do banks price independent directors’ attention?...independent directors for whom the board is one...
TRANSCRIPT
0
Do Banks Price Independent Directors’ Attention?
Henry He Huang
Sy Syms School of Business
Yeshiva University
New York, New York 10033
Tel: (212) 960-0845, E-mail: [email protected]
Gerald J. Lobo
C. T. Bauer College of Business
University of Houston
Houston, Texas 77204-6021
Tel: (713) 743-4838, E-mail: [email protected]
Chong Wang
Von Allmen School of Accountancy
Gatton College of Business and Economics
University of Kentucky
Lexington, Kentucky 40506
Tel: (832) 651-2995, E-mail: [email protected]
Jian Zhou
School of Accountancy
Shidler College of Business
University of Hawaii at Manoa
Honolulu, HI 96822
Tel: (808) 956-7608, E-mail: [email protected]
1
Do Banks Price Independent Directors’ Attention?
Abstract
Masulis and Mobbs (2014a, 2014b) find that directors with multiple directorships allocate
their effort unequally based on a directorship’s relative prestige. We investigate whether
bank loan contract terms reflect such unequal allocation of monitoring effort by directors.
We find that bank loans of firms with a greater proportion of independent directors for
whom the board is among their most prestigious have lower spreads, longer maturities,
fewer covenants, lower syndicate concentration, and lower annual fees. Furthermore,
these firms also have higher bond ratings. Our evidence is consistent with directors’
attention reducing the cost of debt capital.
Keywords: multiple directorships, directors’ attention, cost of debt, bank loan contracting
JEL: G3 G12
2
Do Banks Price Independent Directors’ Attention?
I. Introduction
Bank loans comprise a significant source of corporate financing. According to
Bradley and Roberts (2004), new issuances of private debt, including bank loans, range
from two to three times the amount of new issuances of public debt. Consequently,
understanding the determinants of the cost of bank loans is economically important. In
this study, we examine whether the relative importance of directorships to independent
directors who serve on multiple boards is associated with firms’ bank loan contracting.
Prior literature shows that corporate governance is an important determinant of
the cost of debt. Specifically, studies demonstrate that higher board quality, greater
disclosure quality, and higher institutional ownership can reduce the cost of debt
(Sengupta (1998), Anderson, Mansi, and Reeb (2003), Bhojraj and Sengupta (2003)).
With regard to board quality, the focus has been on traditional board quality measures,
such as board independence and size (Bhojraj and Sengupta (2003), Anderson, Mansi,
and Reeb (2004), Fields, Fraser and Subrahmanyam (2012)). Recently, a few studies have
begun to explore the effect of multiple directorships on governance quality (Masulis and
Mobbs (2014a, 2014b), Huang, Lobo, Wang, and Zhou (2014)). Following these studies,
we focus on the differential economic effects resulting from directors’ unequal
prioritization of their time and effort across multiple directorships.
Multiple directorships are very common for independent directors in U.S. public
firms. More than half of the independent directors at S&P 1,500 companies serve on
more than one directorship (Masulis and Mobbs (2014a)). Although multiple
directorships can signal the talent and quality of a director (Shivdasani and Yermack
3
(1999)), strong board monitoring demands time and energy (Yermack (1996)), and the
limited availability of these resources may constrain a director with multiple directorships
from effectively fulfilling her/his directorial responsibilities (Core, Holthausen, and
Larcker (1999), Fich and Shivdasani (2006)). To minimize the potential reputation
damage that may result from directorship overload and the ensuing ineffectiveness,
directors will rationally distribute their time and energy to different directorships based
on each directorship’s relative importance (and associated prestige) (Masulis and Mobbs
(2014a)). Using firm size to proxy for the relative importance of a directorship, Masulis
and Mobbs (2014a) find that directors have a better attendance record at relatively more
important directorships and are less likely to relinquish their more important director
seats when these firms perform poorly. Further, they show that firms viewed as more
important by their directors enjoy better performance, as measured by return on assets
(ROA) and Tobin’s q. These findings are consistent with Masulis and Mobbs’ conjecture
that firm-size-based reputation incentives motivate directors to prioritize their time and
energy to important directorships.
Directors’ relative attention among multiple directorships can affect the firm’s
cost of bank loans through several channels. First, creditors, such as banks, use
accounting-based numbers to assess firm health and viability. Given that bank loan
contracts are closely tied to accounting numbers (Drucker and Puri (2009)), the integrity
of accounting numbers and the financial reporting process are important for bank loan
contracting. Because board directors can directly monitor firms’ accounting practices,
board effectiveness is critically important in constraining managers’ opportunistic
accounting behavior (Hermalin and Weisbach (1998), Klein (2002), Garcia Lara, Garcia
4
Osma, and Penalva (2009)). Effective boards and audit committees are more diligent and
spend more time and energy in fulfilling their directorial responsibilities (Menon and
Williams (1994), Vafeas (1999), Xie, Davidson, and DaDalt (2003)).1
Given that
directors will unequally distribute their time and effort among their multiple directorships
in accordance with their relative importance (Masulis and Mobbs (2014a, 2014b)), firms
viewed as more important are likely to receive more attention from and monitoring of
managerial accounting practice by directors. Consistent with this argument, Huang et al.
(2014) find that directors are more (less) likely to constrain earnings management when
they serve on more (less) prestigious boards. We expect that such unequal monitoring
among multiple directorships is also reflected in bank loan contracting.
Second, strong board monitoring improves the borrowing firm’s performance and
thereby reduces the cost of borrowing that banks charge. Specifically, the prior literature
documents a negative relation between firm performance and cost of debt (Graham, Li,
and Qiu (2008)). Because firms with a greater proportion of devoted independent
directors have better firm performance (Masulis and Mobbs (2014b)), we conjecture that
these firms obtain bank loans with more favorable terms.
Third, debt holders price the quality of a borrowing firm’s governance and will
rationally require higher loan rates from firms with more opportunistic managerial
behavior (Boubakri and Ghouma (2010)). For example, poor governance increases the
risks that rent-seeking managers would engage in irregular activities that endanger the
wealth of the debt holders, resulting in higher cost of debt. Bhojraj and Sengupta (2003),
Anderson et al. (2004), and Fields et al. (2012) document that firms with higher quality
1 For example, McMullen and Raghunandan (1996) and Garcia Lara et al. (2009) show that more frequent
meetings by boards and audit committees are associated with more conservative accounting and a lower
likelihood of earnings restatement.
5
boards have lower interest rates.2 Given that boards that are viewed as more prestigious
by their directors can more effectively constrain managerial opportunism, we conjecture
that the strong corporate governance associated with such boards will be priced favorably
by banks and, therefore, lead to lower cost of loans.
We base our empirical tests on a sample of 8,512 firm-year observations of S&P
1,500 firms from 1998 to 2011. We find that firms with a greater proportion of
independent directors for whom the board is one of their most prestigious have lower
loan spreads, longer loan maturities, and fewer loan covenants. Further analysis indicates
that the effect of directorship importance on loan spreads is more pronounced when a
majority of board members are independent. This result indicates that independent
directors’ attention is more important when the board relies on the independent directors
to carry out its mission.
To ensure that the differences in loan contract terms are attributable to directors’
attention, we also conduct a first difference analysis. We find that differences in the
proportion of directors who rank the directorship high are associated with differences in
bank loan contracting terms. To address the concern that the results may be driven by the
determinants of the relative importance of a board (e.g., firm size), we employ
propensity-score-matching. We find consistent results with this matched sample analysis.
Additionally, we obtain similar results when we match firm-years by firm size instead of
propensity score. The results of these tests indicate that the relationship between
independent director attention and cost of borrowing does not merely reflect the
differences in firm size and other determinants of directorship importance. Instead, they
2 The board quality measures in Fields et al. (2012) include size, independence, advisory presence, tenure,
busyness, gender, share ownership, and base pay.
6
capture the effect of unequal allocation of attention of board directors on the cost of
borrowing; firms that receive more attention from their independent directors have lower
cost of bank loans.
In additional tests, we extend the effect of directorship importance to independent
directors to audit committee directors and find similar results. Furthermore, we find that
boards viewed as more important by directors exhibit lower loan syndicate concentration
(that is, more diverse lender pools) and pay lower annual loan fees. Finally, we show that
firms with more director attention have higher bond ratings. Overall, we provide robust
evidence that a strong board, as measured by the attention paid by its independent
directors, can reduce the cost of borrowing.
We make several important contributions to the literature. First, despite the
prevalence of directors serving on multiple boards, few studies have explored the impact
of the unequal allocation of effort by independent directors on firm behavior, Masulis
and Mobbs (2014a), (2014b) and Huang et al. (2014) being notable exceptions. We thus
contribute to the emerging literature on the unequal allocation of effort by directors with
multiple directorships, by linking the unequal monitoring to bank loan contracting. Our
results suggest that when retaining a director, a firm needs to take into consideration the
relative importance (and time and effort) it will receive from the director because, among
other effects, this has implications for the pricing of its loans.
Second, we add to the literature linking board governance and cost of debt.
Bhojraj and Sengupta (2003), Anderson et al. (2004), and Fields et al. (2012) find that the
cost of debt is inversely related to traditional measures of board quality such as board
independence, size, and meeting frequency. We contribute to this line of research by
7
focusing on the unequal attention of directors with multiple directorships, which to date is
unexplored in the literature on board governance and cost of debt.
The remainder of the paper is organized as follows. We discuss prior research and
develop our hypotheses in Section 2. Section 3 describes the sample selection process and
presents descriptive statistics. Section 4 discusses the results of our main analyses, and
Section 5 presents the results of additional analyses. Section 6 concludes the study.
II. Prior Literature and Hypotheses Development
A. Multiple Directorship and Board Effectiveness
Research on multiple directorships has primarily focused on how busy directors
are and generally finds that having multiple directorships can make directors too busy to
effectively monitor all the firms under their supervision. For example, Core et al. (1999)
suggest that busy directors are associated with excessive CEO compensation, which leads
to lower firm performance. Shivdasani and Yermack (1999) find that CEOs are likely to
appoint busy directors who serve on multiple boards, and interpret this phenomenon as
CEOs’ attempts to reduce monitoring pressure. Fich and Shivdasani (2006) propose that
firms with busy boards (e.g., firms with a majority of members holding three or more
directorships) have weaker governance and are associated with lower operating
profitability (i.e., ROA) and market valuation (i.e., market-to-book ratio). They also show
that the market responds positively to news of departures of busy directors. However,
Ferris, Jagannathan, and Pritchard (2003) find no evidence that busy directors shirk their
responsibilities to serve on subcommittees; nor do they find an association between
multiple directorships and the likelihood of securities litigation.
Given busy directors’ time constraints, Masulis and Mobbs (2014a) expect that
8
directors differentially value each directorship based on its relative importance to their
reputation, and allocate their time and energy among these boards accordingly. Since firm
size is associated with greater visibility, prestige, compensation, and opportunity to attract
additional external director appointments (Ferris et al. (2003), Ryan and Wiggins (2004),
Adams and Ferreira (2008)), Masulis and Mobbs (2014a) use relative size of each firm
where they serve as independent directors to proxy for the relative importance the
director places on this directorship and the relative amount of time and effort allocated to
this directorship. At the director level, they find that independent directors are more likely
to attend the board meetings of their more important directorships and less likely to
relinquish these board seats, even when these firms perform poorly. At the firm level,
they find that the fraction of independent directors who view the board as relatively more
important is positively associated with the sensitivity of CEO departure to poor
performance as well as to overall firm performance (i.e., ROA) and valuation (i.e.,
Tobin’s q). Further, in a related study, Masulis and Mobbs (2014b) find that greater
reputation incentives of a board lead to less negative outcomes, such as stock delistings,
violations of debt covenants, financial report restatements, options backdating, securities
class actions, and reductions of cash dividends, and more positive firm outcomes.
B. Multiple Directorships, Accounting Quality, and the Cost of Debt
Lenders explicitly rely on financial statement numbers in setting debt covenants
and performance pricing provisions (Dichev and Skinner (2002), Li (2010)). Thus, bank
loan contracting must reflect the quality of a firm’s accounting. Anderson et al. (2004)
propose that creditor reliance on accounting-based debt covenants suggests that debtors
are potentially concerned with board of director characteristics that influence the integrity
9
of financial accounting reports. In a sample of S&P 500 firms, they find that the cost of
debt is negatively related to board independence and board size. They also find that fully
independent audit committees are associated with a significantly lower cost of debt.
Similarly, yield spreads are also negatively related to the size and meeting frequency of
the audit committee. Overall, these results provide market-based evidence that boards and
audit committees are important elements affecting the reliability of financial reports. The
importance of accounting numbers in bank loan contracting and the incentives of
management to manipulate these numbers make banks rationally sensitive to major
changes in governance dimensions that may affect the integrity and reliability of the
financial reporting process (Smith (1993)).
Board monitoring, and especially monitoring by audit committees, is critically
important in constraining managers’ opportunistic accounting behavior (Beasley (1996),
Klein (2002), Xie et al. (2003), Beekes, Pope, and Young (2004), DeFond, Hann and Hu
(2005), Ahmed and Duellman (2007), Larcker, Richardson, and Tuna (2007), Garcia Lara
et al. (2009),). Effective boards and audit committees demand more attention, effort,
time, and energy (Vafeas (1999), Xie et al. (2003)). If directors distribute their attention
unevenly among multiple directorships according to their relative importance, then the
monitoring effectiveness of a firm’s accounting decisions will vary with the amount of
attention given to a directorship. Consistent with this reasoning, Huang et al. (2014) find
that directors are more (less) likely to constrain earnings management when they serve on
more (less) prestigious boards.
