.33
0.
053
-0.0
3%
-0.0
9 31
18
6
Mat
ure
-O
ld
-0.1
2%
-0.6
7 -0
.105
-0
.10%
-0
.70
4176
25
0
0.25
%
1.51
0.
234
0.43
%
3.32
18
3 17
1
0.54
%
2.74
0.
425
0.73
%
4.02
35
16
8
All
Fir
ms
1.02
%
4.26
0.
661
0.00
%
-0.0
2 29
78
977
1.19
%
4.30
0.
667
-0.0
7%
-19
2 16
8 97
7
1.84
%
6.f
2 0.
950
0.23
%
5.26
32
97
7
u>
116
Graph 11. Illiquidity and Firm Age
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100) +100
—•— High Liquidity -"— Med Liquidity —*— Low Liquidity
Figure 2.10 Illiquidity and Firm Age
Percentage of Non Traded Days
Firm Age Mean SD Bottom 5%
Med Top 95% Mean Excess
Med Excess
Avg # of Firms
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
0.197 0.168 0.153 0.137 0.123 0.129 0.117 0.099 0.070 0.055
0.144 0.167 0.165 0.168 0.167 0.172 0.181 0.177 0.147 0.136
0.076 0.019 0.004 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.152 0.125 0.111 0.087 0.059 0.063 0.026 0.007 0.000 0.000
0.396 0.388 0.357 0.358 0.354 0.375 0.375 0.351 0.268 0.211
0.059 0.039 0.026 0.017 0.007 0.016 0.006 -0.011 -0.036 -0.055
0.031 0.007 -0.006 -0.023 -0.042 -0.037 -0.058 -0.064 -0.073 -0.087
96 194 403 445 383 364 490 384 333 290
Percentage of Non Traded Days
Intercept
-0.0077
(1.19) 0.0780 110.36) 0.1937 (22.13)
Log (Age)
0.0973 (6.78) 0.2591 (18.64) 0.1417 (9.36)
Age, Listing Cohorts, & Year Effects
Log (Age2)
-0.0498 (7.35)
-0.1162 (17.69; -0.0745 ((0.64)
Log (ME)
-0.0426 ((66.82) -0.0641 (70.86)
Log (BVME)
-0.0011 (2.12; 0.0054 (3.03)
Log (Age) Loo (ME)
0.0061 (24.31)
Decomposition Log (Age)
Log IBVMEI
-0.0021 (3.73)
R2
0.0868
0.2718
0.2763
Obs
141,895
129.661
129,661
Cohort Effects
Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
0.1304 (5.99)
-0.0507 (5.31) 0.0953
( 7 5 9 ; 0.1799 (6.61)
Log (Age)
-01769 (6.30)
-0 0822 (1.96) 0.1295 (1.03)
-0.0126 (0.24)
Log (Age2)
0.0708 (5.79; 0.0388 (1.94)
-0.0520 (3.40)
0.0070 (0.30)
Log (ME)
-0,0422 ()3.oo; -0.0599 (780)
Log (BVME)
0.0046 (2.87) 0.0110 (4.25)
Log (Age) Log (ME)
0.0049 (3.57)
Log (Age) Log (BVME)
-0.0019 (2.39)
R2
0.0247
0.0939
0.2953
0.3018
Avg # of
Firms 3375
3375
3088
3088
Cohort Effects
No
Yes
Yes
Yes
Percentage of Non Traded Days in Excess to Industry by Firm Ag« i x
S"k
«
-7%
Y e a t E f f o c t s
• '•" XT76 <»< ,« 1991 > ae 2001 *»
117
Equity Duration
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mean
16.86 17.13 16.50 16.00 15.54 15.20 14.81 14.55 14.52 14.62
SD
3.907 3.461 3.231 3.036 2.912 2.809 2.778 2.686 2.601 2.402
Bottom 5%
11.41 11.67 11.27 11.00 10.64 10.43 9.92 9.80 9.88 10.40
Med
16.83 17.10 16.57 16.11 15.63 15.37 15.07 14.75 14.70 14.83
Top 95%
22.79 23.17 21.61 20.51 19.92 19.22 18.66 18.23 18.05 17.77
Mean Excess
1.11 1.33 0.76 0.37 0.04 -0.12 -0.30 -0.42 -0.41 -0.20
Med Excess
0.94 1.19 0.73 0.42 0.09 0.01 -0.11 -0.23 -0.22 -0.05
Avg # of Firms
38 116 287 347 312 300 406 322 291 255
Equity Duration
Intercept
0.1579 (0.84)
-1.3552 (7-64) 1.9378 (9.08)
Log (Age)
8.5754 (19.61) 8.8870 (21.53) 6.0089 (13.48)
Age, Listing Cohorts, & Year Effects
Log (Age2)
-4.3385 (20.79) -4.3603 (22.15) -3.3771 (16.21)
Log (ME)
-0 0640 (11.31) -0.5974 (27.48)
Log (BVME)
-1.3600 (114.96) -0.5491 (12.81)
Log (Age)
Log (ME)
0.1438 (24.24)
Decomposition Log (Age)
Log (BVME)
-0.2520 (19.53)
R2
0.0300
0.1427
0.1545
Obs
112.197
112,176
112,176
Cohort
Effects Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
0.4357 (2.01) 0.5303 (3.83)
-0.8043 (4.83) 1.5816 (5.92)
Log (Age)
6.6475 (5.12) 4.8550 (4.42) 5.2392 (4.75) 3.6486 (3.73)
Log (Age3)
-3.4256 (5.43)
-2.5267 (4.72)
-2.5776 (4.79)
-2.1379 (4.39)
Log (ME)
-0.0304 (1.58)
-0.4654 (8.95)
Log (BVME)
-1.6307 (13.36; -0.9692 (7.76)
Log (Age) Log (ME)
0.1176 (9.98)
Log (Age) Log (BVME)
-0.1945 17.31)
R2
0.0350
0.0417
0.2582
0.2675
Avg # of Firms 2668
2668
2672
2672
Cohort Effects
No
Yes
Yes
Yes
Equity Duration In E X C M S to Industry by Firm Ag«
* * * ^ ^ M ^ ^
Listing Docad* Cohort Effect*
.-
,„ *,*,
BM 1971 197*
Yoar Effects
19(1 t9M 1991 .<» » 0 1 20QS
118
Table 2.8: Default and Debt Structure
The table examines the relation of firm age with respect to: a) Ohlson's Default Probabiltiy, b) Actual Default and, c) Debt Structure. We analyze Ohlson's Default in two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5th and top 95th
percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to 48 industries (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the Ohlson's Default Probability in excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. For this measure we present graphs of the measure's decomposition into age, listing cohort, and year effect. T statistics are reported under the coefficient estimates in parenthesis. The third panel presents actual annualized delisting and merging probabilities tracing them as the firms in an initial age group mature and move to older groups. The fourth panel presents in the same form as the firms panel average, median, standard deviation, etc. of the percentage of long term debt to total debt
Ohlson's Default Probability
Firm Age Mean SD Bottom
5% Med Top 95%
Mean Med Avg # of Excess Excess Firms
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
2.801 2.465 2.187 1.925 1.737 1.548 1.422 1.345 1.302 1.320
3.834 2.721 2.308 1.891 1.529 1.229 0.975 0.998 0.806 0.736
0.849 0.798 0.802 0.794 0.773 0.763 0.737 0.750 0.756 0.784
1.785 1.658 1.477 1.371 1.275 1.197 1.153 1.101 1.087 1.096
7.714 6.944 5.732 4.761 4.190 3.434 2.999 2.702 2.564 2.700
0.922 0.561 0.299 0.067 -0.046 -0.151 -0.161 -0.164 -0.192 -0.145
0.131 -0.037 -0.156 -0.215 -0.246 -0.271 -0.250 -0.213 -0.204 -0.155
81 178 381 425 363 343 452 351 310 270
Ohlson's Default Measure Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.9710 (15.23) 1.6585 (22.53) 3.2722 (40.3d)
Log (Age)
-0.2568 (35.88; 0.0517 (7.82;
-0.4686 (36.28)
Log (ME)
-0.3815 (137.98) -0.8525 (91.46)
Log (BVME)
-0.4540 (79.79) -1.1612 (67.96;
Log (Age) Log (ME)
0.1380 (53.63;
Log (Age) Log (BVME)
0.2289 (39.67;
R2
0.0221
0.1509
0.1726
Obs
133,809
129,747
129,747
Cohort Effects
Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
1.2278 (19.02) 0.7539 (12.79) 1.6745 (16.93) 3.0935 (11.74)
Log (Age)
-1.5020 (3.42)
-0.2367 (15.49) 0.0225 (1.39)
-0.4985 (7.54J
Log (ME)
-0.3305 (10.99) -0.7688 (7.97)
Log (BVME)
-0.3666 0.00
-0.8720 (9.19)
Log (Age) Log (ME)
0.1270 (6.59)
Log (Age) Log (BVME)
0.1652 (37.37;
R2
0.0313
0.0398
0.1818
0.2037
Avg # of Firms 3183
3183
3091
3091
Cohort Effects
No
Yes
Yes
Yes
Default Prob in Excass to Industry by Firm Ag«
Listing Decad* Cohort Effects
+ <J
> y
w t
o -»
-»
-s<
-.
i T
o>
o> 0
) <o
w
o
-±
°§
en en
01 e
n 0
0 -D
i •
pppppppppp
£
oo
po
po
op
op
h
oh
oh
oh
oh
oc
oc
oc
oc
oc
o
^i
pc
n-
jc
oo
-^
ho
ho
ho
oopopoopop
OlCOhOCDOCn-'OO-'O
O*>-h00000C0-»-Cnh0O
oopopoopop
bobobobobo^'-g^^^
MU
lO
iO
lM
lD
^I
Ol
tJ
IO
)
co^aoocoai*o-'0
pppp
pppp
pp
b(
Ob
(O
tO
(D
<o
b(
D(
D
CO
QO
CO
CO
OO
-JC
O-v
lOO
CO
popppooooo
. NiSJ.K°-Aho-&.coco
00
00
00
00
00
bbbbbbbpbb
01
O-
'0
)-
«0
1C
1-
'C
00
)
MM
UM
)J
UU
U-
» O
TC
OC
O-'-
'-'O
lhO
Cn
J
i.
(D
4>
-A
O*
.0
0-
^
•n
5"
>
(D
a 3 <D
fi)
3 </)
O 00
en
©
3
S
(0 a H
0 •D
(O
Ul
S?
m _
x
3
O
(6
CD fi
>
<f>
3
(A m Med
xces Ul
>
•n
<
=T I
Q
3 *
» 0
^t
Lon (Q
H
CD
3 a CD
O-
*• ^5
6~
,2
cn
co
io
-»
-*
-£
;_
g^
tnjn
^c
n^
tn^
jiJ
-^- Ol
CO
-»J
O
*.
-t>.
en
-*
M
--1
^
01 *
. ho
-•
• co
co
a>
a>
« t°
. n
t 7
t 0
1
->•
O
<S
CD
ho
&
• en
3-
£ -»
• CO
*•
en -
»
ho
-J
->
• co
K
CD
00 £
II c 3 cr
CD
76- 0
(56 •i< 1
<*
(26' CO
CO
M
•A
CO
-4
t
ou
.<5
<5
<je
oi!
f
a*
Sif
t 1
P
a
ho en
->i
00 co
bo
ho
oo
00 ho
-j b
3?
CD
'•«
J
0^
00 bo
gs
CD
CO
^
9 >l
.1
CD
b
0^
00
'*• 3?
-J
0 '*.
0s
0 en
3*
_i>
4*
CD
C
O
s?
co
b £ 01 b
Mer (Q
CD
Q. cf
jQ
C
~t'
CD
a Firms
co en
-J
bo b
'*
•
00 0
0
"co b
ui
u
M
^
-»
'2.
Old
! O
IO
UT
|
2S
SK
»^
"
LJ.
Lk
b
-S
-2 -
8
(5^
o^
o^
en e
n *
* p
co
b
en
bo
*>.
co c
o •
** *
. ^
is b
) M
s
* A
O
) M
O
U
L *
i io
b b
*
? >
5 ^
-5
*
5 *
S
&•*
£•*
tf
- 8s*
<?
*
o^
M
CO
O
N^
b
co
b
ho c
o
Ko
bo
3
O 5
3
SL
S£ 0 a 2.
i sting •n
irms
rln
1 OS
K0 •^
z V
CD
its
"-*
Tl
- 3
? >
h
otO
s
®
CD
roup (26-3 S
Q ut!
— CD
to
O
) T
O
rog -55)
^ CD
-~
<fl
cn (
A
? O
Si
-&
0 0 + 0 0
-0
CD
rcenta CO
CD
O
f Fir 3
(A
p*
>
3T
0)
r
* € "~
a CD
O eliste a.
0 S
CD
rge a.
5"
5 eFut ure
M3
o
CN
*"•"•»
CO
o
o CM
• w
C
O
en T
-w
•o iel
>•
*-• .n
0
) Q
a>
re
i-o
Q.
L.
o O
o v>
o> .E
5 u
-<
V)
Med
xces
w
_ w
C
V
) re
a>
o> o
s
i2
s?
a.
o I-0
)
5
E
O
N?
3 S>
m o
CO
c re
a> 5 0) D
l
< 1
Li.
CO
••* CM
h
-•>- C
M
-0.010 0.048
0.034 0.033
co
co
C
i ID
in
•*•
o
o
CO
C
M
CM
00
C
M
•*
oo c
o
••$ CM
If) C
D
CO
C
M
h-
r--
T-
CM
^
co
in
CM
co
m
C
D
CD
o
o c
o
4<= (5-7)
576 0.062 0.035 608
o 705
CO
254
s. 338 644
00 (8-12)
630 0.033 0.054 603
o 535
oo 252
i--328 657
oo (13-18)
524 0.022 0.023 446
o 542
00 252
i^ 315 627
oo (19-25)
462 -0.006 0.024 688
o 325
oo 216
(^ 317 558
co (26-35)
504 -0.095 -0.012 570
o 873
r»-220
r»-285
oo (36-55)
368 -0.141 -0.024 385
o 808
r-235
K 239 291
CO (56-75)
I-- C
O
ID
IT)
CO
CO
-0.189 -0.230
-0.046 -0.132
co
f>
° oS
o
co
in
T
-
r~~ h-
co
•*
o
o>
C
M
T-
r>- i^
1^- T
-O
C
M
CM
i-
00
r-
00 C
M
T-
O
oo c
o
(76-100) +100
_^
CO
o o C
N
i
If) oo o>
T
-10 O
) c
*•
* tRa
A cv
Q
**-o </>
oi .E
3: u
-<
(fl
Med
xces
UJ
_ *>
C
M
re
a v
0
si2
5? u
> en
a o
l-•a CD
s ottom
5%
01 a CO
c re
a S
0)
2>
< §
LL
^(
DO
't
Nt
CO
SK
s
sn
cM
to
oi
om
in
, C
Mm
cD
in-
^m
co
co
co
oo
oo
oo
do
Ifi O
) B
N
ID
t
O
O
O
O
O
O
T-C
MC
OT
rcM
CM
cM
v-C
OC
M
qqqqqqqqq
oooooo'oo'do
ooocDi^o>min->-a5i-
OO
^C
M^
CO
OC
OS
OD
t r
^r
-r
;T
-r
;N
MM
C\
l4
od
od
od
od
od
od
ed
ed
od
ed
CO
OC
DC
MO
OC
Ot
-O
CM
O
cD
m'<
-'*
co
cM
'<-
T-
^r
o
oo
i-~
r^
i^t
»-
oo
cn
i-c
M^
ib
(D
!b
lB
(B
(0
!D
SN
S
iriiriiriiriiriiriwiriiriiri
So
io
ir
on
tO
Nc
oi-
oio
oia
jfflic
mN
i-T
-O
T-
OO
OO
OO
O
co
t^in
co
o5
co
i-cM
'<-m
N
CO
SS
CO
CO
tMT
OC
OO
) (
OS
NS
NO
OO
O'
-C
I c
bc
offlffic
offltd
sss
^IIIIIIP
CO
o
o
CN •
If) C
O
1 Dl .E
(0
1 s
S
x
UJ
C
(A
re a>
« u
S
x
IT)
a>
a o
•o
«
S in
o
CO
n^
(o
o^
r\
i'<r
co
sc
o C
Mt
^t
^C
OC
MC
DO
CD
in
iO
T
-cM
inc
oin
^o
co
co
co
itr
tn
iflM
Din
ioin
c)
oo
oo
oo
oo
oo
ddddddddoci
ooooogo£S8
oooooqopqq
oooooqoqqq
SO
) 0
5
•>*
_
_
in
min
inc
oc
OT
-T
-o m
o
CD co
a> o
ID
M »
r
r .