Given that the cost of borrowing reflects the accounting quality and that the
distribution of directors’ efforts among multiple directorships affects accounting quality,
10
we therefore expect firms that are viewed as more prestigious by their independent
directors to have lower cost of borrowing.
C. Multiple Directorships, Firm Performance, and the Cost of Debt
Strong board monitoring is associated with higher profitability and lower return
volatility, which reduce the cost of borrowing charged by banks. Specifically, firms with
higher profitability have lower default risk and can obtain loans at lower rates and better
terms (Graham et al. (2008)). Prior literature also documents a positive relation between a
firm’s earnings volatility and cost of debt, consistent with banks benefiting from a stable
stream of cash flow that can support bond repayments (e.g., Graham et al. (2008)). Firms
with a greater proportion of devoted independent directors have better firm performance
(e.g., higher ROA). These firms also have a lower likelihood of violating loan covenants
(Masulis and Mobbs (2014b)). We therefore conjecture that these firms will enjoy more
favorable bank loans terms.
D. Multiple Directorships, Firm Governance, and the Cost of Debt
Boubakri and Ghouma (2010) argue that debt holders price the quality of a
borrowing firm’s governance and will rationally require higher interest rates from firms
with more opportunistic managerial behavior. For example, to extract private benefits,
opportunistic managers may divert cash away from paying debt holders, and this
likelihood of managerial opportunism will negatively affect the loan terms. Strong board
monitoring of managers also alleviates the risks that rent-seeking managers would engage
in fraudulent activities that endanger the wealth of debt holders, or that they would divert
assets away from supporting debt cash flow. For example, poor governance can lead to
financial fraud (e.g., Enron, WorldCom, Xerox) that can inflict catastrophic damages on
11
debt holders and also damage a firm’s disclosure credibility about forecasted future cash
flows, resulting in a higher cost of debt. Consistent with this reasoning, Graham et al.
(2008) document that financial irregularities are followed by higher interest rates, shorter
maturities, and more covenant restrictions by banks.
Specific evidence indicates that board quality impacts the cost of bank debt.
Fields et al. (2012) analyze the relation between comprehensive measures of board
quality and the cost as well as the non-price terms of bank loans. They show that firms
that have higher quality boards with a greater advisory presence borrow at lower interest
rates. This relation holds even after controlling for other firm characteristics, such as
ownership structure, CEO compensation policy, and shareholder protection, as well as the
size and financial condition of the borrower. They also document that board quality and
other governance characteristics influence the likelihood that loans have covenant
requirements, but the relations differ by covenant type. When they combine the direct and
indirect costs of bank loans, they find that firms with large, independent, experienced,
and diverse boards, and with lower institutional ownership, borrow more cheaply. Other
studies in the literature also show empirically that measures of corporate governance
quality, such as disclosure quality, institutional ownership, and board independence, are
negatively related to the cost of debt (Sengupta (1998), Anderson et al. (2003), Bhojraj
and Sengupta (2003)).
Given the costly renegotiation following default (references) and the costly
enforcement of loan contracts (Bolton and Scharfstein (1996)), lenders are willing to
extend credit on more favorable terms if they know ex ante that the borrower has strong
governance. The resulting lower governance risk that debt holders perceive will be priced
12
favorably in the firm’s loan terms (Graham et al. (2008), Boubakri and Ghouma (2010),
Qi, Roth, and Wald (2010)). Firms with more attentive directors have better governance
quality (Masulis and Mobbs (2014a)). When directors view the board as more
prestigious, they can better constrain managerial opportunism. As a consequence, we
conjecture that the strong corporate governance associated with these boards will lead
banks to offer more attractive loan features, including lower cost loans.
Based on the aforementioned arguments, we propose the following hypothesis:
H1: Firms associated with more attentive directors receive more favorable bank
loan contract terms.
III. Main Variables, Data Sources, Sample, and Descriptive Statistics
A. Measurement of Board Importance, Sample Selection, and Data Sources
Following Masulis and Mobbs (2014a, 2014b), we rank the relative importance of
directorships for a director sitting on multiple boards using the size of each firm’s equity
market capitalization. Among all the directorships a director serves, the directorship with
the largest (smallest) firm size is deemed the most (least) important and prestigious.
Percent_Ranked_High (Percent_Ranked_Low) captures the percentage of independent
directors for whom this directorship’s rank is 10% larger (smaller) than their smallest
(largest) directorship measured by a firm’s market capitalization. We focus on
independent directors because they are more effective monitors (Beasley (1996), Klein
(2002), Xie et al. (2003), Ahmed and Duellman (2007), Larcker et al. (2007), Garcia Lara
et al. (2009)).
Panel A of Table 1 presents the sample selection procedure. We start by collecting
independent director board data from RiskMetrics and merge this data with bank loan
information from DealScan for the years 1998 to 2011. To ensure that directorship
13
importance drives loan contract terms, we match the loan contract terms with the
preceding year’s director board data. Our initial sample consists of 10,996 firm-year
observations with director data available in RiskMetrics, which covers board information
for the S&P 1,500 firms. We then exclude all 1,348 observations in the financial services
and utility industries. Next, we drop 325 observations with bridge loans and non-fund-
based facilities, such as leases and standby letters of credit. Last, we exclude 811
observations that lack the necessary data to construct the control variables used in our
empirical tests. Our final sample comprises 8,512 firm-year observations. We extract
financial data from Compustat, and stock return data from CRSP.
Pane B of Table 1 presents the sample distribution by year. The sample size varies
widely over time, ranging from a high of 1,187 firms in 2001 to a low of 152 firms in
2011.3
B. Descriptive Statistics
Panel C of Table 1 shows the descriptive statistics for our sample. The first part
shows the descriptive statistics for the bank loan characteristics. The mean and median of
“all-in-drawn” spread are 175 and 150 basis points, respectively. The average maturity is
45.63 months. The mean and median facility amounts are 506.6 and 250 million,
respectively. The percentage of secured bank loans is 39.2%. The averages of total,
general and financial covenants are 3.827, 2.344 and 1.483, respectively. Of the bank
loans in the sample, 48.7% have performance pricing and the average number of lenders
per loan is 10. 3 The main reason for the large drop in our sample after 2007 is that new loans fell by almost 50%
during the financial crisis (2007-2009). This is consistent with Ivashina and Scharfstein (2010),
who also document a large drop-off. Specifically, the full sample in DealScan decreases
significantly, from 6,927 in 2006 to 4,943 in 2007 and 3,843 in 2008. There are only 152 firms in
2011 because most data for 2011 were not available at the time of our data cutoff.
14
The second part of Panel C presents the descriptive statistics for board
characteristics. Across all boards, 19.2 percent of independent directors classify the
directorship as of high importance (Percent_Ranked_High) and 24.2 percent of
independent directors classify the directorship as of low importance
(Percent_Ranked_Low). While 9.3 percent of firms are ranked by the majority of their
independent board members as of high importance (Majority_Ranked_High), 12.3
percent of firms are ranked as of low importance (Majority_Ranked_Low). On average,
CEO ownership is 2.5% of outstanding shares and independent director ownership is
1.1%.
The last part of Panel C presents the descriptive statistics for firm characteristics.
We do not discuss these statistics in detail for the sake of brevity.
IV. Methodology and Empirical Results
A. Methodology
We estimate the relationships between directorship importance and bank loan
contract terms using the following specification:
Bank Loan Contracting =
β0 + β1Percent_Ranked_High + β2Percent_Ranked_Low + β3Majority independent
+ β4CEO ownership + β5CEO ownership squared + β6Independent ownership
+ β7Ln Facility amount + β8Ln Maturity + β9Number of covenants
+ β10Number of lenders + β11Performance pricing + β12Total assets + β13Leverage
+ β14Return on assets + β15Operating cash flow volatility + β16Tangibility
+ β17Z-score + β18MB + β19Credit spread + β20Term spread + Loan Purpose
+ Loan Type + Industry + Year + ε (1)
where, Bank Loan Contracting separately refers to one of the following bank loan
characteristics: cost of borrowing, loan maturity, and number of covenants.4
4 Model (1) does not include Ln (Maturity) as a control variable when the dependent variable, Bank Loan
Contracting, represents loan maturity, and does not include Number of covenants as a control variable when
15
Following Masulis and Mobbs (2014b), we control for several governance
characteristics in our regression analysis. Specifically, we include an independent board,
termed Majority independent, CEO ownership, CEO ownership squared, and
Independent ownership in Equation (1). Drawing on prior studies (Qian and Strahan
(2007), Boubakri and Ghouma (2010), and Qi et al. (2010)), we also control for loan
characteristics and firm characteristics. We use the natural log of the amount of the
facility committed by the facility’s lender pool (Ln (Facility amount)), the natural log of
the number of months to maturity (Ln (Maturity)), the total number of covenants (Number
of covenants), the number of lenders (Number of lenders), and whether the facility has a
performance pricing provision (Performance pricing), to capture other loan
characteristics besides spread.5 We control for firm characteristics, including the natural
log of book value of assets (Ln (Total assets)), total liabilities divided by total assets
(Leverage), income before extraordinary items divided by total assets (Return on assets),
standard deviation of quarterly cash flows from operations divided by total assets
(Operating cash flow volatility), property, plant and equipment scaled by total assets
(Tangibility), Altman’s Z-score (Z-score), and market-to-book ratio (MB), all measured at
the beginning of the fiscal year. We also use two variables to control for macroeconomic
factors, the difference in the yield between BAA- and AAA-rated corporate bonds
measured one month before the loan becomes active (Credit spread), and the difference
in the yield between ten-year and two-year U.S. Treasury bonds measured one month
before the loan becomes active (Term spread). We include Loan Purpose and Loan Type
to control for differences in loan purpose and type. Specifically, Loan Purpose captures
the dependent variable represents the number of covenants. 5 We do not include annual loan fees as a control variable because only 33.6% of our sample facilities have
annual loan fee information.
16
the primary purpose for the loan, including acquisition line, commercial paper backup,
corporate purposes, debt repayment, takeover, working capital, and other purposes. Loan
type includes 364-day facilities, delay draw term loans, revolvers and term loans. Finally,
we include industry and year fixed effects. Appendix A provides detailed descriptions of
all the variables used in the model.
B. Relation between Cost of Borrowing and Importance of Directorship
Table 2, Panel A presents the estimation results relating relative importance of
multiple directorships to the cost of debt, measured using the log of all-in-drawn spread
as the dependent variable. Columns (1)-(3) show the results using Percent_Ranked_High
and Percent_Ranked_Low. Column (1) controls for only board characteristics, Column
(2) controls for all the control variables except loan characteristics, and Column (3)
controls for all the control variables. We find similar results in all three columns.
Specifically, in Column (3), we find that Percent_Ranked_High is negatively related to
the cost of debt, as indicated by the coefficient β1, which equals -0.161 with a t-value of -
5.04. Moving from the first quartile (0.000) to the third quartile (0.333) of
Percent_Ranked_High reduces the cost of debt, measured by the all-in-drawn spread, by
1.131 basis points.6 We also find that Percent_Ranked_Low is significantly positively
related to the cost of debt, as indicated by β2, which has a coefficient estimate of 0.065
with a t-value of 2.78. Moving from the first quartile (0.000) to the third quartile (0.375)
for Percent_Ranked_Low increases the cost of debt by 1.057 basis points for all-in-drawn
spread.7 The difference between β1 and β2 is highly significant (F = 30.21, p = 0.00).
These results show that firms ranked high by directors with multiple directorships are
6 It is calculated as follows: exp((0.333 – 0.000)*0.161) = 1.131.
7 It is calculated as follows: exp((0.375-0.000)*0.065)=1.057.
17
associated with lower cost of debt, while firms ranked low by directors with multiple
directorships are associated with higher cost of debt.8
Columns (4), (5) and (6) present regressions analogous to those in Columns (1) -
(3), except that we use the variables Majority_Ranked_High and Majority_Ranked_Low
in place of Percent_Ranked_High and Percent_Ranked_Low. We find consistent evidence
with this alternative specification; firms with boards valued highly by multiple
directorship directors have lower cost of debt.
Table 2, Panel B examines the interaction effect of directorship importance and
overall board independence (Majority Independence) on cost of debt. Column (1) shows
the results for Percent_Ranked_High and Percent_Ranked_Low, and Column (2) shows
the corresponding results for Majority_Ranked_High and Majority_Ranked_Low. All
columns include all the control variables. The results show significantly negative
coefficients on the interaction term Percent_Ranked_High*Majority_Independent in
Column (1) (β4 = -0.244, t = -4.12, p = 0.00), and
Majority_Ranked_High*Majority_Independent in Column (2) (β4 = -0.160, t = -2.94, p =
0.00). These estimates indicate that the effect of strong director attention on reducing the
cost of debt is more pronounced when the majority of the board is independent.
We also find significantly positive coefficients on the interaction term
Percent_Ranked_Low*Majority_Independent in Column (1) (β5 = -0.146, t = -3.02, p =
0.00), indicating that the effect of weak director attention on increasing the cost of debt is
also more pronounced when the majority of the board is independent. These results
suggest that when the majority of a board is composed of independent directors, the
8 Ln (Maturity), as a control variable, is not significant in this and most of the subsequent tables. This is
because we control for the loan type fixed effect, which is highly correlated with loan maturity. Once we
drop the loan type fixed effect, the coefficient on Ln (Maturity) becomes significant in most regressions.
18
effectiveness of the board depends largely on these independent directors’ efforts and
attention; as a result, these directors’ attention to the board becomes more important in
affecting the cost of debt.9
C. Relation between Loan Maturity and Importance of Directorship
Table 3, Panel A presents the results of the relation between the relative
importance of multiple directorships and loan maturity. Columns (1) - (3) show the
regression results using Percent_Ranked_High and Percent_Ranked_Low. All three
columns yield similar results. Specifically, in column (3), we find that
Percent_Ranked_High is positively related to loan maturity (β1 = 0.052, t = 2.45, p =
0.01). Moving from the first quartile (0.000) to the third quartile (0.333) for
Percent_Ranked_High increases loan maturity by 1.040 months, which represents a
2.279% increase in loan maturity.10
Percent_Ranked_Low is insignificantly positively
related to loan maturity (β2 = -0.003, t = -0.17, p =0.87). The difference between β1 and β2
is significant at the 0.1 level (F = 1.88, p = 0.06). These results show that firms ranked
high by directors with multiple directorships are associated with longer loan maturity
than firms ranked low by directors with multiple directorships.
Columns (4) - (6) present the results of regressions using the variables
Majority_Ranked_High and Majority_Ranked_Low. We find consistent evidence with
this alternative specification; firms valued highly by directors with multiple directorships
9 The results indicate that the coefficients on Percent_Ranked_High, Percent_Ranked_Low,
Majority_Ranked_High, and Percent_Ranked_Low, in Columns (1) and (2) of Table2, Panel B are no
longer significant. One explanation for these results is that these coefficients reflect the effect of
directorship importance for firms that do not have a majority of independent directors on their boards (i.e.,
for 16.5% of sample observations). In other words, directorship importance is relevant for bank loan
contracting only when the board comprises a majority of independent directors (i.e., for 83.5% of sample
observations). 10
The loan maturity increase is calculated as follows: exp((0.333 – 0.000)*0.052) = 1.040, and the
percentage is calculated as follows: 1.040/45.630=2.279%.
19
enjoy longer loan maturity.
Table 3, Panel B presents the results of regressions relating multiple directorships
to loan maturity measured using loan maturity instead of log loan maturity. We find
results qualitatively similar to those reported in Table 3, Panel A.
D. Relation between Number of Covenants and Importance of Directorship
Table 4, Panel A presents the results of regressions relating multiple directorships
to number of total covenants. Columns (1) - (3) show the regression results with
Percent_Ranked_High and Percent_Ranked_Low. The results in all three columns are
similar. For example, in Column (3), Percent_Ranked_High is significantly negatively
related to the number of total covenants (β1 = -0.079; t = -1.85, p = 0.06)11
and
Percent_Ranked_Low is insignificantly negatively related to the number of total
covenants (β2 = -0.017, t = -0.52, p =0.60). These results show that firms ranked high by
directors with multiple directorships are associated with a smaller number of total
covenants. Columns (4) - (6) present results with Percent_Ranked_High and
Percent_Ranked_Low. These results are consistent with the evidence in Columns (1) -
(3); we continue to find that firms valued highly (lowly) by multiple directorship
directors have fewer (more) total covenants.
Table 4, Panel B presents the results of regressions relating multiple directorships
to number of general covenants. We find that firms valued highly (lowly) by multiple
directorship directors have fewer (more) general covenants in Columns (1) and (4), but
not for the models in the other columns.
Table 4, Panel C presents the results of regressions relating multiple directorships
11
The marginal effect of Percent_Ranked_High is -0.25. That is, for a one unit increase in
Percent_Ranked_High, the number of covenants will decrease by 0.25.
20
to number of financial covenants. Except for the model in Column (2), we find that firms
valued highly by multiple directorship directors have fewer financial covenants.
V. Additional Analysis
A. First Differences Analysis
The results reported earlier are based on a levels analysis, which is inherently
vulnerable to omitted variables that may bias the coefficient estimates. To alleviate this
concern, we conduct a first differences analysis by relating the two-year change (from
year t-1 to year t+1) in the loan contract terms to the corresponding two-year change
(from year t-2 to year t) in the percentage of directors that view the directorship as more
or less important (i.e., two-year change in Percentage_Ranked_High and
Percentage_Ranked_Low).12
We also limit the sample to observations with non-zero
changes in directorship importance from year t-2 to year t. Table 5 presents the results of
this first differences analysis. We continue to find results consistent with the results
reported earlier. Specifically, we find that increases (decreases) in
Percentage_Ranked_High are associated with decreases (increases) in loan spread and
increases (decreases) in loan maturity. Furthermore, increases (decreases) in
Percentage_Ranked_Low are associated with increases (decreases) in the number of total
covenants and general covenants. These results provide further evidence that the
differences in the cost of bank loans can be attributed to the changes in importance of
directorships.
B. Using a Propensity-score-matched Sample
Given that we use firm size to proxy for the relative prestige of a firm, and given
12
The results are similar if we use annual changes.
21
that prior research indicates that larger firms are likely to have better credit history and,
therefore, lower cost of debt, our results relating loan characteristics to the relative
importance of multiple directorships could be driven by differences in firm size. Although
we include total assets (Ln(Total Assets)) in our regressions to control for this size
effect,13
we conduct two additional tests to alleviate this potential concern.
First, we select a matched sample using the propensity score matching technique
and compare the results for this sample to those for our test sample. Table 6, Panel A
presents the results for the first-stage regression. The dependent variable, High_Low, is
an indicator variable that equals one if Majority_Ranked_High equals one, and zero if
Majority_Ranked_Low equals one. Firms with larger market value and higher operating
earnings volatility are more likely to have Majority_Ranked_High boards. Firms with
lower percentage of female directors and with majority independent directors are less
likely to have Majority_Ranked_High boards. We use the estimated coefficients from this
first-stage regression to compute the propensity score for each observation in our sample.
We then match each firm-year that is ranked as relatively important by the majority of its
independent directors (i.e., a firm-year with Majority_Ranked_High = 1) with a firm-year
that has the opposite ranking (i.e., Majoroty_Ranked_Low = 1), that is in the same-
industry (based on the Fama-French 48 industry classification), and has the closest
13 Ln(Total Assets), as a measure of size, usually has the same association with the loan contract terms as
Percent_Ranked_High and Majority_Ranked_High. However, in Table 3, loan maturity is negatively
associated with Ln(Total Assets) and positively associated with Percent_Ranked_High and
Majority_Ranked_High. This is not inconsistent with the previous literature, which presents mixed
evidence between loan maturity and firm size. For example, when using loan maturity as the dependent
variable, Graham et al. (2008) and Chan, Chen, and Chen (2013) find that the coefficients on size are
negative but insignificant. Furthermore, Costello and Wittenberg-Moerman (2011) report a positive and
significant coefficient on size in Table 5, but a negative and significant coefficient on size in Table 8.
22
propensity score.14
We use these matched firms as the control sample.
Table 6, Panel B presents the regression results using the propensity-score-
matched sample. Column (1) shows that High_Low is associated with significantly lower
loan spreads. Column (2) shows that High_Low is not associated with loan maturity. The
regression results in Columns (3), (4) and (5) show that High_Low is associated with
significantly lower number of total covenants, general covenants and financial covenants.
These results are qualitatively similar to our primary results, except for loan maturity.
C. Relation between Loan Contract Terms and Importance of Audit Committee
Directorship
Because monitoring by audit committees is critically important in ensuring the
reliability of a firm’s accounting information, we also examine the effect of audit
committee directorship importance on loan contract terms. We define Audit
Percent_Ranked_High (Audit Percent_Ranked_Low) as the percentage of independent
audit committee members to whom this ranked directorship is 10% larger (smaller) than
their smallest (largest) directorship measured by the firm’s market capitalization. We also
define Audit Majority_Ranked_High (Audit Majority_Ranked_Low) as an indicator
variable that equals one if the majority of independent outside directors in the audit
committee classify this directorship as high (low) ranked (i.e., 10% larger (smaller) than
their smallest (largest) directorship by market capitalization of the firm), and zero
otherwise.
Table 7 presents the regression estimates. The results are very similar to our main
findings for the full board. Specifically, the results indicate that greater audit committee
directorship importance is associated with lower loan spread, longer maturity, and fewer
14
The results are similar if we match by firm size (market value of equity) instead of propensity score.
23
covenants. This suggests that the audit committee plays an important role in reducing the
cost of debt, probably because of its importance in monitoring a firm’s accounting
decisions.
D. Relation between Syndicate Concentration, Annual Loan Fees, and Importance of
Directorship
We also examine the effect of relative board importance on two additional
measures of bank loan contracting: syndicate concentration and annual loan fees.
Following Sufi (2007), high information asymmetry forces the lead arranger to take a
larger stake in the loan and form a more concentrated syndicate. We measure syndicate
concentration using the Herfindahl index calculated using each syndicate member’s share
in the loan. A higher index score indicates more concentration by lenders. Table 8
presents the estimation results using Equation (1) with syndicate concentration as the
dependent variable. We find that the coefficients on Percent_Ranked_High and
Majority_Independent_High are significantly negative in Columns (1) and (3), where
only board characteristics are included as control variables. In addition, we find that the
coefficient on Majority_Independent_High is significantly negative in Column (4), where
all the control variables are included in the regression. Overall, the results indicate that
boards with more motivated directors have lower levels of syndicate concentration. In
other words, more lenders are willing to lend to a firm if it has a greater fraction of
directors that view its board as more important.15
Table 9 shows the results using Equation (1) with annual loan fee as the 15
We also examine the effect of directorship importance on the percentage stake of the lead arranger. We
find some evidence that firms with more devoted independent directors have lead arrangers with a lower
percentage stake, consistent with directorship importance reducing loan syndicate concentration.
Furthermore, we examine the effect of directorship importance on the collateral requirement, and find some
evidence consistent with higher (lower) directorship importance decreasing (increasing) the likelihood of
collateral requirement.
24
dependent variable. Annual fee is the natural log of the number of basis points of a
facility commitment amount that a borrower is required to pay on an annual basis
regardless of any loan outstanding. In Columns (1) and (2), Percent_Ranked_High
(Percent_Ranked_Low) is significantly negatively (positively) related to the annual loan
fee. Similarly, in Column (3), Majority_Ranked_High (Majority_Ranked_Low) is
significantly negatively (positively) related to the annual loan fee. These results indicate
higher annual loan fees for firms that are ranked as less important by their directors.
In sum, the results in Tables 8 and 9 suggest that boards that are viewed as more
prestigious by their members have more diffuse lender groups and lower annual loan
fees.
E. Relation between Bond Ratings and Importance of Directorship
We also examine the effect of directorship importance on bond ratings. Since
bond rating is an important determinant of bond cost, this test also indirectly reflects the
relation between director attention and bond cost. We obtain bond data from Mergent
Fixed Income Securities Database (FISD), and convert the bond ratings to numerical
values using the conversion in Becker and Milbourn (2011).16
We compute Ln (Bond
Maturity) as the natural logarithm of the number of days to maturity, and include it as a
control variable. Following Mansi et al. (2011), we include two major determinants of
bond ratings, Idiosyncratic Risk and Firm Risk, as additional control variables.
Furthermore, we include Ln (Total Assets) to control for firm size and information
environment, Leverage to control for capital structure, Return on assets to control for
firm performance, Z-score to control for default risk, and MB to control for growth. We
16
Following Becker and Milbourn (2011), we code AAA as 28, AA+ as 26, AA as 25, AA- as 24, A+ as 23,
A as 22, A- as 21, BBB+ as 20, BBB as 19, BBB- as 18, BB+ as 17, BB as 16, BB- as 15, B+ as 14, B as
13, B- as 12, CCC as 11, CC as 7, and C as 4.
25
also include all the board characteristics in Equation (1).
Table 10 presents the results of estimating the effect of directorship importance on
bond ratings. We find that Percent_Ranked_High is significantly and positively
associated with bond ratings in Columns (1) and (2), and Majority_Ranked_High has a
similar association with bond ratings in Columns (3) and (4). Further,
Percent_Ranked_Low (Majority_Ranked_Low) has a significantly negative association
with bond ratings in Column (1) (Column (3)). These results indicate that firms whose
directorships are viewed as more (less) important by their directors are associated with
higher (lower) bond ratings.
E. Robustness Tests
Loan terms such as loan spread, maturity, and number of covenants are often
determined simultaneously. We conduct robustness tests to address this potential
endogeneity concern but, for brevity, do not tabulate the results. First, we follow Graham
et al. (2008) and employ two stage least squares estimation. In the first stage, we regress
bank loan maturity on the borrower’s total assets maturity.17
In the second stage, we re-
estimate our main model using the fitted value of loan maturity from the first stage as the
instrument for loan maturity. We find results that are consistent with our main results.
Second, to address the endogeneity concern for cost of debt and number of
covenants, we follow Costello and Wittenberg-Moerman (2011) and estimate a
simultaneous equations model, with Prior lending rate as the instrument for cost of debt
17
In the first stage, following Johnson (2003), we measure the total assets maturity as two different
components. The first component, current assets maturity equals current assets (ACT) scaled by cost of
goods sold (COGS). The second component, long-term assets maturity equals gross PPE (PPEGT) scaled
by depreciation expense (OIBDP-OIADP). Total assets maturity equals the weighted (ACT/AT and
PPEGT/AT) sum of these two components.