CM
CM
CM
CM
CM
CM
CM
CM
CM
CN
dddddddddd
mr~-cDCMCM-^T-i-^ir>in
dddddddddd
co
co
sc
oto
ss
co
co
co
inin
min
min
inm
inin
o
oo
oo
oo
oo
o
dd
dd
dd
dd
dd
(D
OI
CD
CO
^O
OO
OJ
CO
r-.r^r^r^r--r--cDcomm
oooooooooo
dddddddddd
mS
KM
NO
Ol
f-
O
CO
CO
CO
CO
CO
CO
CM
CM
CM
CM
OO
OO
OO
OO
OO
.", r»- £2 »-
CN co
m i 2 o
121
Table 2.9: Firm Age and Profitability
The table examines the relation of firm age with respect to three growth measures: a) 1 year Asset Growth, b) 1 year Sales Growth, c) Tobin's Q proxy (market value of Assets / Book Value of Assets). For each measure we present two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5th and top 95th percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to industry which is computed in the following form: 1) compute the industry average following Fama and French 2000, 48 industry classification; 2) subtract the industry average at the firm level from the raw measure; 3) group the firms by age and compute the average and median of the of the excess to the industry measure (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the measure's excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. Finally, for each measure we present graphs of the measure's decomposition into age, listing cohort, and year effect. As a robustness check in all our regressions we add the firms size (log(ME)) and book-to-market ratio (log(BVME)) and the interact those variables with the age (log(Age)) variable. T statistics are reported under the coefficient estimates in parenthesis.
Return on Assets
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75)
(76-100) +100
Mean
-0.009 -0.019 0.011 0.041 0.061 0.074 0.077 0.079 0.079 0.081
SD
0.200 0.209 0.184 0.157 0.125 0.099 0.081 0.068 0.061 0.061
Bottom 5%
-0.410 -0.480 -0.388 -0.273 -0.165 -0.085 -0.045 -0.022 -0.011 -0.001
Med
0.039 0.032 0.055 0.073 0.079 0.081 0.081 0.081 0.080 0.082
Top 95%
0.227 0.246 0.243 0.234 0.227 0.208 0.188 0.177 0.165 0.170
Mean Excess -0.059 -0.060 -0.030 -0.003 0.013 0.019 0.016 0.013 0.010 0.013
Med Excess -0.020 -0.018 0.004 0.018 0.022 0.021 0.016 0.011 0.007 0.007
Avg # of Firms
38 124 297 343 306 306 417 320 276 225
Return on Assets Age, Listing Cohorts, & Year Effects Decomposition
Intercept
0.0060 (0.77)
-0.0033 (0.38)
-0.30O0 (29.74)
Log (Age)
-0.5675 (24.53) -0.7274 (33.4S) -0.0615 (2<*>
Log (Age )
0.2841 (25.79; 0,3504 (33.84) 0.0592 (5.42)
Log.ME, Log.BVME, £ £ g £ • * £ R2
0.0385
0.0165 (28.53)
0.0800 0.1536 (77.92) (75.06)
0.0252 (90.18)
-0.0163 (57.16)
-0.0434 (69.81)
Obs C o h o r t
Effects 111.283 Yes
107.560 Yes
Fama MacBeth Average Estimates Intercept Log (Age) Log (Age2)
0.0365
(1.75) 0.0087
(1.01) -0.0212
P-B7) -0.2694
(10.07)
-0.5011
(7.34) -0.4116
(6.81) -0.5861
-0.0485
(0S1)
0.2489
(7.32;
0.2061
16.87) 0.2821
(7.61) 0.0517
<1-89)
Log«ME, Log.BVME, £ £ « £ £ &
0.0197 0.0028 (11.35) (0.70) 0.0696 0.1001 (10.46) (6.33;
-00147 f9.7?J
-0.0311
0.0317
0.0365
0.1753
0.2184
Avg # of Cohort Firms Effects 2647 No
2647
2562
2562
Yes
Yes
Yes
Rsturn on Asssts in Excess to Industry by Firm Listing Dscods Cohort Effects
Dividend Yield (Dividends / MV Equity)
Firm Age
4<= (5-7) (8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
Mean
0.007 0.007 0.007 0.009 0.011 0.015 0.021 0.028 0.032 0.033
SD
0.015 0.014 0.015 0.016 0.017 0.019 0.021 0.023 0.023 0.021
Bottom 5%
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.004 0.006
Med
0.002 0.002 0.000 0.001 0.004 0.008 0.016 0.025 0.030 0.031
Top 95%
0.022 0.019 0.022 0.029 0.034 0.039 0.050 0.060 0.064 0.063
Mean Excess -0.008 -0.008 -0.007 -0.006 -0.004 -0.003 0.001 0.005 0.007 0.007
Med Excess -0.009 -0.009 -0.008 -0.008 -0.007 -0.005 -0.001 0.004 0.006 0.007
Avg # of Firms
81 178 380 425 364 346 461 358 312 272
Dividend Yield Age, Listing Cohorts, & Year Effects Decomposition
Intercept
-0.0111 (18.97) -0.0119 (15.45) -0.0154 (18.03)
Log (Age)
0.0038 (58.32; 0.0025 (36.42; 0.0033 (24.64;
Log (ME)
0.0014 (48.63; 0.0012 (1196;
Log (BVME)
0.0033 (56.07) -0.0050 (25.60)
Log (Age) Loo (ME)
0.0001 (4.03;
Log (Age) Log (BVME)
0.0027 (44.57)
R2
0.0902
0.1193
0.1335
Obs
133,452
129,397
129,397
Cohort Effects
Yes
Yes
Yes
Fama MacBeth Average Estimates
Intercept
-0.0320 (18.04) -0.0093 (6.76;
-0.0124 (7.34;
-0.0149 (6.15)
Log (Age)
0.0533 (13.38) 0.0043 (12.80) 0.0026 (10.79) 0.0028 (5.71)
Log (ME)
0.0019 (8.52; 0.0016 (3.80;
Log (BVME)
0.0053 0.00
-0.0019 (1.34)
Log (Age) Log (ME)
0.0001 (14.32)
Log (Age) Log (BVME)
0.0022 (18.28)
R2
0.0914
0.1066
0.1592
0.1689
Avg # of Firms 3174
3174
3083
3083
Cohort Effects
No
Yes
Yes
Yes
Dividend Yield in Excess to Industry by Firm Age
Listing Decade Cohort Effect*
3-S49 WO-1959 1960-069 «70- WT9 W60-19B9 B90- B99
1971 1976 SS I <90£ 1991 1996 2001 2006
Tab
le 2
.10:
In
dust
ry C
once
ntra
tion
and
Firm
Age
The
tab
le c
ompu
tes
two
indu
stry
con
cent
ratio
n m
easu
res
follo
win
g H
ou a
nd R
obin
son
and
rela
tes
expe
cted
ret
urns
to
firm
age
. In
dust
ry c
once
ntra
tion
is
prox
ied
by c
ompu
ting
the
Her
find
ahl
inde
x of
fir
m a
sset
s an
d sa
les
usin
g th
ree-
digi
t SI
C n
umbe
rs t
o de
term
ine
indu
stry
mem
bers
hip.
We
aver
age
the
estim
ated
Her
find
ahl
inde
x ov
er t
hree
yea
rs t
o en
sure
tha
t po
tent
ial
data
err
ors
do n
ot i
nflu
ence
the
res
ults
. Fr
om 1
965
to 2
006,
Pan
el A
pre
sent
s re
turn
s fo
r in
dust
ry c
once
ntra
tion
& f
irm
age
equ
al w
eigh
ted
port
folio
s, w
e pa
rtiti
on t
he d
ata
in t
hree
ind
ustr
y co
ncen
trat
ion
grou
ps b
ased
on
Sale
s an
d th
en
furt
her
divi
des
them
acc
ordi
ng t
o fi
rm A
ge.
Pane
l B
for
ms
indu
stry
por
tfol
ios
base
d on
fir
m a
sset
s H
erfi
ndah
l. B
oth
pane
ls p
rese
nt a
vera
ge m
onth
ly r
etur
ns,
stan
dard
de
viat
ions
, t-s
tatis
tics,
Sha
rpe
rati
os, e
xces
s re
turn
s to
Fam
a an
d Fr
ench
por
tfol
ios,
ave
rage
fir
m m
arke
t va
lue,
and
ave
rage
num
ber
of f
irm
s.
Pan
el A
: E
qu
al W
eig
hte
d P
ort
foli
os
So
rted
by
Sal
es H
erfi
nd
ahl
and
Ag
egro
up
c
•D
ra
= &
I!
-1
C
o o
Ind. itration
S 3
E
c u c o
~° I
n
C
2
s> §
x
§ u
Fir
m A
ge
Avg
Ret
urn
S
t. D
ev
t S
har
pe
Rat
io
FF E
xces
s R
etur
n
f-
FF
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f Fi
rms
Avg
Ret
urn
S
t. D
ev
t S
har
pe
Rat
io
FF E
xces
s R
etu
rn
f-
FF
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f F
irm
s
Avg
Ret
urn
S
t. D
ev
t S
har
pe
Rat
io
FF
Exc
ess
Ret
urn
f-
F
F E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Firm
s
7<=
1.39
%
0.09
42
3.31
0.
513
0.01
%
0.03
26
5 10
6
0.99
%
0.08
92
2.49
0.
386
-0.3
7%
-2.3
1 20
7 96
1.08
%
0.08
26
2.93
0.
455
-0.2
0%
-1.4
5 20
0 84
(8-1
2)
1.62
%
0.08
23
4.40
0.
683
0.28
%
2.00
24
6 14
3
1.44
%
0.08
12
3.96
0.
614
0.10
%
0.85
20
1 13
6
1.25
%
0.07
76
3.59
0.
557
-0.1
3%
-1.1
7 23
6 10
2
(13-
18)
1.73
%
0.07
12
5.42
0.
841
0.39
%
4.32
31
7 15
6
1.61
%
0.07
95
4.51
0.
700
0.16
%
14
9 34
7 14
3
1.64
%
0.07
25
5.07
0.
786
0.25
%
2.85
28
4 11
8
(19-
25)
1.58
%
0.06
66
5.28
0.
819
0.26
%
2.95
77
4 12
2
1.58
%
0.07
08
4.97
0.
772
0.21
%
2.41
55
5 11
9
1.68
%
0.06
61
5.66
0.
879
0.27
%
3.38
42
1 11
0
(26-
35)
1.59
%
0.05
81
6.11
0.
948
0.25
%
3.29
10
31
100
1.62
%
0.06
61
5.46
0.
848
0.17
%
2.13
58
9 11
2
1.59
%
0.06
21
5.71
0.
887
0.18
%
2.41
57
0 12
2
(36-
55)
1.41
%
0.04
92
6.38
0.
990
0.08
%
1.12
99
4 12
2
1.39
%
0.05
75
5.41
0.
839
-0.0
1%
-0.2
1 69
1 14
4
1.49
%
0.05
42
6.73
0.
952
0.09
%
1.49
61
5 17
5
(56-
75)
1.23
%
0.04
29
6.38
0.
991
-0.1
2%
-1.3
9 14
94
103
1.23
%
0.05
07
5.40
0.
838
-0.1
3%
-1.8
3 14
43
117
1.31
%
0.05
25
5.55
0.
861
-0.0
4%
-0.6
0 14
54
125
(76-
100)
1.15
%
0.03
90
6.56
1.
018
-0.1
6%
-1.5
1 16
14
97
1.28
%
0.04
95
5.78
0.
897
-0.0
1%
-0.1
3 22
33
95
1.26
%
0.05
00
5.62
0.
872
-0.0
7%
-0.8
8 19
15
112
+100
1.21
%
0.04
08
6.63
1.
029
-0.0
5%
-0.5
6 36
66
93
1.20
%
0.04
92
5.43
0.
843
-0.0
5%
-0.6
9 43
48
80
1.26
%
0.04
84
5.82
0.
903
0.03
%
0.33
42
94
91
Mat
ure
-Y
ou
nq
0.33
%
0.05
54
13
5 0.
209
0.38
%
15
7 29
1 26
2
0.58
%
0.04
12
3.76
0.
491
0.58
%
3.47
38
1 21
5
0.59
%
0.03
87
3.47
0.
530
0.47
%
2.94
31
0 19
5
Mat
ure
- O
ld
0.52
%
0.05
16
2.24
0.
348
0.44
%
3.00
19
92
249
0.38
%
0.04
16
2.04
0.
317
0.27
%
2.77
24
52
199
0.41
%
0.03
51
2.63
0.
408
0.24
%
2.36
23
58
201
All
Fir
ms
1.41
%
0.05
66
5.56
0.
863
0.11
%
2.50
97
9 10
41
1.36
%
0.06
18
4.93
0.
765
0.02
%
0.72
10
30
1043
1.39
%
0.05
82
5.33
0.
827
0.05
%
7.39
10
71
1039
Pan
el B
: E
qu
al W
eig
hte
d P
ort
folio
s S
ort
ed b
y A
sset
Her
find
ahl a
nd
Ag
egro
up
c 0 •o
is
c 2
o
o o
o c 0
c 2
•o
£ 2
c o
d. ition
"I
Ȥ
x c o
Fir
m A
ge
Avg
Ret
urn
S
t. D
ev
t S
har
pe
Rat
io
FF
Exc
ess
Ret
urn
t-
F
F
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f F
irm
s
Avg
Ret
urn
S
t. D
ev
f S
har
pe
Rat
io
FF E
xces
s R
etu
rn
t-
FF
E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Fir
ms
Avg
Ret
urn
S
t. D
ev
/ S
har
pe
Rat
io
FF E
xces
s R
etu
rn
t-
FF
E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Fir
ms
7<=
1.33
%
0.09
08
3.27
0.
508
-0.0
8%
-0.3
8 26
4 10
9
1.01
%
0.08
64
2.61
0.
405
-0.3
1%
-1.9
8 20
8 99
1.13
%
0.08
33
3.04
0.
472
-0.1
6%
-1.1
6 19
8 79
(8-1
2)
1.52
%
0.08
12
4.18
0.
649
0.18
%
13
3 21
8 14
5
1.45
%
0.08
17
3.96
0.
615
0.13
%
10
7 23
9 13
5
1.35
%
0.07
76
3.90
0.
605
-0.0
5%
-0.4
2 22
6 10
1
(13-
18)
1.73
%
0.07
12
5.42
0.
842
0.38
%
4.37
33
1 15
3
1.65
%
0.07
73
4.76
0.
739
0.21
%
2.06
33
8 14
8
1.65
%
0.07
31
5.03
0.
782
0.25
%
2.72
27
6 11
6
(19-
25)
1.62
%
0.06
74
5.37
0.
834
0.31
%
3.53
78
2 11
5
1.62
%
0.06
92
5.23
0.
811
0.26
%
2.75
53
2 13
0
1.59
%
0.06
67
5.33
0.
828
0.18
%
2.18
4
36
106
(26-
35)
1.58
%
0.05
80
6.09
0.
945
0.25
%
3.08
92
2 97
1.67
%
0.06
74
5.53
0.
859
0.21
%
2.27
69
2 11
4
1.52
%
0.06
26
5.43
0.
842
0.10
%
13
4 57
8 12
3
(36-
55)
1.38
%
0.04
91
6.29
0.
976
0.06
%
0.70
93
9 12
2
1.48
%
0.05
71
5.76
0.
895
0.08
%
14
3 70
7 14
6
1.43
%
0.05
55
5.74
0.
891
0.02
%
0.30
6
44
173
(56-
75)
1.19
%
0.04
36
6.10
0.
946
-0.1
2%
-13
2 17
23
103
1.29
%
0.05
10
5.65
0.
878
-0.0
7%
-1.1
3 14
16
114
1.28
%
0.05
24
5.43
0.
843
-0.0
9%
-12
9 13
84
129
(76-
100)
1.17
%
0.03
95
6.63
1.
028
-0.1
2%
-1.1
0 16
15
93
1.26
%
0.04
85
5.87
0.
901
-0.0
5%
-0.6
5 20
19
95
1.23
%
0.05
10
5.39
0.
837
-0.0
8%
-1.1
1 20
97
115
+10
0
1.22
%
0.04
14
6.60
1.
024
-0.0
1%
-0.1
5 44
18
92
1.21
%
0.04
81
5.67
0.
871
-0.0
8%
-1.0
3 40
42
79
1.23
%
0.04
93
5.57
0.
865
0.00
%
0.04
41
88
94
Mat
ure
-Y
ou
ng
0.40
%
0.05
06
1.76
0.
273
0.46
%
2.12
29
8 26
2
0.64
%
0.03
93
3.64
0.
565
0.52
%
3.04
27
3 24
7
0.52
%
0.03
72
3.09
0.
480
0.42
%
2.62
23
7 19
4
Mat
ure
- O
ld
0.51
%
0.05
11
2.21
0.
343
0.40
%
2.80
23
75
245
0.44
%
0.04
86
2.02
0.
313
0.29
%
2.05
21
90
226
0.42
%
0.04
21
2.22
0.
344
0.25
%
2.05
22
32
209
All
Fir
ms
1.38
%
0.0
56
8 5.
44
0.8
44
0.0
9%
1.
95
1018
10
30
1.42
%
0.0
62
0 5.
11
0.7
93
0.0
8%
2.
08
952
1059
1.36
%
0.0
58
8 5.
17
0.8
03
0.01
%
0.20
11
10
1034
Tab
le 2
.11:
Cap
ital
Inv
este
d T
urno
ver
and
Fir
m A
ge
The
tab
le r
elat
es e
xpec
ted
retu
rns
to f
irm
age
and
Cap
ital
Inv
este
d T
urno
ver,
a m
easu
re f
or f
irm
eff
icie
ncy.