26
and Lead Reputation as the instrument for number of covenants.18
Once again, we find
consistent results.
In additional robustness tests, we use other measures of importance (reputation
incentives) of a directorship. First, similar to Masulis and Mobbs (2014), we rank
directorships using the market value of total assets (i.e., sum of book value of liabilities
and market value of equity). Second, we change the threshold of the relative importance
of directorships from 10% of a firm’s market capitalization to 20% and 50%. We find
results under these three alternative measures to be very similar to the primary results
using the 10% threshold. Third, we control for potential nonlinearity in the relation with
firm size (Ln(Total Assets)) by including the square of this variable as an additional
control variable and find consistent results.
We control for the direction of performance pricing provisions in our fourth
robustness test. Our sample is considerably reduced because only 48.7% of the
observations include a performance pricing provision. Following Asquith, Au, Covert,
18
This simultaneous equations model is specified as follows. Prior Lending Rate equals the cost (Ln(All-in-
drawn)) of the previous deal for the same borrower. Lead Reputation is an indicator variable that equals
one if the loan is syndicated by a reputable lead arranger in the syndicated loan market. Following
Bushman and Wittenberg-Moerman (2012), we define a lead arranger bank as reputable if its market share
is larger than 2% in our sample period. The market share is the ratio of total amount of loans syndicated as
a lead arranger to the total amount of syndicated loans in our sample period. All other variables are defined
as in Equation (1).
Ln(All-in-drawn)= β0 + β1Percent_Ranked_High (Majority_Ranked_High) + β2Percent_Ranked_Low
(Majority_Ranked_Low )+ β3Majority independent + β4CEO ownership + β5CEO
ownership squared + β6Independent ownership + β7Ln Facility amount + β8Ln Maturity
+ β9Number of covenants + β10Number of lenders + β11Performance pricing + β12Total
assets + β13Leverage + β14Return on assets + β15Operating cash flow volatility +
β16Tangibility + β17Z-score + β18MB + β19Credit spread + β20Term spread +β21Prior
lending rate + Industry + Year + Loan Purpose + Loan Type + ε (2)
Number of covenants = β0 + β1Percent_Ranked_High (Majority_Ranked_High) + β2Percent_Ranked_Low
(Majority_Ranked_Low )+ β3Majority independent + β4CEO ownership + β5CEO
ownership squared + β6Independent ownership + β7Ln Facility amount + β8Ln(All-in-
drawn) + β9Ln Maturity + β10 Prior Leader + β11Number of lenders + β12 Secured +
β13Total assets + β14Leverage + β15Return on assets + β16Operating cash flow volatility
+ β17Tangibility + β18Z-score + β19MB + β20Credit spread + β21Term spread + + β22 Lead
Reputation + Industry + Year + Loan Purpose + Loan Type + ε (3)
27
and Pathak (2013), we define Interest-Increasing PP as an indicator variable that equals
one if the loan contract contains an interest increasing performance pricing provision, and
zero otherwise. We include Interest-Increasing PP in Equation (1) and find results similar
to our main results, except for the number of covenants which are no longer significant,
possibly due to the reduced sample size. Furthermore, we find some evidence that greater
director attention is negatively associated with the likelihood of having an interest-
increasing performance pricing provision.
Fifth, because dual-class firms have unique voting features that may affect the
reputation incentives of independent directors, we repeat our analyses after excluding
these firms and continue to find similar results.
Sixth, since firms with majority block holders may limit the reputation incentives
of independent directors, we re-estimate the previous test after deleting these firms and
find that our results are robust.
VI. Conclusions
In this study, we examine the relation between reputation incentives in the
director labor market and bank loan contracting. Previous literature, such as Masulis and
Mobbs (2014a, 2014b) and Huang et al. (2014), finds that directors with multiple
directorships distribute their effort unequally based on the directorships’ relative prestige.
Directors with multiple directorships spend more time and effort on their more
prestigious boards. We investigate whether bank loan contracting reflects such unequal
distribution of monitoring effort. We find that firms with a greater proportion of
independent directors for whom the board is one of their most prestigious have lower cost
of debt, longer loan maturity, and fewer covenants. These results are robust to various
28
sensitivity tests, including first differences, propensity score matching, firm-size
matching, and using alternative measures of board importance. In addition, we also find
similar results for the audit committee level. Furthermore, we find that more important
boards are associated with lower syndicate concentration and lower annual fees. Finally,
we find that directorship importance is also significantly related to bond ratings.
To the best of our knowledge, this is the first study to directly link unequal
attention of directors to the cost of debt. We contribute to the understanding of the
economic effects of director attention and how director attention affects bank loan
contracting. By linking the unequal monitoring effort to bank loan contracting, we thus
contribute to the emerging literature on directors’ unequal distribution of their efforts
among multiple directorships (e.g., Masulis and Mobbs (2014a), (2014b), Huang et al.
(2014)). Our results indicate that a firm needs to take into consideration director
reputation incentives because of the importance of this characteristic for loan pricing.
We also add to the literature linking strong board governance to lower cost of debt
(Bhojraj and Sengupta (2003), Anderson et al. (2004), and Fields et al. (2012)). We
contribute to this line of research by focusing on the unequal distribution of directors’
efforts across multiple boards, an area that has not been previously studied. Our study
highlights the importance of building a board that will receive the most attention from its
board members.
29
References
Abbott, L.; S. Parker; and G. Peters. “Audit Committee Characteristics and
Restatements.” Auditing: A Journal of Practice and Theory, 23 (2004), 69–87.
Adams, R., and D. Ferreira. “Do Directors Perform for Pay?” Journal of Accounting and
Economics, 46 (2008), 154–171.
Aggarwal, R., and J. Goodell. “Markets and Institutions in Financial Intermediation:
National Characteristics as Determinants.” Journal of Banking and Finance, 33 (2009),
1770–1780.
Ahmed, A., and S. Duellman. “Evidence on the Role of Accounting Conservatism in
Corporate Governance.” Journal of Accounting and Economics, 43 (2007), 411–437.
Anderson, C.; A. Mansi; and D. Reeb. “Founding Family Ownership and the Agency
Cost of Debt.” Journal of Financial Economics, 68 (2003), 263–283.
Asquith, P.; A. Au; T. Covert; and P. Pathak. “The Market for Borrowing Corporate
Bonds.” Journal of Financial Economics, 107 (2013), 155–182.
Beasley, M. “An Empirical Analysis of the Relation between the Board of Director
Composition and Financial Statement Fraud.” The Accounting Review, 71 (1996), 443–
465.
Beasley, M., and K. Petroni. “Board Independence and Audit-firm Type.” Auditing: A
Journal of Practice and Theory, 20 (2001), 97–114.
Beck, T.; A. Demirguc-Kunt; and R. Levine. “A New Database on Financial
Development and Structure.” World Bank mimeo (1999).
Becker, B., and T. Milbourn. “How did Increased Competition affect Credit Ratings?”
Journal of Financial Economics, 101 (2011), 493-730.
Beekes, W.; P. Pope; and S. Young. “The Link between Earnings Timeliness, Earnings
Conservatism and Board Composition: Evidence from the UK.” Corporate Governance:
An International Review, 12 (2004), 47–59.
Bhojraj, S., and P. Sengupta. “Effect of Corporate Governance on Bond Ratings and
Yields: The Role of Institutional Investors and the Outside Directors.” Journal of
Business, 76 (2003), 455–475.
Bolton, P., and D. Scharfstein. “Optimal Debt Contracts and the Number of Creditors.”
Journal of Political Economy, 104 (1996), 1–25.
Boubakri, N., and H. Ghouma. “Control/Ownership Structure, Creditor Rights Protection,
30
and the Cost of Debt Financing: International Evidence.” Journal of Banking and
Finance, 34 (2010), 2481–2499.
Bradley, M., and M. Roberts. “The Structure and Pricing of Corporate Debt Covenants.”
Working Paper, University of Pennsylvania (2004).
Bushman, R.., and R. Wittenberg-Moerman. “The Role of Bank Reputation in
“Certifying” Future Performance Implications of Borrowers’ Accounting Numbers.”
Journal of Accounting Research, 50 (2012), 883-930.
Cai, Y.; D. Dhaliwal; Y. Kim; and C. Pan. “Board Interlocks and the Diffusion of
Disclosure Policy.” Review of Accounting Studies, 19 (2014), 1086–1119.
Chava, S.; D. Livdan; and A. Purnanandam. “Do Shareholder Rights Affect the Cost of
Bank Loans?” Review of Financial Studies, 22 (2009), 2973–3004.
Chen, K., and J. Zhou. “Audit Committee, Board Characteristics and Auditor Switch
Decision by Andersen’s Clients.” Contemporary Accounting Research, 24 (2007), 1085–
1117.
Chiu, P.; S. H. Teoh; and F. Tian. “Board Interlocks and Earnings Management
Contagion.” The Accounting Review, 88 (2013), 915–944.
Coffee Jr., J. “The Rise of Dispersed Ownership: The Role of Law and the State in the
Separation of Ownership and Control.” Yale Law Journal, 111 (2001), 1–82.
Conroy, S. J., and T. L. N. Emerson. “Business Ethics and Religion: Religiosity as a
Predictor of Ethical Awareness among Students.” Journal of Business Ethics, 50 (2004),
383–396.
Core, J. E.; R. W. Holthausen; and D. F. Larcker. “Corporate Governance, Chief
Executive Officer Compensation, and Firm Performance.” Journal of Financial
Economics, 51 (1999), 371–406.
Cornwall, M.; S. L. Albrecht; P. H. Cunningham; and B. L. Pitcher. “The dimensions of
religiosity: A Conceptual Model with an Empirical Test.” Review of Religious Research,
27 (1986), 226–244.
Costello, A., and R. Wittenberg-Moerman. “The Impact of Financial Reporting Quality
on Debt Contracting: Evidence from Internal Control Weakness Reports.” Journal of
Accounting Research, 49 (2011), 97–136.
DeFond, M., R. Hann; and X. Hu. “Does the Market Value Financial Expertise on the
Audit Committees of Boards of Directors?” Journal of Accounting Research, 43 (2005),
154-194.
31
Dehejia, R.; T. Deleire; and E. Luttmer. “Insuring Consumption and Happiness through
Religious Organizations.” Working Paper, NBER (2005).
Dichev, I., and D. Skinner. “Large-Sample Evidence on the Debt Covenant Hypothesis.”
Journal of Accounting and Economics, 40 (2002), 1091–1123.
Djankov, S.; C. McLiesh; and A. Shleifer. “Private Credit in 129 Countries.” Journal of
Financial Economics, 84 (2007), 299–329.
Drucker, S., and M. Puri. “On Loan Sales, Loan Contracting and Lending Relationships.”
Review of Financial Studies, 22 (2009), 2835–2872.
Edwards, S. “Trade Orientation, Distortions, and Growth in Developing Countries.”
Journal of Development Economics, 39 (1992), 31–58.
Eum, W. “Religion and Economic Development –A Study on Religious Variables
Influencing GDP Growth over Countries.” Working Paper. University of California,
Berkeley (2011).
Fama, E. F., and J. D. MacBeth. “Risk, Return, and Equilibrium: Empirical Tests.” The
Journal of Political Economy, 81(1973), 607–636.
Ferris, S.; M. Jagannathan; and A. Pritchard. “Too Busy to Mind the Business?
Monitoring by Directors with Multiple Board Appointments, Journal of Finance, 58
(2003), 1087–1111.
Fich, E. M., and A. Shivdasani. “Are Busy Boards Effective Monitors?” Journal of
Finance, 61 (2006), 689–724.
Fields, P.; D. R. Fraser; and A. Subrahmanyam. “Board Quality and the Cost of Debt
Capital: The Case of Bank Loans.” Journal of Banking and Finance, 36 (2012), 1536–
1547.
Garcia Lara, J. M.; B. Garcia Osma; and F. Penalva. “Accounting Conservatism and
Corporate Governance.” Review of Accounting Studies, 14 (2009), 161–201.
Ghoul, S.; O. Guedhami; Y. Ni; J. Pittman; and S. Saadi. “Does Religion Matter to Equity
Pricing?” Journal of Business Ethics, 111 (2012), 491–518.
Giannett, M., and Y. Yafeh. “Do Cultural Differences between Contracting Parties
Matter? Evidence from Syndicated Bank Loans.” Management Science, 58 (2012), 365–
383.
Graham, J.; S. Li; and J. Qiu. “Corporate Misreporting and Bank Loan Contracting.”
Journal of Financial Economics, 89 (2008), 44–61.
32
Grullon, G.; G. Kanatas; and J. Weston. “Religion and Corporate (Mis)Behavior.”
Working Paper, Rice University (2010).
Hermalin, B., and M. Weisbach. “Boards of Directors as an Endogenously Determined
Institution: A Survey of the Economic Literature.” Economic Policy Review, 9 (2003), 7–
26.
Hofstede, G. Culture’s Consequences: Comparing Values, Behaviors, Institutions, and
Organizations across Nations. Thousand Oaks, CA: Sage Publications (2001).
Huang, H.; G. J. Lobo; C. Wang; and J. Zhou. “Attention Please: How Does Directors’
Uneven Attention among Multiple Directorships Affect Firms’ Earnings Management.”
Working Paper, Yeshiva University, University of Houston, University of Kentucky, and
University of Hawaii at Manoa (2014).