F
irm
s so
rted
int
o th
ree
Inve
sted
Cap
ital
Tu
rno
ver
ca
tego
ries
(lo
w,
med
ium
, an
d hi
gh)
and
Fir
m A
ge.
Fro
m 1
965
to 2
00
6,
the
pane
l pr
esen
ts t
he a
vera
ge e
qual
wei
ghte
d m
onth
ly r
etur
ns,
stan
dard
dev
iati
on
s, t
-st
atis
tics
, S
harp
e ra
tios
, ex
cess
ret
urn
s to
Fam
a an
d F
renc
h po
rtfo
lios
, av
erag
e fi
rm m
arke
t va
lue,
and
ave
rage
num
ber
of f
irm
s.
Equ
al W
eigh
ted
Por
tfolio
s S
orte
d b
y C
apita
l Inv
este
d T
urno
ver
and
Age
grou
p
>
o
c 3 O >
o
c 3 1—
•D
w
0)
>
o
c 3 .C Hig
Fir
m A
ge
Avg
Ret
urn
f
Sh
arp
e R
atio
F
F E
xces
s R
etu
rn
f-
FF
E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Fir
ms
Avg
Ret
urn
t
Sh
arp
e R
atio
F
F E
xces
s R
etu
rn
f-
FF
E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Fir
ms
Avg
Ret
urn
t
Sh
arp
e R
atio
F
F E
xces
s R
etu
rn
f-
FF
E
xces
s A
vg M
arke
t V
alu
e A
vg #
of
Fir
ms
7<=
1.13
%
3.37
0.
514
-0.2
2%
-7
.23
34
2 59
1.28
%
3.76
0.
491
-0.0
5%
-0
.27
28
2 37
1.54
%
3.5
7 0
.55
4 0
.21
%
0.98
20
5 4
1
(8-1
2)
1.18
%
3.56
0
.55
2 -0
.12
%
-7.0
4 35
8 94
1.52
%
4.07
0
.63
2 0
.12
%
0.84
28
6 81
1.61
%
4.37
0
.67
8 0
.22
%
7.66
22
0 87
(13-
18)
1.49
%
4.94
0
.76
7 0
.13
%
7.30
4
55
106
1.60
%
4.67
0
.72
5 0
.22
%
2.07
32
4 10
3
1.80
%
5.78
0
.80
5 0
.41
%
3.82
33
2 12
2
(19-
25)
1.42
%
5.7
7 0
.80
3 0
.10
%
7.09
6
69
90
1.58
%
5.72
0
.79
5 0
.21
%
2.28
62
0 10
1
1.76
%
5.45
0
.84
5 0
.35
%
3.94
62
2 11
3
(26-
35)
1.30
%
5.26
0.
816
0.0
0%
-0
.01
1177
79
1.61
%
5.59
0.
867
0.2
3%
2.
67
808
107
1.66
%
5.78
0
.89
7 0
.23
%
3.03
4
69
116
(36-
55)
1.20
%
5.59
0
.86
8 -0
.10
%
-7.3
7 11
19
113
1.41
%
5.59
0
.86
8 0
.00
%
-0.0
4 6
99
136
1.60
%
6.37
0
.98
9 0
.19
%
2.96
6
22
159
(56-
75)
1.15
%
6.07
0
.94
2 -0
.13
%
-7.5
0 17
65
110
1.24
%
5.27
0
.80
9 -0
.15
%
-2.2
3 14
75
113
1.29
%
5.66
0
.87
9 -0
.08
%
-0.9
7 13
06
104
(76-
100)
1.11
%
6.46
1.
002
-0.1
5%
-7
.63
2579
11
0
1.25
%
5.46
0
.84
8 -0
.08
%
-1.0
5 18
51
104
1.36
%
6.07
0
.93
4 0
.01
%
0.73
12
06
78
+10
0
1.19
%
6.42
0
.99
6 -0
.05
%
-0.6
0 40
16
112
1.18
%
5.50
0
.85
4 -0
.06
%
-0.7
7 52
62
91
1.31
%
5.77
0
.88
6 0
.03
%
0.34
2
88
7 52
Mat
ure
-Y
ou
ng
0.36
%
7.74
0
.27
1 0
.35
%
7.78
3
98
164
0.3
2%
7.
60
0.2
48
0.2
7%
7.
36
30
3 13
9
0.26
%
7.73
0.
175
0.2
0%
0.
92
26
9 16
3
Ma
ture
-O
ld
0.2
9%
7.
55
0.2
40
0.1
8%
7.
43
22
35
21
7
0.4
2%
7.
95
0.3
03
0.2
8%
7.
89
27
93
194
0.4
9%
2.
33
0.3
62
0.3
8%
2.
45
1609
17
4
All
Fir
ms
1.18
%
5.38
0
.83
5 -0
.10
%
-2.7
4 14
95
872
1.41
%
5.34
0
.82
8 0
.05
%
1.60
12
94
872
1.59
%
5.75
0
.89
2 0
.21
%
5.79
7
65
87
2
Tab
le 2
.12:
"V
alue
Cre
atio
n" a
nd F
irm
Age
(19
85 -
2006
)
The
tab
le r
elat
es e
xpec
ted
retu
rns
to f
irm
age
and
the
spr
ead
betw
een
Ret
urn
on I
nves
ted
Cap
ital
and
the
Cos
t of
Cap
ital
(WA
CC
). R
etur
n on
Inv
este
d C
apita
l (R
OIC
) is
com
pute
d by
mul
tiply
ing
Inve
sted
Cap
ital
Tur
nove
r w
ith N
et O
pera
ting
Pro
fit
Afte
r T
ax P
rofi
t (N
OPA
T).
Fir
ms
with
hig
h m
argi
ns a
nd t
urno
ver
will
hav
e hi
gh R
OIC
. W
e co
mpu
te t
he w
eigh
ed a
vera
ge c
ost
of c
apita
l (W
AC
C)
usin
g bo
nd d
ata
on F
ISD
to
estim
ate
the
cost
of
debt
(w
idel
y av
aila
ble
afte
r 19
85),
tax
rate
is
estim
ated
fol
low
ing
Gup
ta a
nd N
ewbe
rry,
and
cos
t of
equ
ity i
s es
timat
ed u
sing
a th
ree
fact
or m
odel
(M
arke
t, S
MB
, and
HM
L).
T
he s
prea
d be
twee
n R
OIC
and
WA
CC
can
be
seen
a p
roxy
for
val
ue c
reat
ion.
Fro
m 1
985
to 2
006,
eve
ry J
une
we
sort
sto
cks
in t
hree
"V
alue
Cre
atio
n" c
ateg
orie
s (l
ow,
med
ium
, an
d hi
gh)
and
Firm
Age
. T
he P
anel
pre
sent
s eq
ual
wei
ghte
d m
onth
ly r
etur
ns,
stan
dard
dev
iatio
ns,
t-st
atis
tics,
Sha
rpe
rati
os,
exce
ss r
etur
ns t
o Fa
ma
and
Fren
ch p
ortf
olio
s, a
vera
ge f
irm
mar
ket
valu
e, a
nd a
vera
ge n
umbe
r of
fir
ms.
1 J "8 i •s
X
"g 1 I o a 1 ?
s c •a £ ID i r ' c $ s o 3 r
Fir
m A
ge
Avg
Ret
urn
f S
har
pe
Rat
io
FF E
xces
s R
etu
rn
f-
FF
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f Fi
rms
Avg
RO
IC -
WA
CC
Avg
Ret
urn
f
Sh
arp
e R
atio
F
F E
xces
s R
etu
rn
t-
FF
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f F
irm
s A
vg R
OIC
-WA
CC
Avg
Ret
urn
t
Sh
arp
e R
atio
F
F E
xces
s R
etu
rn
t-
FF
Exc
ess
Avg
Mar
ket
Val
ue
Avg
# o
f Fi
rms
Avg
RO
IC-W
AC
C
Equ
al W
eigh
ted
Por
tfol
ios
Sor
ted
by
7<=
0.94
%
17
5 0.
378
-0.3
9%
-1.7
0 31
9 10
0 -2
6.78
%
0.94
%
2.49
0.
536
-0.3
7%
-2.3
0 45
3 60
0.
93%
1.63
%
3.90
0.
841
0.44
%
2.84
63
3 57
19
.71%
(8-1
2)
1.48
%
2.75
0.
594
0.12
%
0.60
27
2 16
0 -2
9.31
%
1.23
%
3.07
0.
661
-0.1
5%
-1.1
6 46
1 10
7 0.
90%
1.51
%
3.47
0.
749
0.32
%
2.33
72
5 12
9 19
.98%
(13-
18)
1.61
%
3.45
0.
743
0.22
%
1.51
34
7 18
1 -2
5.03
%
1.59
%
4.73
0.
890
0.20
%
19
2 46
1 13
4 0.
89%
1.62
%
4.75
0.
896
0.41
%
3.54
99
5 17
7 20
.45%
(19-
25)
1.57
%
3.50
0.
754
0.15
%
7.08
60
7 14
9 -2
0.31
%
1.50
%
4.23
0.
912
0.11
%
7.70
66
6 13
4 1.
01%
1.45
%
4.00
0.
863
0.22
%
2.79
19
99
160
19.0
7%
"Val
ue
Cre
atio
n'
(26-
35)
1.60
%
4.37
0.
930
0.14
%
7.37
65
7 11
6 -1
6.47
%
1.52
%
4.73
1.
020
0.11
%
7.70
94
4 13
6 1.
13%
1.53
%
4.75
1.
024
0.28
%
2.77
20
04
158
17.8
7%
(36-
55)
1.57
%
4.85
1.
045
0.14
%
7.39
89
4 11
2 -1
4.08
%
1.39
%
4.86
1.
049
-0.0
6%
-0.5
7 11
20
166
1.28
%
1.44
%
5.07
1.
094
0.17
%
7.39
16
50
178
16.1
8%
' (R
OIC
-
(56-
75)
1.11
%
3.60
0.
777
-0.3
2%
-2.4
7 15
21
89
-12.
03%
1.19
%
4.65
1.
002
-0.2
3%
-7.8
7 21
32
136
1.20
%
1.21
%
4.65
1.
003
-0.0
6%
-0.4
8 33
59
117
14.9
3%
WA
CC
) an
d A
gegr
oup
(76-
100)
1.15
%
4.27
0.
909
-0.2
8%
-7.9
5 20
39
85
-11.
30%
1.17
%
4.67
1.
007
-0.2
2%
-7.7
3 27
41
141
1.15
%
1.35
%
5.26
1.
135
0.06
%
0.44
37
02
110
13.6
7%
+100
1.28
%
4.49
0.
969
-0.1
0%
-0.7
0 42
46
90
-10.
69%
1.22
%
4.99
1.
076
-0.1
1%
-0.8
6 62
91
136
1.02
%
1.22
%
5.07
1.
080
0.01
%
0.08
11
070
112
15.0
8%
Mat
ure
-Y
ou
na
0.66
%
3.72
0.
672
0.61
%
2.86
33
3 28
2 -2
5.90
%
0.65
%
3.44
0.
742
0.57
%
3.73
45
7 19
4 0.
91%
-0.0
1%
-0.0
7 -0
.016
-0
.03%
-0
.79
814
235
20.0
8%
Mat
ure
-O
ld
0.32
%
0.93
0.
200
0.31
%
7.35
22
96
271
-17.
86%
0.37
%
7.44
0.
311
0.30
%
7.85
33
76
271
0.95
%
0.40
%
7.53
0.
329
0.40
%
2.16
60
32
289
17.7
7%
All
Fir
ms
1.44
%
3.64
0.
785
0.03
%
0.47
98
2 10
83
-20.
09%
1.33
%
4.49
0.
969
-0.0
6%
-0.9
9 18
19
1152
1.
07%
1.46
%
4.66
1.
006
0.22
%
2.92
27
40
1199
17
.58%
Table 2.13: New Duration Estimates
The table examines the relation of firm age with respect to the new re-estimated equity duration measure. Our new re-estimated equity duration follow Dechow, Sloan, and Soliman however instead of assuming mean reversion in sales growth to compute expected sales during the next 10 years, we utilize sales growth estimates based on last year growth and current firm age combined with industry sales growth regression coefficients. We present two panels: The first panel presents the average from 1965 to 2006 of yearly mean, standard deviation, bottom 5th and top 95th
percentiles, and median estimates of the fundamental measure. In addition the panel presents the measure's excess to 48-industry (columns 7 and 8 of the panel). The second panel presents the results of the regression decomposing the measure's excess to the industry by firm age, listing cohorts, and year effects following Deaton 1997, and the Fama MacBeth average estimates for yearly cross-sectional regressions. Finally, for each measure we present graphs of the measure's decomposition into age, listing cohort, and year effect. As a robustness check in all our regressions we add the firms size (log(ME)) and book-to-market ratio (log(BVME)) and the interact those variables with the age (log(Age)) variable. T statistics are reported under the coefficient estimates in parenthesis.
Firm Age
4<= (5-7)
(8-12) (13-18) (19-25) (26-35) (36-55) (56-75) (76-100)
+100
Mean
35.85 28.89 23.36 22.23 19.78 17.26 15.64 14.60 14.04 13.87
SD
19.714 13.843 9.947 12.720 8.909 4.646 3.039 2.820 2.717 2.605
New Bottom
5% 16.51 11.02 8.97 14.17 14.10 12.86 11.63 10.38 9.39 9.33
Equity Duration
Med
31.23 25.29 21.52 18.88 17.57 16.48 15.43 14.60 14.20 14.15
Top 95%
68.94 57.77 41.14 52.00 34.12 24.74 20.47 18.61 17.43 16.83
Mean Excess 15.71 8.51 3.20 2.57 0.75 -0.86 -1.75 -2.17 -2.41 -2.51
Med Excess 11.98 6.23 2.82 1.01 0.03 -0.65 -1.34 -1.70 -1.82 -1.87
Avg # of Firms
38 116 287 347 312 300 406 323 291 255
New Equity Duration Age, Listing Cohorts, & Year Effects Decomposition
Intercept
-2.0804 (2-96) 0.9304 (1.34) 62430 (7.54)
Log (Age)
82.5687 (50.54; 83.5523 (51.62; 90.5326 (52.37)
Log (Age2)
-41.3000 (52.97)
-41.9234 (54 .30
-46.1532 (57.(2;
Log (ME)
-0.1295 (5.83;
-0.7119 (8.45)
Log (BVME)
2.1860 (47.I2J 11.4687 (68.99)
Log l«gei
Loo (ME)
0.1089 (4.74)
i-og («ge| Log (BVME)
-2.8991 (57.95)
R2
0.0912
0.1147
0.1468
Obs
112,250
112,223
112,223
Effects Yes
Yes
Yes
F a m a M a c B e t h A v e r a g e E s t i m a t e s
Intercept
-1.3708 (1.26)
-1.6264 (1.49) 0.6577 (0.84) 6.1581 (7.23)
Log (Age)
66.3318 (9.00)
69.0212 (9.26;
68.6816 (9.90)
739382 (11.30)
Log (Age2)
-33.3675 (9.29;
-34 6981 (9.54;
-34.6212 (10.19;
-38.0222 (11.81)
Log (ME)
-0.0739 (2.25;
-0.8514 (8.54;
Log (BVME)
1.3767 (5.85) 9.4568 (12.44}
Log (Age) Log (ME)
0.1718 (7.35)
Log (Age) Log (BVME)
-24462 (13.11)
R2
0.2462
0.2491
0.2759
0.3297
Avg * of Firms 2670
2670
2673
2673
Cohort Effects
No
Yes
Yes
Yes
Ass«t Growth in Excess to Industry by Firm Age Listing D*cad« Cohort Effect*
. 1«B 2O0O-JO08
128
2.8 References
1. Agarwal, Rajshree, and Dabid B. Audretch, 2001, "Does Entry Size Matter? The Impact of the Life Cycle and Technology on Firm Survival," The Journal of Industrial Economics, 49, 1, 21-43.
2. Agarwal, Rajshree, and Michael Gort, 2002, "Firm and Product Life Cycles and Firm Survival," The American Economic Review, 92, 2, 184-190.
3. Amihud, Yakov, 2002, "Illiquidity and Stock Returns: Cross-section and Time-series Effects," Journal of Financial Markets, 5, 31-56.
4. Altaian, Edward I., "Default Recovery Rates and LGD in Credit Risk Modeling and Practice: An Updated Review of the Literature and Empirical Evidence," Working Paper 2006.
5. Carhart, Mark M., 1997, "On Persistence in Mutual Fund Performance," Journal of Finance, 52, 57-82.
6. Daniel, Kent, Hirshleifer D., and Subrahmanyam A., 1998, "Investor Psychology and Security Market Under- and Overreactions", Journal of Finance, 53, 6, 1839-1885.
7. Daniel, Kent, Hirshleifer D., and Subrahmanyam A., 2001, "Overconfidence, Arbitrage, and Equilibrium Asset Pricing", Journal of Finance, 56, 3, 921- 965.
8. Daniel, Kent, and Sheridan Titman, 1997, "Evidence on the Characteristics of Cross Sectional Variation in Stock Returns," Journal of Finance, 52, 1, 1-33.
9. Daniel, Kent, and Sheridan Titman, "Testing Factor-Model Explanations of Market Anomalies," 2005 Working Paper.
10. Daniel, Kent, Grinblatt M, Titman S., and Wermers R., 1997, "Measuring Mutual Fund Performance with Characteristic-Based Benchmarks," Journal of Finance, 52, 3, 1035-1058.
11. Davis, James L., Eugene F. Fama and Kenneth R. French, 2000, "Characteristics, Covariances and Average Returns: 1929-1997," Journal of Finance, 55, 389-406.
12. Dealers' Digest Publishing Company, 1961, "Corporate Financing, 1950-1960" (Dealers' Digest Publishing Company, New York).
13. Dean, Arthur H., William Piel Jr., and Row H. Steyer, 1951, "Issuer Summaries: Securities Issues in the United States - July 26, 1933 to December31, 1949" (privately printed, New York).
129 14. Dechow, Patricia, Sloan R., and Soliman M., 2004, "Implied Equity Duration: A New Measure of
Equity Risk," Review of Accounting Studies, 9, 197-228.
15. Evans, David S. , 1987, "The relationship between firm growth, size, and age: Estimates for 100 Manufacturing Industries," Journal of Industrial Economics, 35, 4, 567-581.
16. Fama, E. and French K., 1992, "The Cross-Section of Expected Stock Returns," Journal of Finance, 47, 2, 427-466.
17. Fama, E. and French K., 1993, "Common Risk Factors in the Returns on Stock and Bonds," Journal of Financial Economics, 33, 1, 3-56.
18. Fama, E. and French K., 1996, "Multifactor Explanations of Asset Pricing Anomalies," Journal of Finance, 51, 1,55-84.
19. Fama, E. and French K., 1997, "Industry Cost of Equity," Journal of Financial Economics, 43, 153-193
20. Fama, E. and French K., 2004, "New Lists: Fundamentals and Survival Rates," Journal of Financial Economics, 73, 229-269.
21. Fama, E. and French K., 2006, "Profitability, investment and average returns," Journal of Financial Economics, 82, 491-518.
22. Fama, E.F., MacBeth, J., 1973, "Risk, Return, and Equilibrium: Empirical Tests," Journal of Political Economy, 81, 607-636.
23. Fink, Jason, Fink, K. E., Grullon, G. and Weston, J., working paper 2005, "IPO Vintage and the Rise of Idiosyncratic Risk."
24. Frankel, Richard, and Charles M.C. Lee, 1998, "Accounting Valuation, Market Expectation, and Cross-Sectional Stock Return," Journal of Accounting and Economics, 25, 283-319.
25. Gupta, S. and Newberry, K. 1997, "Determinants of the variability in corporate effective tax rates: evidence from longitudinal data," Journal of Accounting and Public Policy, 16, 1 -34.
26. Hall, B. H., A. B. Jaffe, and M. Tratjenberg, 2001, "The NBER Patent Citation Data File: Lessons, Insights and Methodological Tools." NBER Working Paper 8498.
27. Hillstrom, Roger, and Robert King, 1970, "/* Decade of Corporate and International Finance: 1960-1969" (Investment Dealers Digest, New York).
130 28. Hou, Kewei, and David Robinson, 2006, "Industry Concentration and Average Stock Returns," Journal
of Finance, 61, 4, 1927-1956.
29. "International Directory of Company Histories," St. James Press, Vols. 1 to 82.
30. Jovanovic, Boyan, 1982, "Selection and the Evolution of Industry," Econometrica, 50, 3, 649-670.
31. Jovanovic, Boyan and Rousseau, P. L., 2001, "Why Wait? A Century of Life Before IPO," AEA Papers and Proceedigs, 91, 2, 336-341.
32. Kaplan, Steven N., Sensoy B., and Stromberg P., 2005, "What are Firms? Evolution from Birth to Public Companies," Working Paper.
33. Kelley, M. Etna, 1954, "The Business Founding Date Directory", Morgan & Morgan Publishers, New York.
34. Klepper, Steven, 1996, "Entry, Exit, Growth, and Innovation over the Product Life Cycle," The American Economic Review, 86,3,562-583.
35. Lyon, J. D., Barber, B. M. and Tsai, C-L, 1999, "Improved Methods for Tests of Long-Run Abnormal Stock Returns," Journal of Finance, 54, 1,165-201.
36. Moody's Industrial Manuals, various dates (Moody's Investor Services, New York).
37. Mueller Dennis C , 1972, "A Life Cycle Theory of the Firm," The Journal of Industrial Economics, 20, 3, 199-219.
38. Ohlson, J.A., 1980, "Financial Ratios and the Probabilistic Prediction of Bankruptcy," Journal of Accounting Research 18, 109-131.
39. Pastor, Lobus, and Pietro Veronesi, 2003, "Stock Valuation and Learning about Profitability," Journal of Finance, 58, 5, 1749-1789.
40. Plesko, Geoge A., 2003, "An evaluation of alternative measures of corporate tax rates", Journal of Accounting and Economics, 35, 201 - 226.
41. Shumway, T., 1997, "The Delisting Bias in CRSP's Data," Journal of Finance, 52, 1, 327-340.
42. Shumway, T., and Warther, V. A., 1999, "The Delisting Bias in CRSP's Nasdaq and its Implications for the Size Effect," Journal of Finance, 54, 6, 2361-2379.
131
Appendix 2.A Variable Definitions and Construction
1) Firm Growth
• Asset Growth22: dAt I At_x = (At - At_x) / At_x, assets is given by Compustat's data6, and
restricts for assets to be > 0.
• Sales Growth: dS, / <Sf_j = (S, — iS",,, ) /£ ,_ , , sales is given by Compustat's datal2, and
restricts for sales to be > 0.
• Tobin's Q approximation (Market Value of Assets / Book Value of Assets(data6)), where
Market Value of Assets = Market Equity (data24 * data25) + Assets (data6) + Book Value
Equity.23
2) Innovation Edge
• R&D (data46) / Assets (data6): research and development expense as a percentage of total
assets, where assets are greater to zero.
• Patent Originality and Generality, based on citations made and received these measures proxy
the relative level of originality and generality of a patent. They are defined in detail on Hall,
Jaffe, and Tratjenberg 2001.
3) Process Efficiency
• Gross Margin: 1 - COGS(data41) / Sales(datal2).
• Net Operating Profit After Tax Margin (NOPAT): (EBITAD, - Taxes) / Sales,, where
EBITAD is defined as Pretax Income (data 170) - Special Items (data 17) + Interest Expense
(datal5) + Depreciation (datal4). Taxes are computed as Pretax Income times Effective Tax
Rate24
We also computed book value growth, however that limited the sample to firms with positive book values, while asset growth captures more generally firms growth even if at some instances the firms have negative book values.
Book Value of Equity is defined as in Fama & French 92, Shareholders Equity (data60) + Balance Sheet Deferred Taxes (data 74) + Investment Tax Credit (data51).
Effective Tax Rate is defined following Gupta and Newberry, Current Income tax Expense (data 16 -data50) over Book Income before interest and taxes (data 170 - data55 - data 17 + data 15)
132
• Invested Capital Turnover: Average of beginning and ending fiscal year of Sales, / (Assets, -
Current Liabilities, (data5) + Short Term Debt, (data 34) - Long Term Investments, (data31 +
data32) - Excess Cash, which is defined as cash in excess of 3% of Sales.
4) Liquidity and Cash Flow risk
• Illiquidity = [1000 * (ABS(ret) / (ABS(prc) * Vol))"2] , constructed using CRSP daily data,
following Amihud's illiquidity measures.
• Percentage of Non-Traded days, we consider a day as non-traded if volume is zero and if
volume is less than 100 and return is zero.
• Equity Duration, computed following Dechow, Sloan, and Soliman 2004. This measure can
be seen as an approximation of when equity is paid back, a longer (shorter) duration implies
longer (shorter) investment horizon until payback and therefore higher (lower) probability that
such cash flows will not be received.
5) Debt Structure and Default
• Ohlson's Default Probability, is the probability of default on debt, estimated at the end of
fiscal year t, from the logit regression model of Ohlson 1980.
• We used CRSP delisting code to separate delisting, merging, and other firms.
• Long Term debt (data 9) as percentage of total debt (data 9 + data 34)
6) Profitability
• Return on Assets = Earnings,25 / Assets,.i
• Dividends (data21) / Assets, dividend percentage normalized to assets.
Earnings = Earnings before ex-items + interest expense + deferred taxes, data 18 + data 15 + data50.
Appendix 2.B Credit Rating Numerical Conversion
We code the four rating agencies (FISD Rating Type) in the following form: FR (Fitch) = 1;
(Standard & Poors) = 3; and DPR = 4 (which we ignore);
If Rating Type is either Fitch or Standar Poors then:
'AAA+' = 10.33; 'AAA' = 10; 'AAA-' = 9.67;
•AA+' =9.33;'AA' = 9; 'AA-' = 8.67;
*A+' = 8.33; 'A' = 8; 'A-' = 7.67;
'BBB+' = 7.33; 'BBB' = 7; 'BBB-' = 6.67;
'BB+' = 6.33; 'BB* = 6; 'BB-' = 5.67;
'B+' = 5.33;'B' = 5;'B-'=4.67;
'CCC+' = 4.33; *CCC = 4; 'CCC-' = 3.67;
'CC+' = 3.33; 'CC = 3; 'CC-' = 2.67;
'C+' = 2.33;'C' = 2;'C-'=1.67;
'DDD' = 0.67; 'DD' = 0.33; 'D' = 0;
If Rating Type is MR then:
'Aaa' = 10; 'Aal' = 9.33; 'Aa2' = 9; 'Aa3' = 8.67;
'Al' = 8.33; 'A2' = 8; 'A3' = 7.67; 'A' = 8;
'Baal' = 7.33; 'Baa2' = 7; 'Baa' = 7; 'Baa3' = 6.67;
'Bal' = 6.33; 'Ba2' = 6; 'Ba' = 6; 'Ba3' = 5.67;
'Bl' = 5.33; 'B2' = 5; 'B' = 5; 'B3' = 4.67;
'Caal' = 4.33; 'Caa2' = 4; 'Caa' = 4; 'Caa3' = 3.67;
'Ca' = 3; 'C = 2; 'SUSP' = 0;
134
Chapter 3: The Rational Part of Momentum*
3.1 Introduction
The data in this paper suggest that the momentum effect (Jagadeesh and Titman,
1993, 2002) is closely related to changes in a measure of the fundamental value of a
company's equity. We show that the return of each stock return decile tracks the rate of
change in the decile's fundamental value, which we measure as a function of the change
in analysts' earnings estimates. This close co-movement of price and fundamental value
is one aspect of rationality: price should reflect fundamental value.
However, we also find another more subtle aspect of rationality. When we
redefine the momentum deciles to reduce errors in the change in fundamental value, the
stock return deciles appear to predict future changes in fundamental value. This contrasts
with the more frequently discussed result that analysts' estimates tend to predict future
changes in stock price . While we find some evidence of that, the more pronounced
effect works in the opposite direction; stock returns appear to predict future changes in
analysts' estimates.
While there are several possible explanations for this predictability and for the
momentum effect, we believe the most plausible is simple and rational. We conjecture
that informed investors predict changes in fundamental value and move stock prices
before less well informed investors obtain, and react to, the same information. Initially
the informed investors obtain new information and move stock price. Subsequently,
This chapter is from a co-written paper with James H. Scott
26 Ball and Brown, (1968). A growing literature suggests the momentum and PEAD anomalies are linked, e.g., Chan, Jagadeesh and Lakonishok (1999); Van Dijk and Huibers, 2002; Hong, Lee and Swaminathan (2003); Chordia and Shivakumar (JFE, 2006).
135
when the uninformed gain the same information their trades cause a further movement of
stock price in the same direction.
This explanation is inconsistent with Market Efficiency. In its strong form, the
Efficient Markets Hypothesis asserts that "at any time prices fully reflect all available
information" (Fama, 1970, p. 383). Clearly the momentum effect violates this since it
implies that past returns predict future returns. So too does our explanation, since we
believe that uninformed investors affect price, causing it to differ from what informed
investors believe it should be. As we discuss below, our view is not original, but is
consistent with many rational models of capital market equilibrium, and so provides a
rational explanation of the momentum effect.
For example, a heterogeneous expectations model (e.g., Fama and French, 2007,
Rubinstein, 1974, Lintner, 1969) can explain how past returns can predict both future
returns and future fundamental values. In particular, consider a model in which some
investors become "informed" in the sense that they deduce the correct mean of next
year's level of a stock's price. Other investors remain "uninformed" and have incorrect
estimates of the mean. As Fama and French argue the informed investors are likely to
have positive alpha's and the uninformed, negative alpha's (p. 673).
For example, if initially all investors assume that next period the expected price of
a stock is 11 and the appropriate discount rate is 10%, the current market price will be 10.
Next, to add heterogeneous expectations, assume that a subset of investors suddenly
discovers new information that leads them to believe the stock will have higher future
cash flows so that its price next period will have a mean of 14. If the market reopens,
136
then as long as some investors remain uninformed, the current stock price will not rise as
much as the informed investors believe is warranted.
How much the price increases will depend on the relative wealth, risk aversion
and different price expectations of both types of investors as well as their other portfolio
holdings. Suppose that the current price rises to 11.5, thereby creating a positive,
abnormal return. If next period's price equals 14, which is the mean of the informed
investors' distribution, a second abnormally positive return will follow the first.
In sum, if some investors become informed before others, one abnormal return
will be followed by another. That is, there will be a momentum effect. Further, since
prices are determined by discounting future cash flows, the cash flows expected by the
uninformed investors in one period will be the cash flows expected by the informed
investors in the previous period.
An appealing version of this type of equilibrium appears in "On the Impossibility
of Informationally Efficient Markets," (Grossman and Stiglitz, hereafter GS, 1980),
which we believe best captures the essence of what drives the momentum effect. They
begin by showing that, if information is costly, there is no market mechanism to ensure
that prices will fully reflect all available information. Stated differently, if the Efficient
Markets Hypothesis held and prices did reflect all information, there would no incentive
to collect costly information. Hence, prices will not reflect the information. Fama (1991,
p. 1575) acknowledged the point in his second review of the efficient markets literature.
"Since there are surely positive information and trading costs, the extreme version of the
market efficiency hypothesis is surely false."
137
In the noisy rational expectations equilibrium described by GS (see also Lucas,
1972), a subset of investors collects costly information because they expect to profit by
doing so. "Informed investors" in the GS model must be able to buy and sell at prices
that do not reflect the information they generate. When they uncover particularly good
news about a company they must be able to buy at a price that is, on average, less than
the price that fully reflects available information. Similarly, they must be able to sell a
bad-news stock before the bad news is fully incorporated in price. As a result, the prices
behave in a fashion similar to that described in the above example.
Since price partially reflects the information collected by the informed investors,
prices convey useful information. In the GS model, uninformed investors know this and
try to take advantage of it. However, they are hampered by the normal randomness of
prices (which GS model as supply shocks) and their own risk aversion. Although they
know that a stock whose price has just gone up is more likely to have a higher expected
return than a stock that has gone down, because price is an imperfect signal and because
many stocks have gone up, they respond with only a small purchase.
Notice that an uninformed investor in the GS model behaves like a momentum
investor to the extent that he is more likely to favor a stock that has recently gone up than
one that has gone down.
Hong and Stein (1999) present a theoretical model that, like GS, has two types of
investors. While their model is one with boundedly rational agents, rather than a rational
expectations one, Hong and Stein's "news watchers" are like the informed investors in
GS, while their momentum investors are similar to the uninformed investors in GS.
138
In practice, we believe that a highly competent subset of professional investors fill
the role of the informed investors in the above models. Most professional investors
predict future earnings, and many use present value techniques to predict future stock
prices. To the extent they are successful; they buy or sell stocks before changes in
fundamental value are fully incorporated in stock prices. Their trades move prices, but
not so much that prices fully reflect the information these investors have gathered and
interpreted.
Security analysts employed by brokerage firms provide the information to the
uninformed investors. Like professional investors, security analysts carefully research
companies and estimate earnings and future cash flows; but they trade very little on the
basis of the information they generate. Instead, they work to disseminate it widely and
the information they generate is subsequently impounded in prices. As we show below,
over the six month intervals we investigate, changes in our measure of analysts'
expectations are highly correlated with changes in stock price27.
To investigate the earnings/momentum link, we use a modified present value
framework to estimate changes in a stock's fundamental value. Our measure of
fundamental value is similar to but different from earlier present value models of equity
valuation (e.g., Edwards and Bell, 1961, and Miller and Modigliani, 1966), but, like
them, allows the use of security analysts' expectations.
We present the time path of the actual returns of different momentum deciles and
compare them to our estimates of the concurrent percent change in fundamental value for
each momentum decile. The similarity between actual returns and changes in
We show below that analysts' expectations also anticipate future price changes, but that effect is smaller than the correlation between changes in analysts' expectations and concurrent price change.
139
fundamental value is striking and supports our notion that the primary force underlying
the momentum effect is changing company fundamentals as reflected in analysts'
expectations.