Iannaccone, L. “Introduction to the Economics of Religion.” Journal of Economic
Literature, 36 (1998), 1465–1496.
Ivashina, V., and D. Scharfstein. “Bank Lending during the Financial Crisis of 2008.”
Journal of Financial Economics, 97 (2010), 319–338
Jensen, M. C., and W. H. Meckling. “Theory of the Firm: Managerial Behavior, Agency
Costs and Ownership Structure.” Journal of Financial Economics, 3 (1976), 305–360.
Kanagaretnam, K.; C. Y. Lim; and G. J. Lobo. “Effects of National Culture on Earnings
Quality of Banks.” Journal of International Business Studies, 42 (2011), 853–874.
Klein, A. “Audit Committee, Board of Director Characteristics, and Earnings
Management.” Journal of Accounting and Economics, 33 (2002), 375–400.
Krishnan, J.; Y. Wen; and W. Zhao. “Legal Expertise on Corporate Audit Committees and
Financial Reporting Quality.” The Accounting Review, 86 (2011), 2099–2031.
Lagace, M. “Can Religion and Business Learn from Each Other?” Working Paper,
Harvard University (2001).
La Porta, R.; F. Lopez-de-Silanes; A. Shleifer; and R. W. Vishny. “Legal Determinants of
External Finance.” Journal of Finance, 52 (1997), 1131–1150.
La Porta, R.; F. Lopez-de-Silanes; A. Shleifer; and R. W. Vishny. “Law and Finance.”
Journal of Political Economy, 106 (1998), 1113 –1155.
La Porta, R.; F. Lopez-de-Silanes; A. Shleifer; and R. W. Vishny. “The Quality of
Governance.” Journal of Law, Economics, and Organization, 15 (1999), 222–279.
Larcker, D.; S. Richardson; and A. Tuna. “Corporate Governance, Accounting Outcomes,
33
and Organizational Performance.” The Accounting Review, 82 (2007), 963–1008.
Li, N. “Negotiated Measurement Rules in Debt Contracts.” Journal of Accounting
Research, 48 (2010): 1103–1144.
Longenecker, J. G.; J. A. McKinney; and C. W. Moore. “Religious Intensity, Evangelical
Christianity, and Business Ethics: An Empirical Study.” Journal of Business Ethics, 55
(2004), 373–386.
Mansi, S.; W. Maxwell; and D. Miller. “Analyst Forecast Characteristics and the Cost of
Debt.” Review of Accounting Studies, 16 (2011), 116–142.
Masulis, R. W., and S. Mobbs. “Independent Director Incentives: Where Do Talented
Directors Spend Their Limited Time and Energy?” Journal of Financial Economics, 111
(2014a), 406–429.
Masulis, R. W., and S. Mobbs. “Independent Director Reputation Concerns.” Working
Paper, University of New South Wales and University of Alabama (2014b).
McGuire, S.; T. Omer; and N. Sharp. “The Impact of Religion on Financial Reporting
Irregularities.” The Accounting Review, 87 (2012), 645–673.
McCleary, R. M., and R. J. Barro. “Religion and Economy.” Journal of Economic
Perspectives, 20 (2006), 49–72.
McMullen, D., and K. Raghunandan. “Enhancing Audit Committee Effectiveness.”
Journal of Accountancy, 182 (1996), 79–81.
Menon, K., and J. D. Williams. “The Use of Audit Committee for Monitoring.” Journal
of Accounting and Public Policy, 13 (1994), 121–139.
Merriman, S. Religion and the Law in America: An Encyclopedia of Personal Belief and
Public Policy. Santa Barbara, CA: ABC-CLIO (2007).
Merton, R. C. “Theory of Rational Option Pricing.” Bell Journal of Economics and
Management Science, 4 (1973), 141–183.
Milbourn, T., and B. Becker. “How Did Increased Competition Affect Credit Ratings?”
Journal of Financial Economics, 101 (2011), 493–514.
Miller, A., and J. Hoffmann. “Risk and Religion: An Explanation of Gender Differences
in Religiosity.” Journal for the Scientific Study of Religion, 34 (1995), 63–75.
Mintz, S., and R. Morris. Ethical Obligations and Decision Making in Accounting. New
York, NY: McGraw-Hill/Irwin (2011).
34
Montalvo, J., and M. Reynal-Querol. “Ethnic Diversity and Economic Development.”
Journal of Development Economics, 76 (2005), 293–323.
Osoba, B. “Risk Preferences and the Practice of Religion: Evidence from Panel Data.”
Working Paper, West Virginia University (2003).
Parboteeah, K. P.; M. Hoegl; and J. B. Cullen. “Ethics and Religion: An Empirical Test of
a Multidimensional Model.” Journal of Business Ethics, 80 (2008), 387–398.
Qi, Y.; L. Roth; and J. Wald. “Political Rights and the Cost of Debt.” Journal of Financial
Economics, 95 (2010), 202–226.
Qian, J., and P. Strahan. “How Laws and Institutions Shape Financial Contracts: The
Case of Bank Loans.” Journal of Finance, 62 (2007), 2803–2834.
Rajan, R., and L. Zingales. “The Great Reversals: The Politics of Financial Development
in the Twentieth Century. Journal of Financial Economics, 69 (2003), 5–50.
Roe, M., and J. Siegel. “Political Instability: Its Effects on Financial Development, Its
Roots in the Severity of Economic Inequality.” Working Paper, Harvard University
(2008).
Roth, L. M., and J. C. Kroll. “Risky Business: Assessing Risk Preference Explanations
for Gender Differences in Religiosity.” American Sociological Review, 72 (2007), 205–
220.
Ryan, H., and R. Wiggins. “Whose Is in Whose Pocket? Director Compensation, Board
Independence, and Barriers to Effective Monitoring.” Journal of Financial Economics,
73 (2004), 497–524.
Sengupta, P. “Corporate Disclosure Quality and the Cost of Debt.” The Accounting
Review, 73 (1998), 459–474.
Shi, L.; R. Dharwadkar; and D. Harris. “Board Interlocks and Earnings Quality.”
Working Paper, SUNY Binghamton and Syracuse University (2014).
Shivdasani, A., and D. Yermack. “CEO Involvement in the Selection of New Board
Members: An Empirical Analysis.” Journal of Finance, 54 (1990), 1829–1853.
Smith, C. “A Perspective on Accounting-Based Debt Covenant Violations.” The
Accounting Review, 68 (1993), 289–303.
Stulz, R., and R. Williamson. “Culture, Openness, and Finance.” Journal of Financial
Economics, 70 (2003), 313–349.
Sufi, A. “Information Asymmetry and Financing Arrangements: Evidence from
35
Syndicated Loans.” Journal of Finance, 62 (2007), 629–668.
Vafeas, N. “Board Meeting Frequency and Firm Performance.” Journal of Financial
Economics, 53 (1999), 113–142.
Weaver, G. R., and B. R. Agle. “Religiosity and Ethical Behavior in Organizations: A
Symbolic Interactionist Perspective.” Academy of Management Review, 27 (2002), 77–
97.
Weber, M. The Protestant Ethic and the Spirit of Capitalism. New York, NY: Harper
Collins (1930).
Williamson, O. E. “The Institutions of Governance.” American Economic Review, 88
(1998), 75–79.
Williamson, O. E. “The New Institutional Economics: Taking Stock, Looking Ahead.”
Journal of Economic Literature, 38 (2000), 359–613.
Xie, B.; W. N. Davidson; and P. DaDalt. “Earnings Management and Corporate
Governance: The Roles of the Board and the Audit Committee.” Journal of Corporate
Finance, 3 (2003): 295–316.
Yermack, D. “Higher Market Valuation of Companies with a Small Board of Directors.”
Journal of Financial Economics, 40 (1996), 185–211.
Zahn, G.C. “The Commitment Dimension.” Sociology of Religion, 31 (1970), 203–208.
Zheng, X.; S. Ghoul; O. Guedhami; and C. Kwok. “National Culture and Corporate Debt
Maturity.” Journal of Banking and Finance, 36 (2012), 468–488.
36
Appendix A
Variable name Variable definitions and constructions
Board Characteristics
Percent_Ranked_High
Percentage of independent directors for whom this ranked
directorship is 10% larger than their smallest directorship, measured
by firm market capitalization. Source: Risk Metrics.
Percent_Ranked_low
Percentage of independent directors for whom this ranked
directorship is 10% smaller than their largest directorship, measured
by firm market capitalization. Source: Risk Metrics.
Majority_Ranked_High
Indicator variable that equals 1 if the majority of independent
directors classify this as a high ranked directorship (i.e., 10% larger
than their smallest directorship by market capitalization of the firm),
and is 0 otherwise. Source: Risk Metrics.
Majority_Ranked_low
Indicator variable that equals 1 if the majority of independent
directors classify this as a low ranked directorship (i.e., 10% smaller
than their largest directorship by firm market capitalization), and 0
otherwise. Source: Risk Metrics.
High-Low Indicator variable: 1 if Majority_Ranked_High equals1 and 0
if Majority_Ranked_Low equals 1.
Majority independent Indicator variable: 1 if the majority of directors are independent
directors, and 0 otherwise. Source: Risk Metrics.
CEO ownership Percentage of common shares outstanding held by the CEO at year-
end, including stock options. Source: Risk Metrics.
CEO ownership squared Square of percentage of common shares outstanding held by the CEO
at year-end, including stock options. Source: Risk Metrics.
Independent ownership Percentage of common shares outstanding held by the independent
directors at year-end, including stock options. Source: Risk Metrics
Audit Committee
Characteristics
Audit_Percent_Ranked_High
Percentage of independent audit committee members for whom this
ranked directorship is 10% larger than their smallest directorship,
measured by firm market capitalization. Source: Risk Metrics.
Audi_ Percent_Ranked_Low
Percentage of independent audit committee members for whom this
ranked directorship is 10% smaller than their largest directorship,
measured by firm market capitalization. Source: Risk Metrics.
Audit_Majority_Ranked_High
Indicator variable that equals 1 if the majority of independent outside
directors in the audit committee classify this as a high ranked
directorship (i.e., 10% larger than their smallest directorship by firm
market capitalization), and 0 otherwise. Source: Risk Metrics.
Audit_Majority_Ranked_Low
Indicator variable that equals 1 if the majority of independent outside
directors in the audit committee classify this as a low ranked
directorship (i.e., 10% smaller than their largest directorship by firm
market capitalization), and 0 otherwise. Source: Risk Metrics.
Bank Loan Characteristics
Ln(All-in-drawn) Natural logarithm of the all-in drawn spread of each facility. Source:
DealScan
Ln (Facility amount) Natural logarithm of the actual amount (in million) of the facility
committed by the facility's lender pool. Source: DealScan
Ln(Maturity) Natural logarithm of the number of the months to maturity. Source:
DealScan
Secured Indicator variable that equals 1 if the facility has secured asset, and 0
otherwise. Source: DealScan
Number of total covenants Number of total covenants, measured as the sum of number of
general covenants and number of financial covenants. Source:
37
DealScan
Number of general covenants Number of general covenants. Source: DealScan
Number of financial covenants Number of financial covenants. Source: DealScan
Performance pricing Indicator variable that equals 1 if the loan contract includes a
performance pricing provision, and 0 otherwise. Source: DealScan
Number of lenders Natural log of the number of lenders in a loan deal. Sources:
DealScan
Loan purpose Indicator variable that equals 1 for primary loan purposes, including
acquisition line, commercial paper backup, corporate purposes, debt
repayment, takeover, working capital, and other purposes, and 0
otherwise. Source: DealScan
Loan type Indicator variable that equals 1 for loan types, including 364-day
facility, delay draw term loan, revolver, and term loan, and 0
otherwise. Source: DealScan
Syndicate concentration Herfindahl index calculated using each syndicate member’s share in
the loan. Source: DealScan
Annual fee Natural log of the number of basis points of a facility commitment
amount that a borrower is required to pay on an annual basis,
regardless of any loan outstanding. Source: DealScan
Firm characteristics
Ln (Total assets) Natural log of the book value of assets at the beginning of the fiscal
year. Source: Compustat
Leverage Total liabilities divided by total assets, measured at the beginning of
the fiscal year. Source: Compustat
Operating cash flow volatility Cash flow volatility, measured as the standard deviation of quarterly
cash flows from operations (change in quarterly Compustat data item
108) divided by total assets (Compustat data item 6) over the past
five fiscal years. Source: Compustat
Return on assets Income before extraordinary items divided by total assets, measured
at the beginning of the fiscal year. Source: Compustat
Tangibility Property, plant, and equipment (PP&E) divided by total assets,
measured at the beginning of the fiscal year. Source: Compustat
Sales growth Change in sales from prior fiscal year divided by prior fiscal year
sales. Source: Compustat
Z-score Modified Altman (1968) Z-score=(1.2 working capital + 1.4 etained
earnings + 3.3EBIT + 0.999sales)/total assets. We use a modified Z-
score, which does not include the ratio of market value of equity to
book value of total debt, measured at the beginning of the fiscal year,
because a similar term, market-to-book, is included in the regressions
as a separate variable. Source: Compustat
MB Ratio of market value of equity to book value of equity, measured at
the beginning of the fiscal year. Source: Compustat
Macroeconomic Factors
Credit spread Difference in the yield between BAA-rated and AAA-rated corporate
bonds measured one month before the loan becomes active. Source:
Federal Reserve Board of Governors
Term spread Difference in the yield between ten-year and two-year U.S. Treasury
bonds measured one month before the loan becomes active.