This linkage casts doubt on the primacy of several behavioral and/or technical
explanations of momentum. In particular, prospect theory and the disposition effect
suggest investors sell winners too early and hold losers too long so that prices slowly
react to fundamentals and thus, significantly diverge from fundamentals (e.g., Kahneman
and Tversky, 1979; Shefrin and Stattman, 1985; Grinblatt and Han, 2005). In contrast,
our findings suggest that prices mirror current changes in fundamentals, and anticipate
future changes in fundamentals.
Other theories posit that momentum occurs because technical traders react to
unusual price movements by seeking to ride a trend and over-extrapolate underlying
fundamentals (e.g., Delong et al, 1990). Daniel et al (1998) suggest that the behavior of
over-confident investors causes over-extrapolation. Again, our work suggests that, by
and large, prices track and anticipate underlying fundamentals. While these
behavioral/technical explanations may account for some portion of the momentum effect,
they do not seem to be the driving force.
This paper shares similarities with recent research on futures prices by Boudoukh
et al (2007) who contest the view of Delong et al (1990). Delong et al argued that
fundamentals had little effect on the frozen concentrated orange juice futures market.
They based their argument on the fact that regressions relating futures prices to
temperature that appeared in Roll (1984) have low R2's.
140
Later research by Boudoukh et al show that when temperatures remain above 36°,
Delong et al are correct; temperature change has little effect on futures prices. However,
movements in temperature near and below the freezing point have a significant impact on
supply because they can destroy orange crops. These freezing temperatures also have
dramatic effects on price, just as economic theory would predict. In concert with our
findings, changing fundamentals drive prices.
In a similar way, separating stock returns into momentum deciles and focusing on
the extreme deciles forces a researcher to study stocks at a time when something highly
significant affected their prices. Neoclassical finance suggests the significant event
should relate directly to the valuation of future corporate cash flows. Behavioral finance
might allow for some impact by fundamentals but would hold that there should be
additional and significant behavioral effects that would be considered irrational in a
neoclassical context. We find that fundamentals appear to be the predominant force.
The findings of Hong, Lee and Swaminathan (2003) (HLS) can be interpreted as
added evidence supporting a noisy rational expectations equilibrium. In particular, if
information is costless and the price system highly informative, there is unlikely to be a
noisy rational equilibrium or a momentum effect. An important instance of this can
occur when corporate insiders, who do have costless information about their own
corporations, can trade freely on that information. HLS studied 11 countries. In those
countries where investor protection was low and corruption was high (and, presumably,
insider trading largely unimpeded), there was neither post-earnings-announcement-drift
nor a momentum effect, as the GS analysis would suggest.
141
Section 3.2 describes our estimate of the change in fundamental value. Section
3.3 discusses the data we use. Section 3.4 provides initial evidence suggesting the
potential of our measure of fundamental value to explain the momentum effect. Section
3.5 investigates the momentum effect directly by showing how prices and fundamental
values of different momentum deciles change together over time from a year before the
formation of the momentum deciles until a year and a half after. Section 3.6 presents a
similar story for deciles ranked on change in value. Section 3.7 provides a check to see
whether our results could be explained by assuming that analysts' expectations simply
track past rates of return. Section 3.8 tests momentum hypotheses based on behavioral
arguments and estimates how far into the future professional investors appear to predict
fundamental value. Section 3.9 contains concluding comments and final remarks on the
literature.
3.2 Representing Changes in Fundamental Value with Analysts'
Estimates
Since our tests concern rates of return, we do not need to measure fundamental
value directly. We only need to measure the rate of change in fundamental value. We
then relate that to the rate of change in price. However, in order to motivate our rate of
change measure we will begin with a multi-period dividend discount formula. In
particular, we assume that V, the fundamental value per share of a firm's equity is given
by
v=Y-^— [l] ' £ ( l + r)' M
142
Where Dit = annual dividends per share expected by the average or representative
investor, and r = risk-adjusted discount rate , or equivalently,
V, = - + 1±T + > " ", [21 ' 1 + r (1 + r)2 tl(l + r)1 LJ
Where Eit = annual earnings per share for the i,h firm in period t as expected by
the representative investor, and A.jt = dividend payout ratio as expected by the
representative investor, where we have assumed the first two X.'s are equal for simplicity.
We next assume that in many instances, news that effects fundamental value in a
cross-sectional analysis also effects expected earnings within in the next two years.
Implicit in this view is also the assumption that valuation-sensitive news that only affects
information beyond two years is less likely and often more difficult to assess, so its
impact on valuation is less.
Mechanically, denote the first two terms on the right hand side of [2] A and the
last term B. We assume B is proportional to A, i.e., B = yA, or
r,=(i + r) XEn + XEi2
1 + r (l + ry [3]
The discount rate used was fixed at 10%. Similar results were obtained using a fixed industry discount rate suggested by Fama & French, 1997, and with a time varying discount rate equaling 6% plus the 10 year Treasury bond rate.
143
We define Ritv as the rate of change in fundamental value, or smoothed earnings
estimates. In a cross-sectional of stock returns, macro factors, such as the level of the
overall stock market, will be common for all stocks. As a result we assume that, cross-
sectionally, Rjtv will be proportional to the change in a firm's fundamental value .
l+Rltv~(Vlt/VlM). [4]
Notice that since X and (1+y) appear multiplicatively in both numerator and
dominator of [4] they are not required in the definition of Rtv. In sum, we approximate
the cross-sectional change in fundamental value by changes in expectations about firm-
specific, near-term events.
Analysts commonly estimate earnings for the current fiscal year, the following
fiscal year, and sometimes more. They typically estimate a long-run growth estimate as
well. The time interval we use in estimating Rjtv is six months, i.e., in [4], t is six months
after t-1. Our measure requires estimates for earnings one year hence and two years
hence. In each month, for each firm in our sample we estimate expected earnings one
year hence and two years hence using the following formulas.
Eit+1 = w-FYl + (1 - w)-FY2 and [5]
29 For a practitioner-oriented view of Rjt
v consider the following: let Pmt represent the average prices of the stocks in the market at period t. Let V^ represent the average of Vj, from [2], for all the stocks in the market. Then Pmt/Vmt is a type of market P/E ratio. Assume that the corresponding P/E ratio for each stock moves proportionally with the market's P/E ratio. Equation [2] then implies
1+Ritv = (Vi/Vu.0[(Pm/Vmt)/(PmtyVmt.I)]. [4']
Since we use [4'] cross-sectionally, the market term is the [4'] is the same for every firm. This implies that in cross-sectional analysis l+R," is proportional to the firm specific term, (VJ/VJM), or equation [4] above.
144
Eit+2 - wFY2 + (l-w)-FY2-(l + LTG) [6]
Where Ejt is company i's expected earnings for period t constructed each month
with the weighted average of FY1 and FY2, the analyst average estimates30 for earnings
for fiscal year 1 and 2 respectively. LTG is the consensus long term growth31, and w is
the fraction of months remaining in the current fiscal year. Using [5] and [6], we create
monthly proxies for earnings one and two years ahead.
Defining Rjtv in this manner has several advantages. A six month interval allows
us to use a time interval used in many previous studies. A six month interval is also long
enough to allow meaningful changes in fundamental value.
In addition, smoothing earnings over two years allows us to update changing
expectations on a regular basis and track the relation between prices and expectations as
stocks in different momentum deciles evolve over time. The smoothing process also
makes it more likely that our measure will capture changes in relating to the firm's longer
run profitability. Thus, it helps mitigate issues relating to transitory earnings (Kothari,
2001).
However, because our measure of expectations looks out only two years, it may
miss the full impact of some important changes. For example, the discovery of an oil
field or a new drug may impact long-run fundamentals but may not affect near-term
30 We also computed very similar results using the median instead of the mean of FY1 and FY2.
31 When LTG is not available we assign the industry (three digit SIC) average long term growth rate, similar results obtained when assigning 0% to the missing LTG estimates.
145
earnings estimates. In Section V we use an averaging process to reduce error in our
measure.
3.3 Data and Variable Construction
Our sample is composed of firms listed in the Center for Research in Securities
Prices (CRSP) tapes for the period of 1985 to 2006 for which there are also security
analyst estimates of future earnings in the IBES database. To be consistent with previous
literature, we focus on common equity and exclude REIT's, ADR's, limited partnerships,
and closed-end funds. Since the tests in this paper are based on holding periods of six
months,32 we include in our sample only firms that have analyst estimates for both fiscal
year 1 and 2 six months before the formation date of our momentum deciles and six
months after formation.
Table 3.1 contains the characteristics of our sample. In an average every month
we cover 1,977 firms. Before 1990, the average number of firms was 1,036. It rose to
2,674 in the second half of the 90's and then decreased to 2,318 in the following decade.
In terms of size and book-to-market quintiles, the sample appears slightly skewed toward
smaller low book-to-market stocks. However, the apparent skew is largely attributable to
the fact that the Fama French breakpoints are based on NYSE stocks, while our sample
includes Nasdaq and AMEX stocks as well. Panel F shows that the sample covers a wide
range of industries. Over the sample period the percent of manufacturing and non
durable industries decreased slightly, while business equipment increased.
Test periods of 3, and 9 months were also conducted with similar results.
146
3.4 An Initial Look at the Evidence
In this section we show that, as expected, the momentum effect occurs in our data.
We also show the power of our rate of change in fundamental value to explain
contemporaneous returns.
First, both the momentum literature and the noisy rational expectations model
suggest that past returns should predict future returns. That is, bi in the following
regression should be positive (the momentum effect).
R i t=a, +b,Ri t., +e t [7]
Table 3.2 presents average monthly cross-sectional regression estimates, using six
month returns, so that Rjt represents one six month return and Rit.i represents the six
month return immediately preceding it with a one month lag between the two periods.
Given that the holding periods are 6 months long and the regressions are estimated every
month, the observations overlap. Therefore reported ^-statistics are computed with
Newey-West (1987) standard errors, using a lag length of one less than the holding-
period horizon. This statistic is designed to correct for moving average errors induced by
the overlapping observations. Equation [7] appears as the first regression in Panel A of
Table 2. The R2 is low 0.014, but the t Statistic is significant (5.19).
The remaining regressions in Panel A show that Rjt.2 is insignificant and Rit_3 has
a significant, but negative sign. The negative sign is consistent with the familiar reversal
effect in the momentum literature (e.g., Jegadeesh and Titman, 1993). We shall return to
it later in the discussion of momentum.
147
The second regression of Panel B shows that when returns are regressed against
last period's change in fundamental value, Ri.tV, the R2 is even lower (.0046), though the
t Statistic is significant at 2.52. However, when last period's return is included in the
regression, its coefficient remains significant while the coefficient on Rj,t-iv falls to a third
of its former value and becomes insignificant (t = .97). This suggests that much of the
information in Ri,t-iv is subsumed in past return. Finally, Panel C includes regressions
where return is regressed on concurrent R / .
Ri t=a2 + b2Ri tv + e2t. [8]
R2 equals .11 and the t Statistic on R;/ equals a highly significant 16.9. We
believe this suggests that investors, whose activities are reflects in return, and analysts,
whose expectations are reflected in R;/, are reacting to the same news about future
corporate profitability. Further, when past return is added to the regression, the
coefficient on past return becomes insignificant, suggesting that the information in past
return is largely subsumed by the current change in fundamental value.
A noisy rational expectations interpretation of these results is that in period t-1,
informed investors correctly anticipated much of the information contained in next
period's R / . The trading activity of these informed investors causes Rt.i to partially
reflect next period Rv. Then in period t, analysts publish their expectations, and
uninformed investors trade based on the now public information in Rjtv. This explains the
momentum effect.
148
In sum, Table 3.2 provides initial evidence consistent with the momentum effect,
the noisy rational expectations hypothesis and the power of our measure of fundamental
value to explain concurrent stock returns. The next section presents a more dramatic view
of these issues by tracing the time paths of both fundamental value and the returns of
stocks sorted into momentum deciles.
3.5 The Time Path of Momentum Deciles
3.5.A Returns
We create overlapping momentum deciles by ranking stocks each month
according to their trailing 6-month returns. We will call each 6-month ranking period,
Period 0, e.g., for the six month period beginning January, 2000 and ending June, 2000,
we rank the stocks as of their six month return at the end of June. We then calculate 6
month returns for the stocks in each Period 0 decile for five additional periods. The 6-
month period immediately proceeding the ranking period is Period -1 (in the example,
July, 1999 through December, 1999); the 6-month period before that is Period -2. We
skip one month between period 0 and the next 6-month period (Period 1), as is common
in the momentum literature, to avoid bid-ask bounce problems (as well as lags in
analysts' earnings changes). Immediately following Period 1 are the two final 6-month
periods, Periods 2 and 3. Counting the month we skipped after the ranking period, our
data covers 37 months for each momentum decile. Each month begins another
(overlapping) ranking period, In our example, the next Period 0 would start at the
beginning of February, 2000. To adjust for overlapping observations, we use Newey-
West standard errors to calculate our t-statistics.
149
If the data met the test of strong form market efficiency, and assuming each decile
is well-diversified, a plot of the momentum deciles would look like Figure 3.1. In Period
0, the stocks are ranked into deciles by returns. Therefore in the ranking period, the
deciles differ by construction. In an efficient markets' interpretation, the return
differences in Period 0 reflect the idiosyncratic returns of the underlying stocks,
presumably the result of a news event. The market should fully reflect the news event in
Period 0, and that news should not affect returns in other periods.
In periods other than the ranking month, each decile simply represents a more or
less random sample of 10% of the stocks in the market. In these non-ranking periods,
each decile should earn the market rate of return. Only in the ranking period should the
deciles, by construction, differ from the market return.
Before and after Period 0, the returns of all deciles should equal the average
return in the market. Table 3.3 presents the actual results using our data. Figure 3.2,
which is based on the first panel in Table 3.3, presents a graphical summary of the actual
returns of the momentum deciles. It is similar to Figure 3.1, and, visually, suggests rough
correspondence with market efficiency. But Figure 3.2 differs from Figure 3.1 in a
number of important respects. We will start with Period 1, the ranking period, and move
forward in time before considering the earlier periods.
Period 0 is the ranking period. As expected, deciling stocks on 6-month returns
results in huge return differentials. Return differentials this large are most likely the
result of investors reacting to significant new information. The heterogeneity of the
returns suggests that the news is company-specific, or at most industry-specific.
150
Period 1 shows the familiar momentum effect. The deciles in Period 1 line up
just as they did in Period 0. Stocks in decile 10 had higher returns than stocks in decile 9
and so on, monotonically down to decile 1. The difference in returns between deciles 10
and 1 is large, 8.1% per six month period and the t statistic of the difference is a
statistically significant 5.4.
A noisy rational expectations interpretation is that some investors became
informed and traded stocks in Period 0. Those stocks then subsequently out- or under-
performed in Period 1. We investigate this interpretation more fully below when we
present the changes in fundamental value over these periods.
By Period 2 the t statistic of top decile average return minus the bottom decile
return is negative and insignificant, and the returns within the deciles are roughly similar.
Deciles returns in Period 1 do not predict decile returns in Period 2. By the end of this
period, 13 months after the ranking period, the information that moved prices in Periods 0
and 1 seem to be fully imbedded in stock prices.
In Period 3, the familiar reversal effect is evident (as it was in the regressions of
Table 2). The decile returns are again monotonic, but in the opposite direction. Stocks
that 19 months earlier had the highest returns, now have the lowest and underperformed
the lowest momentum decile by a statistically significant -4.5% (t statistic = -3.7).
In part the reversal effect is probably due to survivorship. To be in the Period 3
analysis, a firm must survive until Period 3. The Period 3 returns in Table 3.3 and Figure
3.2, are not realizable returns, rather they are conditional returns, conditional on the
survival of the firm until Period 3. It is not surprising that bottom decile firms, who may
have flirted with bankruptcy, but survived, have high returns. However, survivorship
151
does not seem to explain the swan dive taken by the former high flyers in the top deciles.
In the next subsection, we show the close relation of the reversal effect to fundamental
value and resume discussion of survivorship issues.
Next, consider the returns in Period -1, the period before the ranking period. The
decile returns in Period -1 or any period before Period 0 are not realizable as an
investment strategy since they are based on a sort that occurs in Period 0. Nevertheless,
as we argued above, under the hypothesis of Market Efficiency as depicted in Figure 1,
the decile returns in Period -1 should be independent of those in Period 0. However they
are not, and the actual relationship between Period 0 and Period -1 might be called the
reverse momentum effect. The top decile resulting from the ranking in Period 0
outperforms the bottom decile in Period -1 by 4.7% (t statistic = 3.12). Though
somewhat peculiar, this reverse momentum effect would seem as great a challenge to
market efficiency as the more familiar momentum effect. On the other hand, a noisy
rational expectations interpretation of this reverse momentum effect is that, in Period -1 ,
informed investors successfully traded stocks that subsequently out- or under-performed
in Period 0.
Although the payoffs to these informed Period -1 investors seem large, they
reflect the mechanical construction of the momentum deciles constructed in Period 0.
Many stocks had returns as high (or low) as the extreme decile portfolios did in Period -1.
On average these stocks did not earn extreme returns in Period 0. We are simply looking
at the stocks that subsequently did earn high returns. Nonetheless, even for these stocks,
Market Efficiency implies there should be no reverse momentum effect.
152
Period -2 is odder still. A year before the ranking period, the extreme deciles all
had high returns while the middle deciles had below average returns, the difference
between top and bottom deciles is -2.8% but it is statistically insignificant. To cast
further light on the time path followed by the momentum deciles, the next subsection
shows the relation between momentum and R / , our measure of the change in
fundamental value.
The third panel of Table 3.3 shows that adjusting for risk has little effect on the
results. Excess returns relative to Fama French size and book-to-value portfolios show
the same results discussed above, although in most cases the t statistics, though
significant, are less so. For example, in Period 1 the excess return from top minus bottom
momentum deciles is 6.34% with a t-statistic of 4.01; while in Period 2 the alphas are
statistically insignificant, but in Period 3 they show a reversal effect of -3.52% with a
significant t-statistic of 3.04.
3.5.B Fundamental Value by Momentum Deciles
For each of the above momentum deciles, Figure 3.3 and the second panel of
Table 3.3 present each decile's average Rv, our estimate of the change in fundamental
value. They trace the path of the Rv's for each momentum decile from a year and a half
before the formation of each momentum decile (period -2) until a year and seven months
after formation (period 3).
If there were no predictability in the time series of the Rv's, Figure 3.3 would look
like Figure 3.1. However, though Figure 3.3 resembles Figure 3.2, it does not look like
Figure 3.1. The monotonicity of the Rv's in Periods 0 and 1 suggest predictability in the
153
time series of analysts earnings revisions and thus in our Rv's. Further, there seems to be
greater predictability in the high momentum deciles, which had above-average Rv's for
Periods -1,0, 1 and 2, or a little over two years.
The individual periods tell an interesting story. As before, we will begin with
Period 0 and work forward before considering the pre-ranking periods. In Period 0 the
Rv's for the different momentum deciles line up in exactly the same order as did their
stock returns. The highest momentum decile had the highest increase in fundamental
value exceeding that of the lowest decile by 56.2%. The difference is statistically
significant (t statistic = 28.3)33. This suggests that the primary driver of the large
differences in the Period 0 returns is captured by the estimated change in fundamental
value that also occurred in Period 0.
As mentioned in the previous section, there is modest evidence that analysts'
estimates, and thus our Rv's, may be affected by stock price changes. While some trend
following may account for a little of our measure of changing fundamental value, it is
unlikely to explain much. The increases (decreases) in fundamental value in the extreme
deciles in Period 0 appear too large for an analyst to justify if there is no fundamental
evidence supporting such dramatic changes.
Further, the monotonicity of the Rv's would require that most, if not virtually all
analysts, ignore fundamentals and change their earnings estimates to correspond to price
moves. Finally, if price movements were driving changes in analysts' expectations, we
would expect to find long discussions of momentum in analysts' reports. Instead, they
Compared with the actual returns for the difference in momentum deciles, this return is relatively low. This phenomenon is similar to the low values of "earnings response coefficients" in the accounting literature (Kothari, 2001). It also reflects that the extreme decile stocks are largely growth stocks as we show below.
154
focus on company prospects for future earnings, cash flow and the relation of price to
fundamentals.
Similarly, the changes in fundamental value in Period 1 mirror the returns in
period 1. However, the Period 1 fundamental returns are larger than the price returns in
Period 1, and are monotonically in line with the momentum deciles of Period 0. A noisy
rational expectations interpretation is that the price changes in Period 0 reflect not only
the change in fundamental value occurring in Period 0 but also the expectations by
informed investors that fundamental value will continue to change in Period 1. As the
information underlying those expectations becomes more widely apparent, analysts, and
thus Rv, reflect it. The informed investors benefit from their foresight and trading activity
as the returns in Period 1 validate their Period 0 trades (the momentum effect).
There may be a second, more mechanical reason, for the wider spread in the Rv
deciles in Period 1 that is related to the way we construct Rv. As equation [3] shows, our
measure depends only on expected earnings over the next two years. It may be that some
of information expected by analysts in Period 0 effect earnings shortly beyond two years.
By Period 1 some of that information affects our measure of Rv and causes the wider
dispersion of the Period 1 deciles. This argument may also affect Rv in Period 2.
A similar phenomenon, "prices lead earnings," is familiar in the accounting
literature (Kothari, 2001). There it refers to reported earnings, which are reported later
than the earnings expectations we analyze. Figure 3.3 suggests that "prices lead analysts'
expectations of future earnings." A final interpretation is that analysts expectations
simply follow past price change, an issue to which we return in Section 3.7.
155
In Period 2 the pattern of the Rv's, partially extends the pattern of Period 1, but
the extreme deciles have begun moving toward the mean. By Period 3, which ends over
a year and a half after the momentum deciles were established, a reversal in the Rv's is
evident, which corresponds with the reversal in their stock returns. The bottom four
deciles have an average Rv of 7.3%, while the top six have an average of 3.9%.
Fundamental value in Period 3 suggests that the reversal in price returns is not a
return to rational pricing after a period when prices overshoot fundamentals (as suggested
by e.g., Daniel et al, 1998). To the extent that the results are not due to survivorship
bias34, investors appear to react to changing fundamentals. The expectations of these
fundamentals may be over-extrapolations, but this data suggests the return reversal
mirrors changing fundamentals. Notice too that these changes in fundamental value
cannot be interpreted as the result of analysts changing their earnings expectations to
ratify past returns.
3.6 Ranking Stocks into Rv Deciles
In this section instead of ranking stocks by returns in Period 0, we rank them into
deciles based on Rv, the percent change in fundamental value. Then, just as we did in the
previous sections, we follow these deciles through time in terms of both Rv and R. If our
hypotheses are correct, a more powerful link between stock price and fundamental value
should be apparent when we sort the stocks into Rv deciles. This is because our measure
We were able to reduce the reversal effect to a degree by revising the initial way we did the study. Initially, to be included in any period, a firm had to be in the sample for the entire period. In our revision, we required only that a firm be in the sample during periods 0 and 1. We assumed any proceeds from the last price at which the firm traded were invested in the remaining stocks in its decile. Its revised Rv equaled its R. This reduced survivorship and reversal, but as is apparent in the Tables and Figures, the reversal remains.
156
of fundamental value, Rv, is based on near term earnings estimates and will be most
powerful when information about fundamentals changes near-term earnings (the next two
or three years).
When we ranked on returns, the extreme deciles likely contained, not only stocks
with large changes in near-term earnings estimates, but also stocks whose valuation-
sensitive information focused only on longer term cash flows. That is, the extreme
deciles in the R ranking likely contained stocks that either had won or lost long-term
contracts, had drugs or other products approved or not, found extensive mineral deposits,
etc. Since our measure is unlikely to respond to that type of information, we would
expect that in momentum rankings the link between stock price and fundamentals will
appear weaker than it actually is.
However, when we rank stocks by Rv, it is likely that stocks whose returns have
only been affected by changing long run fundamentals will be randomly distributed
among the deciles. In an Rv ranking, Rv is more likely to be the primary driver of stock
returns, and if the noisy rational expectations hypothesis is correct we should expect that
prices should appear to be a better predictor of changes in fundamental value. Note: Jim I
think that you saying that Rv is the driver could be interpreted as predictability????
Table 3.4 and Figures 3.4 and 3.5 show the time path of Rv and R. Figure 3.4
shows the path of fundamental value. In Figure 3.4 the changes in fundamental value
seem less predictable than in the momentum ranking. Here it looks as if analysts were
surprised, particularly in the extreme deciles. If the Rv's were independent of each other,
Figure 3.4 would look like Figure 3.1. However, although it is similar, the highest
deciles in Period 0 are also the highest in Periods -2 through Period 1. The lowest two
157
deciles in Period 0 were also the lowest in the two years beginning in Period -2.
However mirroring Figure 3.3, by Periods 1 and 2, the lowest decile in Period 0 is now
the highest in terms of changing fundamental value.
The reversal in the lowest decile appears sooner than was apparent in the
momentum sorts. In particular, the lowest decile bounces back in the period immediately
following the ranking period. Further, as Table 3.4 and Figure 3.5 show, the
corresponding stock return for that decile do not reverse until periods 2 and 3. This may
be related to excessive pessimism by analysts about bottom decile stocks in Period 0 or to
major earnings write-downs by managers, which analysts subsequently corrected in
Period 1. It is important to note that returns in Period 2 and 3 in excess of Fama French
portfolios are all statistically insignificant.
Figure 3.5 shows the returns corresponding to the value deciles and these returns
appear consistent with a noisy rational expectations market. Compared to Figure 3.2,
prices predict the forthcoming Period 0 changes sooner and to a larger extent.
Recall that the returns in Figure 3.2, where stocks were ranked on the returns
themselves, approximated market efficiency. On the other hand, Figure 3.5, which shows
the returns when stocks are ranked by the change in fundamental value, does not look like
market efficiency at all. High (low) ranking deciles in Period -2 continue to be high
(low) ranking deciles for the next three six month periods.
An interpretation of Figure 3.5 in the spirit of noisy rational expectations is that,
as early as Period -2, some informed investors predicted the large changes in fundamental
value that security analysts reported in Period 0. Then, in Period -1 , additional investors
158
became informed (and/or the previously informed investors become more confident)
causing Period -1 returns to deviate even more from average.
Security analysts then disseminate the information in Period 0 and the price
moves dramatically as most of the uninformed investors react. Finally, a few of the
uninformed do not react until Period 1, causing a small momentum effect in Period 1.
Simple correlations between decile returns are consistent with a stronger link
between current returns and future changes in value when using an Rv ranking. Consider
the correlations of R in one period with Rv one or two periods later. For the two period
ahead prediction, this indicates how well returns in one six month period, for clarity, say
January to June, predict changes in fundamental value the following January to June. For
the momentum deciles (Figures 3.2 and 3.3), the correlation for one period predictions is
.63; for two periods, .17. For the fundamental deciles (Figures 3.4 and 3.5) the
correlations rise to .73 and .29.
3.7 Do Prices Predict Fundamentals or Do Analysts Chase Prices?
The results so far, and particularly Figure 3.5, suggest that prices lead analysts'
expectations and, thus our measure of Rv. We believe the data suggests a noisy rational
expectations equilibrium. That is informed, or professional, investors either obtain
information more quickly or analyze existing data better than the average analyst. These
professional investors make their decisions and move prices before analysts publish their
opinions.
Nevertheless, another interpretation is that analysts' expectations of earnings
simply follow prices. If say, a stock's return relative to the market is positive, analysts
159
increase their published expectations. Under this interpretation, the high concurrent
correlation between returns and our measure of the change in fundamental value simply
reflects analysts publishing expectations that ratify recent price changes.
We will call this possibility the extrapolative hypothesis. That is, current changes
in analysts' expectations move in the same direction as lagged returns . It is not possible
to cleanly distinguish the extrapolative hypothesis from a noisy rational expectations
explanation because both hypotheses imply some degree of extrapolation. The
extrapolative hypothesis implies that all a security analyst does, On the other hand, if
rational expectations are noisy, analysts base forecasts on their assessment of company
fundamentals. However, since they realize that returns convey information about current
and future fundamentals, their expectations should also reflect lagged returns.
It is relatively easy to cast substantial doubt on an extreme version of the
extrapolative hypothesis, particularly over the six month intervals we have investigated.
In particular, suppose that over the past six months a stock's return has been relatively
high. If the extrapolative hypothesis is true, then analysts' should increase their
expectations of earnings, regardless of what is happening currently.
For example, suppose we look at the change in earnings expectations one month
after the end of the six month ranking period. If a company's lagged return was
relatively high, then regardless of what is happening currently, analysts' expectations
should rise and should be independent of the actual return over that one month interval.
35 There is some evidence prior stock movements partially explain changes in analyst's earnings expectations, but that fundamentals are more important. For example, Stickel (1990) shows that a stock's return over the interval between an analyst's prior estimates and her current estimate has statistically significant explanatory power. However, he also shows that its importance is small and weak (R2 increases from .37 to .38), and that the behavior of other analysts is far more dominant. Subsequent work, such as that by Keane and Runkel (1998) and others (e.g., Lim, 2001) suggest that analysts' earnings estimates are focused on predicting earnings and not driven primarily by stock prices.
160
That is, if we look at high momentum stocks, the change in analysts' earnings estimates
should be reflect past return, but not current return.
On the other hand, if the noisy rational expectations view is correct, we should
expect the current change in analysts' expectations to reflect both past and current
returns. And as the observation period after the ranking period lengthens from one month
to three and six months, the more analysts' expectations should appear to track
concurrent rather than past returns.
To investigate this, we independently sort stocks each month into deciles based
their past six month return (momentum deciles) as well as their realized returns over the
next 1, 3 and 6 months. We then calculate the average realized Rv's (change in value, as
measured by the change in analysts' expectations) for each of the 100 bins for the
following 1, 3 and 6 month periods. Each bin has between 12 and 30 firms on average.
Table 3.5 shows how Rv, our measure of fundamental value changes over each of those
intervals.
Table 3.5 does not support an extreme version of the extrapolative hypothesis
over any subsequent horizon. Panel B summarizes the results of the first panel in Table
3.5. Instead of deciles, it simply divides the sample in half, the five top deciles and the
bottom five. The six numbers in the top left corner show what happens to the stocks that
had relatively high rates of return in the last six months, and also had relatively high
returns in the subsequent 1,3 or 6 month period. The top line shows that stocks that did
well in both the initial six month period and in the subsequent one month period a return
in that subsequent month of 7.7% above the equally weighted average of all stocks in our
161
sample. Rv was also above average for those stocks, as we would expect either under
either hypothesis.
However, follow the top line across the table to the low return quadrant to the
high momentum stock that underperformed in the subsequent one month period. Already
in this first subsequent month the difference in analysts' expectations suggests a noisy
rational expectation explanation rather than the extrapolative hypothesis. The Rvt+i is
below average and significantly below the average for all high momentum stocks (t stat =
-12.6). If analysts were simply extrapolating past returns, the t statistic would have been
0 (and the change in value would be above average).
The remaining cells in the Table 3.5 also support the noisy rational expectations
explanation. For stocks where momentum works (the upper left six cells and the lower
right six cells) fundamental value tracks both past and present return. For the cells in the
off diagonals, fundamental value tracks concurrent price change more closely as the
observation interval increases. By the time six months have elapsed, which is the
window we used for our analysis, the differences in Rvt+i are striking.
Nevertheless, even at six months, Table 3.5 suggests that the stocks for which
momentum did not work still reflect some of the initial momentum. Perhaps some
analysts are still extrapolating the optimism or pessimism in past returns, or perhaps they
have not yet observed the evolution of fundamentals expected by professional investors.
3.8 The Prediction Horizon of Informed Investors
The theoretical arguments and empirical results suggest that, because of the
activity of professional investors, prices anticipate future changes in fundamental value.
162
If so, current stock price should reflect not only current fundamental value but future
fundamental value as well. Current price should also reflect the activity of noise traders
whose actions are independent of any estimates of fundamental value.
It is simplest to capture this in a multiplicative model, so we can then take
logarithms and derive a regression equation. Let Pt and Vt represent current price and
fundamental value respectively.
In a perfect market, with homogeneous expectations, where Vt accurately
measures those expectations, Pt = Vt and, over any period of time, Rt = Rvt . However,
suppose that all investors know the current Vt while a few can accurately predict Vt+i.
Then current price will be a function involving the beliefs of those who know only Vt
and those who know Vt and Vt+i (and thereby Rvt+i). The pricing function will also
reflect the number of investors in each group, their wealth levels, aversion to risk and any
possible constraints they face. In this case we might represent current price as
Pt = V t(l+Rvt+i)pl, [9]
Where and Pi reflect the relative effect of the two types of investors on price. If
all investors know Vt and no one knows Vt+i, then (3, = 0. If all investors know both Vt
and Vt+i, then p, - 1. If some know Vt+i and other do not, then p, will lie between 0 and
1. Notice that this is a different way of presenting our view of how the momentum effect
works. At any time, the current price reflects not only current fundamental value (as
perceived by the representative investor) but future value as well.
163
We can generalize equation [9] to allow some investors who may not know even
Vt, as well as some investors who can predict not only Vt+i but Vt+2 or Vt+3. Assuming
that a noisy rational expectations equilibrium is a reasonable approximation to reality, it
is interesting to ask how far into the future can investors and thus current prices anticipate
future fundamental value. Let p b (52 and P3, reflect the relative importance of future
changes in fundamental value in determining current price. Also let p0 and X be
parameters to reflect the relative importance of traders whose decisions do not depend on
any estimates of fundamental value ("noise traders"). In the following equation, it is
reasonable to expect Pi > P2 > P3-
Pt = 7LtVtP(,(l+Rv
t+i)p,(l+RV2)P2(l+Rvt+3)P3et [10]
Since 1+Rt = Pt /Pt_i, taking logarithms, differences, and using lower case letters
to denote a logarithmic variable, e.g., r = ln(l+R), yields
r, = a + p0rvt + pi Ovm -rvt) + p2(rvt+2-rvt+i) + P3(rvt+3- rVt+2) + u t , [11]
Where we reflect the influence of noise traders by a, a constant, and ut, a random
error term. Since ut = ln(et) - ln(et), we would expect it to display negative
autocorrelation if we were to estimate [11] in a time series. However, we estimate the
equation cross-sectionally.