Source:Federal Reserve Board of Governors
Bonds characteristics
Bond Rating Coded as follows (Becker and Milbourn, 2011; (Table 2, page 500):
AAA as 28, AA+ as 26, AA as 25, AA- as 24, A+ as 23, A as 22, A-
as 21, BBB+ as 20, BBB as 19, BBB- as 18, BB+ as 17, BB as 16,
38
BB- as 15, B+ as 14, B as 13, B- as 12, CCC as 11, CC as 7, and C as
4. Source: Mergent Fixed Income Securities Database.
Ln (Bond Maturity) Natural log of number of the days to maturity (maturity date minus
offering date). Source: Mergent Fixed Income Securities Database.
Additional Firm
Characteristics for Bonds
Regressions
Idiosyncratic Risk Standard deviation of the firm’s abnormal return over the calendar
year, which equals daily stock return minus value-weighted
market return. To calculate this variable, we require at least 50
trading days each year for a given stock. Source: CRSP.
Firm Risk ROA volatility, measured as the standard deviation of quarterly
Income Before Extraordinary Items divided by total assets over the
past five fiscal years. Source: Compustat.
39
Table 1: Sample Distribution and Descriptive Statistics
Panel A: Sample Development
The sample consists of 8,512 firm-year observations from 1998 to 2011. Variable definitions are in the
Appendix
Number of firm
years
Total firm-year independent director observations available in Risk Metrics and
merged with DealScan 10,996
Less:
Observations from financial services and utilities industries (1,348)
Observations with bridge loans and non-fund-based facilities, such as leases and
standby letters of credit (325)
Observations with insufficient data to calculate control variables (811)
Final sample 8,512
Panel B: Distribution by year
Year Frequency Percentage Cumulative
1998 619 7.270 7.270
1999 676 7.940 15.21
2000 800 9.400 24.61
2001 1,187 13.95 38.56
2002 641 7.530 46.09
2003 702 8.250 54.34
2004 724 8.510 62.84
2005 600 7.050 69.89
2006 710 8.340 78.23
2007 359 4.220 82.45
2008 248 2.910 85.36
2009 398 4.680 90.04
2010 696 8.180 98.21
2011 152 1.790 100
Total 8,512 100
40
Panel C: Descriptive statistics
Variables N Mean Std. P25 Median P75
Bank Loan Characteristics
All-in-drawn 8125 175.50 149.10 62.50 150.00 250.00
Ln(All-indrawn) 8125 4.798 0.921 4.135 5.011 5.521
Maturity (months) 8293 45.63 22.49 25.00 59.00 60.00
Ln(Maturity) 8293 3.631 0.698 3.219 4.077 4.094
Facility amount(millions) 8512 506.60 745.60 100.00 250.00 550.00
Ln (Facility amount) 8512 5.494 1.262 4.605 5.521 6.310
Secured 8512 0.392 0.488 0.000 0.000 1.000
Number of total covenants 8512 3.827 3.823 0.000 3.000 6.000
Number of general covenants 8512 2.344 2.704 0.000 1.000 3.000
Number of financial covenants 8512 1.483 1.432 0.000 2.000 2.000
Performance Pricing 8512 0.487 0.500 0.000 0.000 1.000
Number of lenders 8504 10.02 9.00 4.00 8.00 13.00
Board Characteristics
Percent_Ranked_High 8512 0.192 0.245 0.000 0.111 0.333
Percent_Ranked_Low 8512 0.242 0.280 0.000 0.167 0.375
Majority_Ranked_High 8512 0.093 0.291 0.000 0.000 0.000
Majority_Ranked_Low 8512 0.123 0.328 0.000 0.000 0.000
Majority Independent 8452 0.835 0.371 1.000 1.000 1.000
CEO ownership 8512 0.025 0.063 0.001 0.007 0.020
CEO ownership squared 8512 0.005 0.030 0.000 0.000 0.000
Independent ownership 8512 0.011 0.043 0.000 0.002 0.005
Firm characteristics
Ln (Total assets) 8512 7.802 1.454 6.731 7.622 8.803
Leverage 8512 0.587 0.202 0.461 0.580 0.702
Operating cash flow volatility 8512 0.041 0.036 0.015 0.027 0.058
Return on assets 8512 0.035 0.091 0.013 0.046 0.081
Tangibility 8512 0.574 0.357 0.288 0.508 0.804
Z-score 8512 3.353 2.580 1.804 2.746 4.151
MB 8512 3.046 3.831 1.339 2.129 3.431
Credit spread 8512 1.011 0.353 0.810 0.910 1.130
Term spread 8512 1.256 1.025 0.190 1.520 2.190
41
Table 2: Relation between Directorship Importance and Cost of Debt
This table presents the OLS estimation results of the impact of directorship importance on the cost of bank
loans. The following model is used in Panel A: Ln(All-in-drawn) = β0 + β1Percent_Ranked_High
(Majority_Ranked_High) + β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The following
model is used in Panel B: Ln(All-in-drawn) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +
β2Percent_Ranked_Low (Majority_Ranked_Low) + β3 Majority Independent + β4Percent_Ranked_High
(Majority_Ranked_High)* Majority Independent + β5Percent_Ranked_Low (Majority_Ranked_Low)*
Majority Independent + Controls + ε. The dependent variable is Ln (All-in-drawn). Panel B presents the
results for testing the interaction between directorship importance and board independence with the
dependent variable Ln (All-in-drawn). All models include year, industry, loan purpose, and loan type fixed
effects. The t-statistics reported in parentheses are based on standard errors that are heteroskedasticity
robust. To conserve space, we do not report the coefficient estimates for the year, industry, loan purpose,
and loan type dummies. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% level (two-
tailed), respectively. All variables are defined in Appendix A.
42
Panel A: Directorship Importance and Cost of Bank Loans (Ln (All-in-drawn))
(1) (2) (3) (4) (5) (6)
Ln(All-in-
drawn)
Ln(All-in-
drawn)
Ln(All-in-
drawn)
Ln(All-in-
drawn)
Ln(All-in-
drawn)
Ln(All-in-
drawn)
Percent_Ranked_High -0.731*** -0.155*** -0.161***
(-19.99) (-4.62) (-5.04)
Percent_Ranked_Low 0.341*** 0.073*** 0.065***
(12.60) (2.98) (2.78)
Majority_Ranked_High -0.333*** -0.040 -0.069***
(-11.35) (-1.54) (-2.77)
Majority_Ranked_Low 0.239*** 0.045** 0.043**
(10.26) (2.20) (2.18)
Majority Independent -0.093*** -0.061*** -0.049*** -0.095*** -0.060*** -0.048***
(-4.34) (-3.36) (-2.80) (-4.41) (-3.30) (-2.75)
CEO Ownership 1.018*** 0.337* 0.192 1.207*** 0.334* 0.185
(4.09) (1.71) (0.99) (4.76) (1.69) (0.96)
CEO Ownership Square -1.703*** -0.535 -0.312 -1.950*** -0.507 -0.282
(-3.37) (-1.45) (-0.84) (-3.88) (-1.37) (-0.76)
Independent Ownership -0.203 -0.351** -0.309** -0.122 -0.338** -0.299**
(-1.20) (-2.56) (-2.21) (-0.69) (-2.44) (-2.12)
Ln (Facility amount) -0.101*** -0.100***
(-12.36) (-12.33)
Ln(Maturity) 0.000 -0.001
(0.01) (-0.04)
Number of total covenants 0.047*** 0.047***
(24.16) (24.25)
Number of lenders -0.002** -0.002**
(-2.37) (-2.40)
Performance Pricing -0.185*** -0.186***
(-13.56) (-13.57)
Ln (Total Assets) -0.211*** -0.121*** -0.222*** -0.131***
(-31.50) (-14.67) (-35.61) (-16.39)
Leverage 0.502*** 0.453*** 0.508*** 0.458***
(10.88) (10.20) (10.91) (10.25)
Return on assets -1.660*** -1.382*** -1.676*** -1.399***
(-18.09) (-15.50) (-18.20) (-15.62)
Operating cash flow volatility 0.0450 0.271 0.0470 0.271
(0.24) (1.52) (0.25) (1.52)
Tangibility -0.121*** -0.099*** -0.120*** -0.098***
(-5.29) (-4.48) (-5.24) (-4.44)
Z-score -0.043*** -0.043*** -0.045*** -0.044***
(-10.91) (-11.05) (-11.24) (-11.30)
MB -0.017*** -0.016*** -0.018*** -0.017***
(-8.59) (-8.57) (-9.10) (-9.05)
Credit_ Spread 0.247*** 0.212*** 0.249*** 0.214***
(10.59) (9.35) (10.64) (9.39)
Term_ Spread 0.164*** 0.150*** 0.165*** 0.151***
(14.65) (13.26) (14.75) (13.33)
Intercept 4.330*** 5.636*** 5.485*** 4.271*** 5.705*** 5.546***
(57.13) (55.74) (50.38) (55.27) (56.91) (51.41)
No. of observations 8066 8066 7852 8066 8066 7852
Adjusted R2 0.539 0.669 0.701 0.514 0.668 0.700
Test of β1=β2
F
524.73***
27.93***
30.21***
224.70***
6.40**
11.99***
43
Panel B: Interaction between Directorship Importance and Board Independence
(1) (2)
Ln(All-in-drawn) Ln(All-in-drawn)
Percent_Ranked_High 0.002
(0.05)
Percent_Ranked_ Low -0.040
(-0.98)
Majority_Ranked_High 0.055
(1.14)
Majority_Ranked_Low -0.015
(-0.39)
Majority Independent -0.036 -0.040**
(-1.55) (-2.05)
Percent_Ranked_High* Majority Independent -0.244***
(-4.12)
Percent_Ranked_ Low* Majority Independent 0.146***
(3.02)
Majority_Ranked_High* Majority Independent -0.160***
(-2.94)
Majority_Ranked_Low* Majority Independent 0.071
(1.60)
CEO Ownership 0.183 0.196
(0.95) (1.01)
CEO Ownership Square -0.246 -0.275
(-0.67) (-0.75)
Independent Ownership -0.319** -0.310**
(-2.31) (-2.21)
Ln (Facility amount) -0.100*** -0.100***
(-12.31) (-12.28)
Ln(Maturity) -0.001 -0.002
(-0.06) (-0.10)
Number of total covenants 0.047*** 0.047***
(24.19) (24.25)
Number of lenders -0.002** -0.002**
(-2.36) (-2.40)
Performance Pricing -0.187*** -0.187***
(-13.69) (-13.58)
Ln (Total Assets) -0.117*** -0.130***
(-14.18) (-16.34)
Leverage 0.433*** 0.448***
(9.76) (10.05)
Return on assets -1.371*** -1.395***
(-15.50) (-15.66)
Operating cash flow volatility 0.287 0.270
(1.61) (1.51)
Tangibility -0.101*** -0.098***
(-4.55) (-4.43)
Z-score -0.043*** -0.044***
(-11.10) (-11.33)
MB -0.016*** -0.017***
(-8.39) (-9.11)
Credit_ Spread 0.215*** 0.214***
(9.47) (9.44)
Term_ Spread 0.150*** 0.151***
(13.35) (13.45)
44
Intercept 5.468*** 5.547***
(50.20) (51.52)
No. of observations 7852 7852
Adjusted R2 0.702 0.700
Test of β4=β5
F 24.13*** 9.52***
45
Table 3: Relation between Directorship Importance and Loan Maturity
This table presents the OLS estimation results in Panel A and Poisson regression results in Panel B of the
impact of directorship importance on the loan maturity. The following model is used:
Ln(Maturity) (Maturity) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +
β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Ln (Maturity)
in Panel A and Maturity in Panel B. All models include year, industry, loan purpose, and loan type fixed
effects. The t-statistics and z-statistics reported in parentheses are based on standard errors that are
heteroskedasticity robust. To conserve space, we do not report the coefficient estimates for the year,
industry, loan purpose, and loan type dummies. *, **, and *** indicate statistical significance at the 10%,
5%, and 1% levels (two-tailed), respectively. All variables are defined in Appendix A.