164
The following two regressions show that Pi and (32 > 0, but p3 < 0. The positivity
of pi and P2 suggests that current prices impound (accurate) predictions of future changes
in expected fundamental value one year hence (two periods).
Months 250
244
R i t = a
Intercept 0.0704 4.89
0.0740 5.09
+ /?0Rv,t
Rv,t 0.3433 18.94
0.3456 16.90
+ /?1dRVit+1
dRvt+ i 0.1434 13.46
0.1291 9.79
,+ /?2dRv
dRv.f2 0.0383
6.27 0.0109
1.08
,t+2 + /?3dRv,t+3 + e i , t
dRvt+3 Avg # Obs 1900
-0.0183 1720 -2.45
R2
0.12339
0.1329
The negativity of P3 is consistent with the familiar reversal effect and is
inconsistent with rational expectations. However, as different regressions involve more
and more periods in the future, there is a survivorship issue - the sample size keeps
decreasing. Although we have used standard procedures to deal with this, it should be
borne in mind that a regression requiring rvt+3 is a conditional regression. That is, the
observations only involve firms that have survived more one year after the observation of
rt. If a number of firms have gone bankrupt, or were delisted, during this period,
although they were properly represented in prior regressions, they are not represented in
this one. This suggests that p3 may be biased downward and that the suggested reversal
may be more apparent than real.
3.9 Conclusion
This paper argues that stock returns are closely linked to current changes in
fundamental value and anticipate future changes in fundamental value. The behavior of
165
price and value is consistent with what one would expect in a rational well-functioning
stock market where information is costly, expectations are heterogeneous and some
investors are better informed than others. More narrowly, our data is consistent with a
heterogeneous expectations equilibrium, and more narrowly, a noisy rational expectations
equilibrium.
The noisy rational expectations model is attractive in terms of assumptions and
implications. In the first place, it requires the plausible assumption that the acquisition
and interpretation of information about equity pricing requires skill and resources. In the
second, it implies that the less well informed investors base their investment decisions, in
part, on past returns
While our work supports the notion that the evolution of prices is consistent with
the above models as well as with present value theory, it does not support the efficient
markets hypothesis. Prices do not fully reflect all available information. Instead, the data
suggests that prices partially reflect two types of information: (1) information that is
readily available (from analysts' expectations) and interpretable by a large group of
investors, and (2) the predictions of a group of "informed" investors, who rely on
information that is harder to obtain and/or more difficult to interpret.
The momentum effect in this interpretation is due to the fact that informed
investors move prices in anticipation of information that other investors will learn of in
subsequent periods.
This interpretation of price and value has implications beyond momentum, post
earnings announcement drift and the broader anomalies literature. It suggests that the
behavioral hypotheses that rely on investor reactions to past price movements may be less
166
important than hypotheses about how investors form expectations about stock
fundamentals.
It may also relate to the literature on the excess volatility of the equity market
(see, e.g., Shiller, 2003). Much of that work is based on a measure of fundamental value
equal to the present value of the future dividends of companies that survive. Stock prices
appear to move far too much relative to the slow and smooth movements of discounted
actual dividends. The results here suggest that, at the individual stock level, the volatility
of the fundamental value is far larger than suggested by discounted dividends. It may be
that, even in the aggregate, fundamental values are more volatile and more closely tied to
stock prices. If better estimates of fundamental value are highly volatile, perhaps the
stock market behaves appropriately, in that stock prices reflect estimates of fundamental
value. However, it may also be true that investors' estimates of fundamental value are
excessively volatile.
167
Table 3.1: Sample size and distribution mapped into Fama French quintiles and 12 industry classifications
The sample is composed of common equity firms that have both IBES earnings estimates for fiscal year 1 and year 2, and that can be matched to CRSP tapes. Every month firms are classified into their corresponding Fama French book to market and size quintile based on the information available on the last month of June. Panel A presents the average number of firms analyzed per month, the 5x5 size - book to market grid and the average for each quintile. Panels B to E show the sample distribution for different time periods. Panel F shows the sample distribution when firms are classified using Fama French 12 industry sectors.
Panel A: Sample Distribution from 1985 to 2006 Avg # Firms
1977 Low B/M High B/M Total Size Small
2 3 4
Big
6.1% 6.5% 5.7% 4.7% 5.4%
6.4% 5.6% 4.2% 3.6% 3.1%
6.1% 5.0% 3.7% 3.0% 2.3%
6.0% 3.9% 2.6% 2.2% 1.8%
5.9% 2.5% 1.5% 1.2% 1.1%
30.5% 23.5% 17.7% 14.6% 13.7%
Total B/M 28.5% 22.8% 20.2% 16.4% 12.0% 100.0%
Panel B: Sample Distribution from 1985 to 1989 Avg # Firms
1036 Low B/M 2 3 4 High B/M Total Size Small
2 3 4
Big
7.3% 8.0% 6.6% 4.1% 4.5%
7.4% 5.8% 3.9% 3.5% 3.3%
5.3% 4.7% 3.5% 3.3% 3.1%
3.8% 2.7% 2.5% 2.4% 2.7%
3.8% 2.3% 1.9% 1.5% 1.8%
27.7% 23.6% 18.5% 14.8% 15.5%
Total B/M 30.6% 24.0% 19.9% 14.2% 11.3% 100.0%
Panel C: Sample Distribution from 1990 to 1994 Avg # Firms
1739 Low B/M 2 3 4 High B/M Total Size Small
2 3 4
Big
6.6% 7.2% 5.7% 4.4% 4.6%
6.2% 6.5% 4.4% 3.2% 3.1%
4.7% 5.3% 3.8% 3.1% 2.6%
4.8% 3.9% 2.8% 2.4% 1.6%
6.3% 3.3% 1.4% 1.2% 0.7%
28.6% 26.2% 18.2% 14.4% 12.6%
Total B/M 28.6% 23.5% 19.6% 15.5% 12.8% 100.0%
Panel D: Sample Distribution from 1995 to 2000 Avg # Firms
2674 Low B/M 2 3 4 High B/M Total Size Small
2 3 4
Big
5.3% 6.0% 5.6% 5.1% 6.0%
5.8% 4.9% 4.2% 3.6% 3.0%
7.1% 4.9% 3.5% 2.8% 1.8%
7.5% 4.3% 2.6% 2.0% 1.2%
7.4% 2.4% 1.3% 1.1% 0.8%
33.1% 22.5% 17.1% 14.6% 12.7%
Total B/M 28.0% 21.4% 20.1% 17.6% 12.9% 100.0%
Panel E: Sample Distribution from 2001 to 2006 Avg # Firms
2318 Low B/M 2 3 4 High B/M Total Size Small
2 3 4
Big
5.6% 5.0% 5.0% 5.0% 6.3%
6.3% 5.2% 4.4% 3.9% 3.0%
7.3% 5.2% 3.9% 2.9% 1.9%
7.5% 4.7% 2.4% 2.0% 1.6%
5.8% 1.9% 1.3% 1.0% 1.1%
32.4% 22.0% 17.0% 14.7% 13.9%
Total B/M 26.9% 22.7% 21.2% 18.2% 11.0% 100.0%
Panel F: Sample Distribution Per Industry Sector Industry (1985-2006) (1985-1989) (1990-1994) (1995-2000) (2001 - 2006)
Non-Durables Durables Manufacturing Energy Chemicals BusinessEquip Telecom Utils Shops
Health Finance Other
8.45% 2.03% 10.26% 3.74% 2.51% 16.99% 1.52% 3.89% 12.49%
7.31% 18.83% 11.99%
10.68% 2.35% 11.68% 3.03% 3.48% 15.89% 1.16% 4.16% 12.18%
6.04% 18.24% 11.12%
9.49% 1.99% 10.33% 4.27% 2.75% 14.68% 1.36% 4.17% 13.43%
8.18% 18.18% 11.18%
7.29% 1.99% 10.07% 3.71% 2.01% 18.95% 1.55% 3.62% 12.63%
7.13% 18.42% 12.64%
6.70% 1.80% 9.06% 3.94% 1.95% 17.97% 1.98% 3.69% 11.74%
7.88% 20.45% 12.84%
169
Table 3.2: Returns and Change in Value, Cross-Sectional Evidence
The table presents time series average of monthly cross-sectional regression coefficients used to test the validity of the Grossman Stiglitz hypothesis. The test is conducted for return periods of six months as well as change in value over six months, therefore each subscript indicates a six month period i.e. R;,t-i is the return over the last six months for firm i and Rit.2 is the return from t - 12 months to t - 6 months. The means are computed with overlapping observations, therefore t-statistics are computed with Newey-West (1987) standard errors with a lag length of one less than the holding period horizon in months. Panel A presents mean estimates for the regression explaining current return by past returns (momentum). Panel B presents estimates for the regression explaining current return by past change in value, concurrent change in value, past returns and concurrent SIC industry return.
Panel A: Rj,t = a + KiRi,n + K2RU-2 + K3RU-3 +e\,t
Months Intercept RjiM R,^ Ri,t-3 Avg # Obs R
251
245
239
245
239
239
0.0653 4.43
0.0704
4.81 0.0744
5.04 0.0636
4.44 0.0738
5.09 0.0670
4.73
0.0688 5.19
0.0623 4.69
0.0584 3.95
0.0053
0.44
0.0039 0.34
0.0063 0.50
0.0047 0.39
-0.0368 -4.27
-0.0357 -4.34
-0.0340 -4.20
1956
1764
1615
1763
1605
1604
0.0140
0.0106
0.0085
0.0233
0.0185
0.0315
Panel B: RM = a + /3,Rv>t., + &Rv,t.2 + /33RV),.3 + YiRi,,-, + yi^i,t-i + Y3RM-3 + f M
Months Intercept RM_, R,,t.2 R ^ RV,M R ^ j Rv?,_3 Avg # Obs R2
251 0.0653 0.0688 1956 0.0139
4.43 5.19 251 0.0717 0.0147 1956 0.0046
4.78 2.52 251 0.0653 0.0659 0.0046 1956 0.0171
4.45 5.12 0.97 245 0.0633 0.0596 0.0066 0.0037 -0.0115 1765 0.0286
4.48 4.39 0.55 0.77 -2.38 239 0.0668 0.0540 0.0038 -0.0317 0.0087 -0.0058 1616 0.0361
4.73 3.65 0.31 -4.13 1.59 -1.11 239 0.0670 0.0538 0.0032 -0.0332 0.0086 -0.0051 0.0027 1616 0.0389
170
«-<J + tf" • ^
+
t-3
tf p-+ *V **
P? »r* <-+
^ H
of r H
<-+
<S • W
> & m oa. +
^ H
+•1
>" *
<£ + +*
> *
o oa. + 53 II
C? a •
u ^^ 4> a « ft-
< S
tf
i »
Xs O % M > <
* t
*
f * }
r-oS
f • i
eT
V • i
tf
2 p?
2 > *
* t
> Pi
- W
S1 y
a> g3
M
i »
A *^ a o s
• *
t~ o ~* ©
o NO I—<
^* d
^ H
f l t^ i—*
d
NO ON r~-' — i
d
© m O <N
d
m W") O CS
d
• *
o »—< ( N
d
»—c
o\ ON 1 — <
d
rn u-i • — 1
<N
d
NO i n 0 \
NO i n ON
NO m OS
NO i n OS
NO *—1
NO
NO m NO O NO O NO r - NO
O . O N . r ^ . r~ . N o . t ^ . » n r ^ t ^ N O O O N O N O N O N O N O ' O N O l ^ N O d r y ' > d r r , d c r > d ' y ' > d r r > d r o ©
r-
O ON
P ~
O en ^ --, © --, CN <3\ <N °O
^ <N->
ON O • - o-l ON "O
p -; P O ' O
<N
en
O ^; »—I cv-j
P -; o
-H O NO 00 NO
— i © o o < N a \ f x © ° O r t ~ - i © - - - . © t \ © O o o l S > ^ < S
• * - H o m »> in " i Jo d O ON O ^O o S p c> P - ; © <s O • © ' o
NO ON , -^- _ en m o - i ^ - N o f o O N O N i N rtKr-KrtO\r-lOO P < N P < s i P c s j P n S © © C O
2 S5 <N No
d ~"
^ No -* ~_ © ( \ en 2 § 2 5 g *• -<N No - ' C N N o < N ° O r N l ° o < N t ^ c s t ^ C " l t \
© ~ ~ , © ^ © ~ ~ , © ~ ~ , d " ~ , © ~ ~ '
r\ \i' \j-y i o\ ON NA
(S * -' '
NO oo oo i/i r~ en n o o v j t * n " i » , o v £ ) P \ f p > ^ 0 - ; 0 - ^ 0
© © © ©
NO NO
<N ©
d <N
—• -3-in *o r— •* ~ ON --. <N P - P <S
-* Wl
<N ^ >n CN
,—( i n <N
^ >n CN
ON CO <N
ON
m <N
ON ro (N
i n -3-(N
ON rn tN
Figure 1 Hypothetical R e t i m s for Momentim Deciles Under the AssLrrption of Market Efficiency
100%
75%
50%
25%
0%
-25%
-50% Reriod-2 Fteriod-1 FteriodO Period 1 Period2 Fteriod3
Period 0 is the Ranking Period
-LowNbm- - 6 - - 8 9 -Hg f iNbm
Table 3.1: Hypothetical Returns for Momentum Deciles
Figure 2 Returns for Momentum Deciles
Period -2 Period -1 Period 0 Period 1 Period 2 Period 3
Period 0 is the Ranking Period
- Low Mom -•— 2 -8 — 9 - High Mom
Table 3.2: Returns for Momentum Deciles
Figure 3
Change in Value (Rv) for Momentum Deciles
Period -2 Period -1 Period 0 Period 1 Period 2
Period 0 is the Ranking Period
Period 3
- Low Mom -•— 2 -8 - Ugh Mom
Table 3.3: Change in Value for Momentum Deciles
172
Table 3.3: Returns and Change in Value for Momentum Portfolios
The table presents average returns and average change in value (Rv) for ten momentum portfolios. In period 0 stocks are sorted in deciles based on their six month returns. Periods -2 and -1 show the returns and change in value that the portfolios formed in Period 0 had prior to being formed. Period 1 is the six month holding period, which to avoid bid-ask problems and any lags in analysts' earnings changes starts one month after the formation period. Periods 2 and 3 correspond to six months post holding period.
Returns for Momentum Deciles (1985 -2006)
Low Mom 2 3
4
5
6
7 8 9
High Mom High - Low
Period -2
t-18 tot-12 14.57% 11.11% 9.54%
9.29%
9.33%
9.39%
9.68% 10.10% 11.38% 11.74% -2.82%
Period -1
t -12tot-6 10.08% 8.90% 8.56%
8.95%
8.43%
8.63%
8.88% 9.57% 10.85% 14.80% 4.73%
Period 0
t-6 to t -37.32% -18.19% -9.15%
-2.59%
3.07%
8.58%
14.61% 22.19% 33.96% 73.09% 110.40%
Period 1
t+1 to t+7 3.23% 5.09% 6.26%
7.03%
7.17%
7.45%
7.86% 8.36% 8.87% 11.33% 8.10%
Period 2
t+7 to t+13 7.24% 6.76% 6.97%
7.41%
7.32%
7.35%
7.55% 7.69% 7.03% 7.15% -0.09%
Period 3
t+13 to t+19 10.14% 8.28% 8.01%
7.72%
7.47%
7.63%
7.26%
6.98% 7.02% 5.63% -4 .51%
Low Mom 2 3
4
5
6 7 8 9
High Mom High - Low
Change in Period -2
t-18 tot-12 18.56% 11.78% 10.29%
8.97%
8.40%
8.31%
9.31% 9.27% 10.92% 12.02% -6.55%
Value (Rv) for Momentum Deciles (1985 - 2006) Period -1
t-12 to t-6 14.94% 10.91% 8.30%
8.33%
7.58%
8.38%
8.76% 9.66% 12.30% 17.52% 2.58%
Period 0
t-6 t o t -12.86% -1.17%
2.59%
5.59%
7.52%
10.18%
11.67% 14.80% 22.47% 43.36% 56.22%
Period 1
t+1 to t+7 -5.24% 0.71% 3.01%
4.32%
6.22%
6.71%
7.72% 9.75% 12.90% 20.21% 25.46%
Period 2
t+7 to t+13 0.95% 2.88% 3.50%
4.92%
3.54%
4.40%
6.05% 6.45% 7.24% 9.86% 8.91%
Period 3
t+13 to t+19 8.42% 7.54% 6.06%
7.02%
3.99%
4.27%
3.81% 3.77% 3.59% 3.93% -4.49%
Excess Returns to Fama French Size & Value Porfolios (1985 -2006)
Low Mom 2 3 4
5
6
7 8 9
High Mom High - Low
Period -2
t-18 tot-12 3.29% 0.48% -0.42% -0.54%
-0 .61%
-0.55%
-0 .21% -0.22% 0.62% 0.00% -3.29%
Period -1
t-12 to t-6 1.08%
-0.54% -0.45% -0.13%
-0.50%
-0.42%
-0.35% 0.26% 1.20% 4.09% 3.02%
Period 0
t-6 t o t -42.80% -24.18% -15.87%
-9.80%
-4.70%
0.21%
5.70% 12.53% 23.26% 60.25% 103.05%
Period 1
t+1 to t+7 -2.29% -1.36% -0 .61% -0.02%
0.05%
0.30%
0.49% 0.99% 1.58% 4.05% 6.34%
Period 2
t+7 to t+13 0.39% -0.22% -0.15% -0.19%
-0.08%
-0.02%
0.16% 0.26% -0.24% 0.02% -0.37%
Period 3
t+13 tot+19 2.76% 0.50% 0.49% 0.16%
-0.02%
0.02%
-0.12% -0.15% -0.18% -0.76% -3.52%
173
Table 3.4: Returns and Change in Value for RV,M> Value Portfolios
The table presents average returns and average change in value (Rv) for ten value portfolios. In period 0 stocks are sorted in deciles based on their change in value over six months. Periods -2 and -1 show the returns and change in value that the portfolios formed in Period 0 had prior to being formed. Period 1 is the six month holding period, which to avoid bid-ask problems and any lags in analysts' earnings changes starts one month after the formation period. Periods 2 and 3 correspond to six months post holding periods.
Low R ^
2 3 4 5 6 7 8 9
High Rv,w
High - Low
Returns for Rv
Period -2 t-18 tot-12
3.22%
5.68% 7.03% 8.00% 9.26% 11.21% 13.56% 16.09% 18.43% 13.96% 10.74%
Period -1 t -12tot-6
-8.16% -1.10% 2.93% 5.82% 8.23% 10.71% 13.59% 18.21% 23.70% 26.50% 34.66%
^ Value Decile (1985 - 2006) Period 0 t - 6 t o t
-15.07%
-5 .91% -0 .01% 3.97% 6.85% 9.73% 13.05% 17.05% 23.31% 35.15%
50.22%
Period 1 t+1 to t+7
5.42% 5.94% 6.34% 6.79% 7.04% 7.72% 7.63% 7.66% 8.88% 9.21%
3.79%
Period 2 t+7 to t+13
8.79% 7.49% 7.32% 6.99% 7.20% 7.20% 6.99% 7.04% 7.09% 6.35% -2.45%
Period 3 t+13 to t+19
9.49% 8.05% 7.32% 7.35% 7.28% 7.35% 7.21% 7.28% 7.14% 7.56% -1.92%
Low RViM
2 3 4 5 6 7 8 9
High Rv,,.6 High - Low
Change in Period -2
t-18 tot-12 10.21%
6.79% 5.69% 6.00% 7.10% 8.44% 10.59% 13.08% 18.82% 24.22%
14.01%
Value (Rv) for Rv>6 Value Decile (1985 - 2006) Period -1 t -12tot -6
10.17%
6.93% 6.63% 6.46% 7.13% 8.49% 10.09% 13.16% 18.04% 21.35% 11.18%
Period 0 t - 6 t o t
-43.65% -15.30% -5.29% 0.32% 4.15% 7.48% 11.20% 16.48% 26.64%
102.32% 145.96%
Period 1 t+1 to t+7
19.19% -0.29% 0.25% 1.86% 2.83% 4.48% 5.87% 7.29% 10.02%
14.73% -4.46%
Period 2 t+7 to t+13
13.45% 4 .11% 1.36% 1.97% 2.46% 3.23% 4.23% 4.53% 5.67%
10.73% -2.72%
Period 3 t+13 to t+19
16.78%
6.75% 3.69% 2.98% 4.45% 3.51% 3.82% 3.83% 3.87%
4.08% -12.70%
Excess Returns to Fama French Size & Value Porfoli
Low R v M
2 3 4 5 6 7 8 9
High R v W
High - Low
Period -2 t-18 tot-12
-7.12% -4.33% -2.80% -1.63% -0.53% 1.21% 3.07% 5.36% 6.35% 1.81% 8.93%
Period -1 t -12tot -6 -16.74%
-9.68% -5.73% -3.14% -0.75% 1.57% 3.99% 7.99% 12.96% 14.96% 31.70%
Period 0 t - 6 t o t
-22.61% -13.47% -8 .01% ^ . 0 7 % -1.12% 1.43% 4.64% 8.22% 13.93% 25.84%
48.45%
Period 1 t+1 to t+7
-1.19%
-0.82% -0.78% -0.38% -0 .01% 0.45% 0.74% 0.90% 1.73% 2.53% 3.72%
ios (1985 -2006) Period 2
t+7 to t+13 1.12%
0.08% -0.07% -0.43% -0.11% 0.01% 0.04% -0.07% -0.24% -0.29% -1.42%
Period 3 t+13 to t+19
1.67% 0.43% -0.09% -0.27% -0.07% 0.12% -0.25% 0.16% 0.22% 0.51% -1.16%
Figure 4 Change in Value (Rv) for Rv,t-6 Value Deciles
Period -2 Period -1 Period 0 Period 1 Period 2
Period 0 is the Formation Period
• Low Rv,t-6 -6 • 8 9
Period 3
High Rv,t-6
Table 3.4: Change in Value for Past Value Deciles
Figure 5
Returns for Rv,t-6 Value Deciles
(A
E 3
•4-*
o a. •a a N O. -
Period -2 Period -1 Period 0 Period 1 Period 2
Period 0 is the Formation Period
• Low Rv,t-6 • 8
Period 3
High Rv,t-6
Table 3.5: Returns for Past Value Deciles
u
o E
a a o
s a ©
©
> Q
3 B
a S
«
s s fa
o
a «
^ ^
* -o
C3 -O 1)
-a
2 S)
& *j
o 5
•9 3
<u S * ••=
•^000000000*-;
^boododddod
oooooooooo
oooobooooo
0)lOtDT-0)t-NO)0
r*.^o>-<roa>T-com
ooooooooo
dd
dd
dd
dd
d
S2
<0
N.-*J
-C0
*-O*fC
S|C
0 g
jo
iN
'-
so
nf
fl
m
h. OdT-rrr-NNNB)
PO
OO
OO
OO
OO
^d
dd
dd
dd
dd
i-
fl
OU
Ol
tf
lO
ON
N
(0 OOOOT-T-T-CNJCMI-
OO
OO
OO
OO
OO
dddddddddd
PPoooooooo
oooodddddci
Ppooooooo
096066006
feaaMO
Orlllr-
3 g
s3
Nin
on
or
-n
rt
go
go
oo
T-
eM
co
Tf
Po
po
oo
oo
oo
9 d d d 0
00
0
ggOO*-T-T-CN5
qpooooooo
090666666
11
lO
nr
tM
O^
Or
JS
^
qqoqoqoooo
O oooooooooo
t I Nr
t^io
toK
co
o)
o
• m c
Do
oc
Mo
mo
o-^
t K
66
66
66
66
66
Tto
cN
Oc
oo
o)
^c
nc
o ^
•i-
tD
0)
T-
MT
-(
MG
)lO
O
OO
OO
OO
T-
^I
-;
dddddddddd
floifiO
T-n
oonn
OO
OO
OO
OO
T-
T-
6666666666
•»- ro
oo
co
if5T
-o>
T-o
o>
00
00
00
00
01
-d
dd
dd
dd
dd
d
to
nm
wo
co
vt
oo
oD
O
OO
OO
OO
OO
T-
;
dd
dd
dd
dd
dd
ddoddod'i
1
dd
dd
dd
dd
d
- g
oo
dd
dd
dd
dd
K2J2f2<"''D'ot^-ooo>co
N-cvioT-T-cMcgrt(0'-
OqqoOOOOO^—
ooooo'dddoo"
o o
d d
P o 6 o 6 *-
o
d
d
d
d
d
"• B
tO
OO
ll
Di
nO
NO
i)
£ T
-T
-T
-O
OO
OO
OO
o
dd
dd
dd
dd
dd
* o
f
1
I
N0
00
30
)M
O^
(D
O(
0 (
OC
Min
nu
lff
lNIO
MN
06
66
66
66
6
inc
vU
fN
Mi-
tn
co
. N
CO
On
oS
ff
lS
Nr
tQ
O
O*
-"--*
-*--"
--"--"
-
dd
dd
dd
dd
d
SC
OO
-r-CN
OC
O'd
-Tl-
T-
^C
OS
ff
lO
JO
^S
*fr O
OO
'-
" —
~
' 0
00
0 o
o
0
06
06
0)C
0O
0)(N
(0C
M0
0O
(D
to o
cv
jTr^
-mr^
OT
t-co
co
OO
OO
OO
OT
-T
-C
M
dd
dd
dd
dd
d
•"a-S
OT
t-in-tr^
cO
T-io
in2
o>
oo
TfC
M^
)-o-<
-oo
T-o
,_[O
T)-m
t^O
)cg
^-
OP
OO
OO
OO
T-
CM
dd
dd
dd
dd
d
fg^toeoaiocor^r^T-
"^T-Nomoiooj©
PP
OO
OO
OO
T- T-
oo
dd
dd
dd
d
PSS'-S'"051*'"'1''
t5SSS<N'K<°''
-ooo
|^
^PPOPOOOOT-
066060666
TTPPPPPqo?
OO
OO
OO
OQ
'Q
Q
•fcDTj-mocOT-T-aio,,
£
CMCSICMT-T-T-T-T-OO
o dodddddddo
on
nv
in
uM
oo
o
s 2
!
--I
I
176
Low Mom (R,.,) 2 3 4 5 6 7 8 9
High Mom (R,^)
Low R,+, 31 22 16 14 12 13 13 14 17 25
2 19 18 18 17 17 17 18 18 18 19
Average Number of Firms fort 3 15 18 18 19 19 19 20 19 18 16
4
13 17 19 20 20 20 20 19 18 14
5 12 16 19 20 21 21 20 19 17 13
+1 Months 6 12 17 19 20 21 21 20 19 17 13
7
13 17 19 20 20 21 20 19 17 14
8
15 17 19 19 19 19 19 19 18 17
9
19 19 18 17 17 17 17 18 19 20
High R^,
31 19 15 13 13 13 13 15 19 28
Average Number of Firms for t + 3 Months
Low Mom (R,,*) 2 3 4 5 6 7 8 9
High Mom (R,4)
Low Rt+3
31 21 16 14 13 13 13 15 18 25
2 20 19 17 17 16 17 18 18 18 19
3 16 17 18 19 19 19 19 19 18 16
4 13 17 19 20 20 20 20 19 17 14
5 13 16 19 20 21 21 20 19 17 13
6 13 17 19 20 21 20 20 19 17 13
7
13 17 19 20 20 20 20 19 17 13
8 15 18 19 19 19 19 19 19 18 15
9 18 18 18 17 17 17 17 18 19 20
High R,tJ
27 18 15 14 13 13 14 16 20 30
Low Mom (R,^) 2 3 4 5 6 7 8 9
High Mom (Rw )
Low R M
32 22 16 14 12 12 13 14 17 26
2
22 20 18 16 16 16 17 17 18 19
Average Number of Firms for t 3
17 18 18 18 19 19 18 18 18 15
4 14 17 19 20 21 20 19 19 17 13
5 12 17 19 21 21 21 21 19 16 12
+ 6 Months 6 12 17 19 21 21 20 20 19 17 13
7
13 17 20 20 20 20 19 19 17 13
8 14 17 18 19 19 19 19 19 19 15
9 17 17 17 17 17 17 18 19 20 20
H ighR,« 24 17 15 14 13 14 14 17 21 31
Panel B: Summary of the Evolution of Fundamental Value High Rt+1 Low Rt+1
V+1 Rv t+1 M+1 t+1
Average Rt+1 Average Rvt+1 Average Rt+1 Average R' t+i
of .c if
5 o
1 month
3 months
6 months
1 month
3 months
6 months
7.70%
13.97%
20.46%
8.29%
14.48%
21.52%
0.92%
3.64%
8.22%
1.36%
4.96%
10.34%
-7.74%
-14.24%
-21.71%
-8.31%
-14.34%
-20.39%
-0.93%
-3.72%
-8.71%
-1.37%
-4.88%
-9.74%
3.10 References
177
1. Ball, Ray and Philip Brown, 1968, An Empirical Evaluation of Accounting Income Numbers, Journal of Accounting Research, 6, 159-178.
2. Boudoukh, Jacob, Matthew Richardson, YuQing (Jeff) Shen, Robert F. Whitelaw, 2007,Do Asset Prices Reflect Fundamentals? Freshly Squeezed Evidence from the OJMarket Journal of Financial Economics, 83, 397-412.
3. Chan, Louis K. C, Narasimhan Jegadeesh and Josef Lakonishok, 1999, The Profitability of Momentum Strategies, Financial Analysts Journal, 55, 80-90.
4. Chordia, Tarun and Lakshmanan Shivakumar, 2006, Earnings and Price Momentum, The Journal of Financial Economics, 80, 627-656.
5. Daniel, Kent, David Hirshleifer and Avanidhar Subrahmanyam, 1998, Investor Psychology and Security Market Under- and Overreactions, The Journal of Finance, 53, 1939-1885.
6. De Long, J. Bradford, Andrei Shleifer, Lawrence H. Summers and Robert J. Waldmannn, 1990, Positive Feedback Investment Strategies and Destabilizing Rational Speculation, The Journal of Finance, 45, 379-395.
7. Edwards, E. O. and P. W. Bell, 1961, The Theory and Measurement of Business Income, University of California Press.
8. Fama, Eugene F., 1970, Efficient Capital Markets: A Review of Theory and Empirical Work, The Journal of Finance, 25, 383-417.
9. Fama, Eugene F., 1991, Efficient Capital Markets: II, The Journal of Finance, 46, 1575-1617.
10. Fama, Eugene F., 1998, Market Efficiency, Long-Term Returns, and Behavioral Finance, The Journal of Financial Economics, 49, 1998, 283-306.
11. Fama, Eugene F. and Kenneth R. French, 1997, Industry Costs of Equity, The Journal of Financial Economics, 43, 153-193.
12. Fama, Eugene F. and Kenneth R. French, 2007, Disagreement, Tastes, and Asset Prices, The Journal of Financial Economics, 83, 667-689.
13. Grinblatt, Mark and Bing Han, 2005, Prospect Theory, Mental Accounting and Momentum, Journal of Financial Economics, 78, 311-339.
14. Grossman, Sanford J., and Stiglitz, Joseph E., 1980, On the Impossibility of Informationally Efficient Markets, The American Economic Review, 70, 393-408.
15. Hong, Dong, Charles Lee and Bhaskaran Swaminathan, 2003, Earnings Momentum in International Markets, Working Paper, Cornell University.
16. Jagadeesh, Narasimhan and Sheridan Titman, 1993, Returns to Buying Winners and Selling Loser: Implications for Stock Market Efficiency, The Journal of Finance, 48, 65-91.
17. Jagadeesh, Narasimhan and Sheridan Titman, 2002, Cross-Sectional Determinants of Momentum Returns, The Review of Financial Studies, 15, 143-156.
18. Kahneman, Daniel and Amos Tversky, 1979, Prospect Theory: An Analysis of Decision under Risk, Econometrica, 47, 263-292.
19. Keane, Michael P., and David E. Runkle, 1998, Are Financial Analysts' Forecasts of Corporate Profits Rational? Journal of Political Economy, 106, 768-805
20. Kothari, S.P., 2001, Capital Markets Research in Accounting, Journal of Accounting & Economics, 31, 105-231
21. Kyle, Albert S., 1985, Continuous auctions and insider trading, Econometrica 53, 1315— 1335.
22. Lim, Terence, 2001, Rationality and Analysts' Forecast Bias, Journal of Finance, 56, 369-385.
23. Lintner, John, 1969, The Aggregation of Investors' Diverse Judgments and the Selection of Risky Investments in Stock Portfolios and Capital Budgets, Journal of Financial and Quantitative Analysis, 4, 347-400.
24. Lucas, Robert E., Jr., 1972, Expectations and the Neutrality of Money, Journal of Economic Theory, A, 103-124.
179
25. Miller, Merton H. and Franco Modigliani, 1961, Dividend Policy and the Valuation of Shares, The Journal of Business, 34,411-433.
26. Newey, Whitney and Kenneth West, 1987, A Simple Positive Semi-definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix, Econometrica, 55, 3, 703-708.
27. Roll, Richard, 1988, Orange Juice and Weather, American Economic Review, 74, 861-880.
28. Rubinstein, Mark, 1974, An Aggregation Theorem for Securities Markets, Journal of Financial Economics, 1, 225-244.
29. Shefrin, Hersh and Meir Statman, 1985, The Disposition to Sell Winners Too Early and Ride Losers Too Long," The Journal of Finance, 40, 777-790.
30. Shiller, Robert J., 2003, From Efficient Markets Theory to Behavioral Finance, Journal of Economic Perspectives, 17, 83-104.
31. Stickel, Scott E., 1990, Predicting Individual Analyst Earnings Forecasts, Journal of Accounting Research, 28, 409-417.
32. van Dijk, Ronald and Fred Huibers, 2002, European Price Momentum and Analyst Behavior, Financial Analysts Journal, 58, 96-105.