46
Panel A: Directorship Importance and Loan Maturity (Ln (Maturity))
(1) (2) (3) (4) (5) (6)
Ln(Maturity) Ln(Maturity) Ln(Maturity) Ln(Maturity) Ln(Maturity) Ln(Maturity)
Percent_Ranked_ High 0.051*** 0.048** 0.052**
(2.58) (2.26) (2.45)
Percent_Ranked_Low 0.003 -0.007 -0.003
(0.17) (-0.39) (-0.17)
Majority_Ranked_High 0.046*** 0.050*** 0.051***
(3.20) (3.41) (3.48)
Majority_Ranked_Low -0.009 -0.015 -0.015
(-0.56) (-1.00) (-1.00)
Majority Independent -0.032** -0.029** -0.031** -0.031** -0.028** -0.030**
(-2.49) (-2.28) (-2.45) (-2.41) (-2.23) (-2.39)
CEO Ownership 0.093 0.131 0.111 0.080 0.133 0.109
(0.62) (0.89) (0.76) (0.53) (0.90) (0.75)
CEO Ownership Square -0.067 -0.175 -0.146 -0.046 -0.178 -0.143
(-0.26) (-0.69) (-0.57) (-0.17) (-0.70) (-0.56)
Independent Ownership 0.060 0.031 0.015 0.058 0.030 0.014
(0.78) (0.40) (0.21) (0.75) (0.39) (0.19)
Ln (Facility amount) 0.063*** 0.062***
(9.69) (9.65)
Number of total covenants 0.002 0.0010
(1.04) (0.95)
Number of lenders 0.001 0.0010
(1.15) (1.17)
Performance Pricing 0.017* 0.019*
(1.70) (1.84)
Ln (Total Assets) -0.010** -0.049*** -0.009** -0.048***
(-2.29) (-8.34) (-2.26) (-8.39)
Leverage 0.149*** 0.117*** 0.151*** 0.121***
(4.83) (3.86) (4.91) (3.99)
Return on assets 0.560*** 0.441*** 0.565*** 0.445***
(8.10) (6.45) (8.16) (6.50)
Operating cash flow volatility -0.463*** -0.484*** -0.455*** -0.476***
(-3.45) (-3.66) (-3.39) (-3.60)
Tangibility -0.021 -0.022 -0.022 -0.022
(-1.37) (-1.44) (-1.39) (-1.45)
Z-score -0.012*** -0.011*** -0.012*** -0.011***
(-4.48) (-4.18) (-4.52) (-4.21)
MB -0.003*** -0.003*** -0.003*** -0.004***
(-2.59) (-2.85) (-2.61) (-2.88)
Credit_ Spread -0.139*** -0.126*** -0.139*** -0.125***
(-7.58) (-7.09) (-7.56) (-7.07)
Term_ Spread -0.033*** -0.029*** -0.033*** -0.029***
(-4.33) (-3.86) (-4.35) (-3.88)
Intercept 2.497*** 2.701*** 2.638*** 2.500*** 2.697*** 2.633***
(34.04) (32.34) (31.79) (34.08) (32.40) (31.87)
No. of observations 8236 8236 8228 8236 8236 8228
Adjusted R2 0.706 0.716 0.722 0.706 0.716 0.722
Test of β1=β2
F 2.88* 3.52* 3.52* 6.36** 8.92*** 9.35***
47
Panel B: Directorship Importance and Loan Maturity (Maturity)
(1) (2) (3) (4) (5) (6)
Maturity Maturity Maturity Maturity Maturity Maturity
Percent_Ranked_High 0.055*** 0.047** 0.050***
(3.24) (2.50) (2.73)
Percent_Ranked_Low 0.006 -0.003 -0.000
(0.41) (-0.22) (-0.01)
Majority_Ranked_High 0.057*** 0.055*** 0.055***
(4.18) (3.98) (3.95)
Majority_Ranked_Low -0.001 -0.008 -0.008
(-0.09) (-0.62) (-0.63)
Majority Independent -0.028** -0.025** -0.025** -0.027** -0.023** -0.024**
(-2.55) (-2.28) (-2.31) (-2.41) (-2.17) (-2.20)
CEO Ownership 0.002 0.039 0.033 -0.005 0.045 0.036
(0.02) (0.34) (0.30) (-0.04) (0.39) (0.32)
CEO Ownership Square 0.017 -0.082 -0.070 0.030 -0.090 -0.073
(0.09) (-0.42) (-0.36) (0.15) (-0.46) (-0.38)
Independent Ownership 0.046 0.018 -0.000 0.046 0.017 -0.001
(0.71) (0.27) (-0.00) (0.71) (0.26) (-0.02)
Ln (Facility amount) 0.057*** 0.057***
(11.34) (11.28)
Number of total covenants 0.002 0.002
(1.59) (1.47)
Number of lenders 0.000 0.000
(0.73) (0.77)
Performance Pricing -0.009 -0.008
(-1.09) (-0.96)
Ln (Total Assets) -0.007** -0.040*** -0.007** -0.039***
(-2.04) (-8.34) (-2.00) (-8.41)
Leverage 0.140*** 0.113*** 0.140*** 0.114***
(5.77) (4.73) (5.80) (4.81)
Return on assets 0.438*** 0.346*** 0.445*** 0.352***
(7.96) (6.36) (8.12) (6.49)
Operating cash flow volatility -0.409*** -0.424*** -0.400*** -0.416***
(-3.71) (-3.92) (-3.63) (-3.85)
Tangibility -0.023* -0.026** -0.024* -0.026**
(-1.86) (-2.07) (-1.87) (-2.07)
Z-score -0.010*** -0.009*** -0.010*** -0.009***
(-4.41) (-4.21) (-4.49) (-4.27)
MB -0.002** -0.003** -0.002** -0.003**
(-2.22) (-2.41) (-2.23) (-2.41)
Credit_ Spread -0.145*** -0.131*** -0.145*** -0.130***
(-9.95) (-9.21) (-9.92) (-9.18)
Term_ Spread -0.033*** -0.031*** -0.033*** -0.031***
(-5.54) (-5.27) (-5.57) (-5.30)
Intercept 2.498*** 2.682*** 2.617*** 2.499*** 2.677*** 2.612***
(48.64) (43.43) (42.10) (48.55) (43.67) (42.33)
No. of observations 8236 8236 8228 8236 8236 8228
Pseudo R2 0.465 0.474 0.480 0.465 0.474 0.480
Test of β1=β2
Chi2 4.47** 4.08** 4.28** 9.63*** 11.06*** 11.22***
48
Table 4: Relation between Directorship Importance and Number of Covenants
This table presents the Poisson regression results of the impact of multiple-directorship on the number of
covenants. The following model is used: Number of Total Covenants (Number of General Covenants or
Number of Financial Covenants) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +
β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Number of total
covenants in Panel A, Number of general covenants in Panel B, and Number of financial covenants in Panel
C. All models include year, industry, loan purpose, and loan type fixed effects. The z-statistics reported in
the parentheses are based on standard errors that are heteroskedasticity robust. To conserve space, we do
not report the coefficient estimates for the year, industry, loan purpose, and loan type dummies. *, **, and
*** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. All variables
are defined in Appendix A.
49
Panel A: Directorship Importance and Number of Total Covenants
(1) (2) (3) (4) (5) (6)
Number of
Total
Covenants
Number of
Total
Covenants
Number of
Total
Covenants
Number of
Total
Covenants
Number of
Total
Covenants
Number of
Total
Covenants
Percent_Ranked_High -0.359*** -0.063 -0.079*
(-7.27) (-1.24) (-1.85)
Percent_Ranked_Low 0.108*** 0.005 -0.017
(2.98) (0.15) (-0.52)
Majority_Ranked_High -0.179*** -0.026 -0.096**
(-4.21) (-0.62) (-2.51)
Majority_Ranked_Low 0.075** -0.001 -0.017
(2.42) (-0.04) (-0.62)
Majority Independent -0.049* -0.024 -0.015 -0.052* -0.024 -0.017
(-1.82) (-0.93) (-0.64) (-1.94) (-0.93) (-0.75)
CEO Ownership 1.187*** 0.788** 0.296 1.247*** 0.784** 0.277
(3.52) (2.32) (0.98) (3.69) (2.31) (0.91)
CEO Ownership Square -2.239*** -1.526* -0.568 -2.291*** -1.507* -0.538
(-2.64) (-1.82) (-0.77) (-2.68) (-1.80) (-0.74)
Independent Ownership 0.170 -0.102 0.090 0.194 -0.102 0.090
(0.85) (-0.49) (0.47) (0.98) (-0.49) (0.47)
Ln (Facility amount) 0.054*** 0.054***
(4.70) (4.73)
Ln(Maturity) 0.015 0.016
(0.59) (0.64)
Prior Leader -0.016 -0.016
(-0.94) (-0.92)
Number of lenders 0.012*** 0.012***
(9.54) (9.54)
Secured 0.828*** 0.830***
(37.31) (37.36)
Ln (Total Assets) -0.155*** -0.145*** -0.159*** -0.147***
(-15.70) (-11.90) (-17.45) (-12.49)
Leverage 0.442*** 0.204*** 0.442*** 0.205***
(7.46) (3.83) (7.45) (3.86)
Return on assets 0.124 0.591*** 0.118 0.580***
(1.04) (5.39) (0.99) (5.28)
Operating cash flow volatility -1.463*** -1.316*** -1.465*** -1.328***
(-4.89) (-5.02) (-4.90) (-5.06)
Tangibility -0.104*** -0.059* -0.104*** -0.062*
(-2.83) (-1.82) (-2.85) (-1.91)
Z-score -0.024*** -0.003 -0.025*** -0.003
(-4.30) (-0.73) (-4.41) (-0.67)
MB -0.001 0.000 -0.002 -0.000
(-0.56) (0.04) (-0.71) (-0.06)
Credit_ Spread 0.019 0.026 0.019 0.025
(0.56) (0.90) (0.57) (0.88)
Term_ Spread 0.043*** 0.029* 0.044*** 0.029*
(2.62) (1.90) (2.64) (1.89)
Intercept 1.121*** 2.136*** 1.408*** 1.089*** 2.163*** 1.417***
(6.34) (11.41) (7.81) (6.09) (11.63) (7.91)
No. of observations 8452 8452 8228 8452 8452 8228
Pseudo R2 0.161 0.178 0.259 0.159 0.178 0.260
Test of β1=β2
Chi2 60.05*** 1.18 1.37 23.57*** 0.22 2.88*
50
Panel B: Directorship Importance and Number of General Covenants
(1) (2) (3) (4) (5) (6)
Number of
General
Covenants
Number of
General
Covenants
Number of
General
Covenants
Number of
General
Covenants
Number of
General
Covenants
Number of
General
Covenants
Percent_Ranked_High -0.359*** -0.060 -0.073
(-6.20) (-0.99) (-1.47)
Percent_Ranked_Low 0.148*** 0.0110 -0.021
(3.61) (0.27) (-0.58)
Majority_Ranked_High -0.135*** 0.021 -0.062
(-2.76) (0.43) (-1.38)
Majority_Ranked_Low 0.088** -0.010 -0.031
(2.53) (-0.27) (-1.03)
Majority Independent -0.032 -0.002 0.015 -0.033 0.000 0.013
(-1.03) (-0.05) (0.54) (-1.06) (0.01) (0.49)
CEO Ownership 1.110*** 0.796** 0.208 1.164*** 0.795** 0.192
(2.82) (2.00) (0.59) (2.97) (2.00) (0.55)
CEO Ownership Square -1.730* -1.174 -0.054 -1.753* -1.152 -0.024
(-1.85) (-1.25) (-0.07) (-1.88) (-1.24) (-0.03)
Independent Ownership 0.157 -0.116 0.127 0.178 -0.121 0.121
(0.63) (-0.45) (0.56) (0.72) (-0.48) (0.54)
Ln (Facility amount) 0.063*** 0.064***
(4.89) (4.92)
Ln(Maturity) 0.028 0.028
(0.97) (0.99)
Prior Leader -0.011 -0.011
(-0.58) (-0.56)
Number of lenders 0.011*** 0.011***
(8.39) (8.41)
Secured 1.076*** 1.077***
(39.21) (39.22)
Ln (Total Assets) -0.155*** -0.126*** -0.161*** -0.129***
(-13.61) (-9.32) (-15.43) (-9.92)
Leverage 0.534*** 0.249*** 0.533*** 0.251***
(7.70) (3.94) (7.70) (3.97)
Return on assets -0.067 0.475*** -0.073 0.462***
(-0.49) (3.79) (-0.53) (3.70)
Operating cash flow volatility -1.665*** -1.493*** -1.659*** -1.498***
(-4.63) (-4.80) (-4.61) (-4.82)
Tangibility -0.087** -0.040 -0.087** -0.044
(-2.06) (-1.07) (-2.06) (-1.16)
Z-score -0.034*** -0.007 -0.036*** -0.007
(-4.95) (-1.19) (-5.14) (-1.22)
MB -0.002 0.000 -0.002 -0.000
(-0.55) (0.14) (-0.79) (-0.01)
Credit_ Spread 0.051 0.061* 0.052 0.060*
(1.33) (1.89) (1.35) (1.88)
Term_ Spread 0.040** 0.019 0.040** 0.019
(2.09) (1.13) (2.11) (1.11)
Intercept 0.669*** 1.661*** 0.632*** 0.635*** 1.705*** 0.654***
(3.21) (7.57) (2.95) (3.02) (7.82) (3.07)
No. of observations 8452 8452 8228 8452 8452 8228
Pseudo R2 0.158 0.175 0.272 0.155 0.175 0.272
Test of β1=β2
Chi2 52.04*** 0.91 0.70 13.61*** 0.25 0.31
51
Panel C: Directorship Importance and Number of Financial Covenants
(1) (2) (3) (4) (5) (6)
Number of
Financial
Covenants
Number of
Financial
Covenants
Number of
Financial
Covenants
Number of
Financial
Covenants
Number of
Financial
Covenants
Number of
Financial
Covenants
Percent_Ranked_High -0.358*** -0.066 -0.081*
(-7.51) (-1.35) (-1.81)
Percent_Ranked_Low 0.042 -0.005 -0.013
(1.16) (-0.13) (-0.38)
Majority_Ranked_High -0.245*** -0.099** -0.142***
(-5.72) (-2.30) (-3.45)
Majority_Ranked_Low 0.053* 0.013 0.006
(1.71) (0.41) (0.22)
Majority Independent -0.077*** -0.058** -0.056** -0.083*** -0.060** -0.059**
(-2.91) (-2.21) (-2.25) (-3.14) (-2.30) (-2.40)
CEO Ownership 1.345*** 0.795** 0.458 1.413*** 0.785** 0.437
(3.89) (2.35) (1.40) (4.04) (2.31) (1.34)
CEO Ownership Square -3.206*** -2.200** -1.546* -3.298*** -2.185** -1.520*
(-3.50) (-2.54) (-1.81) (-3.53) (-2.53) (-1.79)
Independent Ownership 0.216 -0.037 0.077 0.242 -0.031 0.083
(1.16) (-0.18) (0.38) (1.30) (-0.15) (0.41)
Ln (Facility amount) 0.034*** 0.034***
(2.80) (2.82)
Ln(Maturity) -0.010 -0.008
(-0.38) (-0.30)
Prior Leader -0.024 -0.024
(-1.34) (-1.33)
Number of lenders 0.014*** 0.014***
(10.44) (10.44)
Secured 0.480*** 0.483***
(22.39) (22.51)
Ln (Total Assets) -0.155*** -0.172*** -0.156*** -0.171***
(-15.73) (-13.19) (-16.88) (-13.57)
Leverage 0.276*** 0.114** 0.274*** 0.114**
(4.55) (1.99) (4.55) (2.00)
Return on assets 0.469*** 0.776*** 0.466*** 0.770***
(3.83) (6.48) (3.82) (6.43)
Operating cash flow volatility -1.172*** -1.075*** -1.187*** -1.096***
(-4.04) (-3.89) (-4.09) (-3.96)
Tangibility -0.135*** -0.098*** -0.135*** -0.100***
(-3.67) (-2.82) (-3.68) (-2.87)
Z-score -0.013** -0.001 -0.013** -0.000
(-2.41) (-0.26) (-2.35) (-0.09)
MB -0.001 -0.000 -0.001 -0.000
(-0.39) (-0.16) (-0.37) (-0.12)
Credit_ Spread -0.039 -0.038 -0.040 -0.039
(-1.17) (-1.17) (-1.20) (-1.21)
Term_ Spread 0.048*** 0.043*** 0.048*** 0.043***
(2.99) (2.76) (3.00) (2.75)
Intercept 0.0500 1.132*** 0.869*** 0.0210 1.132*** 0.856***
(0.34) (6.57) (4.99) (0.14) (6.62) (4.96)
No. of observations 8452 8452 8228 8452 8452 8228
Pseudo R2 0.104 0.115 0.143 0.103 0.115 0.143
Test of β1=β2
Chi2 47.22*** 1.04 1.51 32.49*** 4.44** 8.86***
52
Table5: Relation between Directorship Importance and Loan Contract Terms: Changes Analysis
This table presents estimation results relating changes in directorship importance (from t-2 to t year) to
changes in bank loan contracting (from t-1 to t+1 year). The following model is used: Bank Loan
Contracting = β0 + β1Percent_Ranked_High (Majority_Ranked_High) + β2Percent_Ranked_Low
(Majority_Ranked_Low) + Controls + ε. The t-statistics reported in parentheses are based on standard
errors that are heteroskedasticity robust. *, **, and *** indicate statistical significance at the 10%, 5%, and
1% levels (two-tailed), respectively. All variables are defined in Appendix A.
(1) (2) (3) (4) (5)
Ln(All-in-
drawn) Maturity Number of
Total
Covenants
Number of
General
Covenants
Number of
Financial
Covenants
Percent_Ranked_High -0.189** 0.135* -0.455 -0.264 -0.202
(-2.29) (1.78) (-1.42) (-1.13) (-1.43)
Percent_Ranked_Low 0.096 0.102 0.913** 0.666** 0.253
(0.94) (1.22) (2.19) (2.15) (1.50)
Controls in Equation (1) YES YES YES YES YES
No. of observations 1557 1687 1675 1675 1675
Test of β1=β2
F/ Chi2 4.98** 0.09 7.37*** 6.93*** 4.71**
53
Table 6: Relation between Directorship Importance and Loan Contract Terms: Propensity Score
Matching
This table presents estimation results of the impact of directorship importance on bank loan contracting
after explicitly controlling for firm size differences using a propensity score matched control sample. Panel
A presents the probit estimation results of the following first-stage propensity score model: High_Low = β0
+ β1Ln(MV) + β2 Leverage + β3 Return on assets + β4 Operating earnings volatility + β5 Percent Female +
β6 Majority Independent + ε. Panel B reports results of the following model using this matched sample to
examine the relations between directorship importance and loan contract terms: Bank Loan Contracting = β0 + β1High_Low + Controls + ε. In Panel B, Column (1) reports OLS regression results, and Columns (2)-
(5) report Poisson regression results. In the first stage, all models include year and industry fixed effects. To
conserve space, we do not report the coefficient estimates for the year and industry dummies. The t- and z-
statistics reported in parentheses are based on standard errors that are heteroskedasticity robust. *, **, and
*** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. All variables
are defined in Appendix A.
Panel A: First Stage Propensity Score Matching
First Stage:
High_Low
Ln(MV) 1.549***
(14.97)
Leverage -0.409
(-0.73)
Return on assets 1.022
(0.93)
Operating earnings volatility 3.529**
(2.18)
Percent Female -2.680***
(-3.68)
Majority Independent -0.540**
(-2.14)
Intercept -10.166***
(-10.51)
Industry fixed effect Included
Year fixed effect Included
No. of observations 1569
Pseudo R2 0.590
54
Panel B: Directorship Importance and Loan Contract Terms: Propensity Score Matched Control
Sample
(1) (2) (3) (4) (5)
Ln(All-in-
drawn) Maturity Number of
Total Covenants
Number of
General Covenants
Number of
Financial Covenants
High_Low -0.118** 0.032 -0.236*** -0.232** -0.251***
(-2.37) (1.25) (-2.90) (-2.38) (-3.08)
Controls in Equation (1) YES YES YES YES YES
Intercept 5.806*** 2.819*** -0.435 -0.772 -1.886**
(10.88) (12.26) (-0.55) (-0.78) (-2.47)
No. of observations 591 614 614 614 614
55
Table 7: Relation between Directorship Importance of Audit Committee and Loan Contract Terms
This table presents estimation results of the impact of audit committee directorship importance on bank
loan items based on the following model: Bank Loan Contracting =β0 + β1 Audit Percent_Ranked_High
(Audit Majority_Ranked_High) + β2 Audit Percent_Ranked_Low (Audit Majority_Ranked_Low) + Controls
+ ε. Panel A reports the relation between percentage of audit committee directorship importance and loan
contract terms. Panel B reports the relation between majority of audit committee directorship importance
and loan contract terms. Column (1) reports OLS regression results, and Columns (2)-(5) report Poisson
regression results. All models include year, industry, loan purpose, and loan type fixed effects. The t-
statistics reported in parentheses are based on standard errors that are heteroskedasticity robust. To
conserve space, we do not report the coefficient estimates for the year, industry, loan purpose, and loan type
dummies. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed),
respectively. All variables are defined in Appendix A.
Panel A: Relation between Percentage of Audit Committee Directorship Importance and Loan
Contract Terms
(1) (2) (3) (4) (5)
Ln(All-in-
drawn) Maturity Number of
Total
Covenants
Number of
General
Covenants
Number of
Financial
Covenants
Audit Percent_Ranked_High -0.042 0.022 -0.103*** -0.095** -0.108***
(-1.61) (1.47) (-2.90) (-2.28) (-2.83)
Audit Percent_Ranked_Low 0.062*** -0.018 0.008 0.008 0.004
(2.90) (-1.41) (0.28) (0.24) (0.13)
Controls in Equation (1) YES YES YES YES YES
No. of observations 7362 7714 7714 7714 7714
Test of β1=β2
F/ Chi2 8.95*** 3.67* 5.05** 3.19* 4.60**
Panel B: Relation between Majority of Audit Committee Directorship Importance and Loan
Contract Terms
(1) (2) (3) (4) (5)
Ln(All-in-
drawn) Maturity Number of
Total
Covenants
Number of
General
Covenants
Number of
Financial
Covenants
Audit Majority_Ranked_High -0.038* 0.031** -0.073** -0.068* -0.077**
(-1.66) (2.55) (-2.24) (-1.76) (-2.18)
Audit Majority_Ranked_Low 0.050** -0.017 -0.018 -0.038 0.012
(2.55) (-1.46) (-0.66) (-1.23) (0.40)
Controls in Equation (1) YES YES YES YES YES
No. of observations 7362 7714 7714 7714 7714
Test of β1=β2
F/ Chi2 7.97*** 7.48*** 1.55 0.33 3.54*
56
Table 8: Relationship between Directorship Importance and Syndicate Concentration
This table presents OLS estimation results of the impact of directorship importance on syndicate
concentration based on the following model: Syndicate Concentration = β0 + β1Percent_Ranked_High
(Majority_Ranked_High) + β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The
dependent variable is Syndicate Concentration. All models include year, industry, loan purpose, and loan
type fixed effects. The t-statistics reported in parentheses are based on standard errors that are
heteroskedasticity robust. To conserve space, we do not report the coefficient estimates for the year,
industry, loan purpose, and loan type dummies. *, **, and *** indicate statistical significance at the 10%,
5%, and 1% levels (two-tailed), respectively. All variables are defined in Appendix.
(1) (2) (3) (4)
Syndicate
Concentration
Syndicate
Concentration
Syndicate
Concentration
Syndicate
Concentration
Percent_Ranked_High -0.041*** -0.003
(-7.03) (-0.54)
Percent_RankedLow -0.001 0.003
(-0.16) (0.42)
Majority_Ranked_High -0.024*** -0.007*
(-6.69) (-1.81)
Majority_Ranked_Low 0.004 0.006
(0.73) (1.09)
Controls Board
characteristics
Controls in
Equation (1)
Board
characteristics
Controls in
Equation (1)
No. of observations 8444 8147 8444 8147
Test of β1=β2
F/ Chi2 20.55*** 0.42 17.13*** 3.40*
57
Table 9: Relation between Directorship Importance and Transaction Fees: Annual Fees
This table presents OLS estimation results of the impact of directorship importance on annual fee based on
the following model: Ln(Annual Fee) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +
β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Ln(Annual
Fee). All models include year, industry, loan purpose, and loan type fixed effects. The t-statistics reported
in parentheses are based on standard errors that are heteroskedasticity robust. To conserve space, we do not
report the coefficient estimates for the year, industry, loan purpose, and loan type dummies. *, **, and ***
indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. All variables are
defined in Appendix A.
(1) (2) (3) (4)
Ln(Annual
Fee)
Ln(Annual
Fee)
Ln(Annual
Fee)
Ln(Annual
Fee)
Percent_RankedHigh -0.640*** -0.108**
(-14.08) (-2.51)
Percent_Ranked_Low 0.295*** 0.069
(6.18) (1.58)
Majority_Ranked_High -0.276*** -0.017
(-10.07) (-0.73)
Majority_Ranked_Low 0.158*** 0.030
(4.17) (0.92)
Controls Board
characteristics
Controls in
Equation (1)
Board
characteristics
Controls in
Equation (1)
No. of observations 2708 2659 2708 2659
Test of β1=β2
F/ Chi2 205.25*** 7.81*** 89.08*** 1.41
58
Table 10: Relation between Directorship Importance and Bonds Rating
This table presents OLS estimation results of the impact of directorship importance on bond rating based on
the following model: Bond Rating = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +
β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Bond Rating.
All models include year, industry fixed effects. The t-statistics reported in parentheses are based on
standard errors that are heteroskedasticity robust. To conserve space, we do not report the coefficient
estimates for the year, industry dummies. *, **, and *** indicate statistical significance at the 10%, 5%,
and 1% levels (two-tailed), respectively. All variables are defined in Appendix A.
(1) (2) (3) (4)
Bond Rating Bond Rating Bond Rating Bond Rating
Percent_Ranked_High 4.958*** 0.629*** (46.14) (6.13)
Percent_Ranked_Low -2.722*** -0.136
(-23.36) (-1.41) Majority_Ranked_High 1.956*** 0.323***
(35.58) (6.46) Majority_Ranked_Low -1.594*** 0.098
(-16.14) (1.29) Majority Independent 0.560*** 0.285*** 0.716*** 0.298***
(6.22) (4.34) (8.03) (4.57)
CEO Ownership -18.248*** -10.516*** -22.173*** -10.396*** (-17.04) (-12.36) (-19.75) (-12.12)
CEO Ownership Square 24.540*** 13.787*** 30.381*** 13.328*** (11.67) (7.76) (14.14) (7.39)
Independent Ownership 1.462*** 3.461*** 1.663*** 3.623***
(3.28) (8.05) (3.26) (8.27) Idiosyncratic Risk -104.446*** -105.258***
(-28.30) (-28.48) Ln (Bonds Maturity) 0.109*** 0.109***
(3.02) (3.04) Ln (Total Assets) 1.217*** 1.251***
(53.66) (60.77)
Leverage -2.024*** -2.047*** (-11.58) (-11.74)
Return on assets 4.744*** 4.842*** (9.76) (9.96)
Firm Risk -2.936** -2.666**
(-2.13) (-1.97)
Z-score 0.597*** 0.602***
(25.49) (25.71)
MB 0.016** 0.018***
(2.34) (2.67) Intercept 20.445*** 10.861*** 21.775*** 10.688***
(77.92) (21.38) (70.61) (20.96)
No. of observations 21199 11377 21199 11377 Adjusted R2 0.345 0.718 0.271 0.718
Test of β1=β2 F 2740.31*** 29.41*** 1004.89*** 5.89**