regulation of compensationregulation of compensation anya kleymenova*...
Post on 15-Jan-2020
1 Views
Preview:
TRANSCRIPT
Regulation of Compensation
Anya Kleymenova*
Anya.Kleymenova@chicagobooth.edu
University of Chicago Booth School of Business
İrem Tuna*
ituna@london.edu
London Business School
This version: March 2017
Abstract This paper studies the consequences of regulating executive compensation at financial institutions using
the setting of the introduction of the UK Remuneration Code and the EU bonus cap regulation. Our
analysis indicates that while the initial reaction to the Remuneration Code was positive, the stock market
reacted negatively to the EU bonus cap regulation, suggesting that equity market investors perceive at
least some costs from regulating executive compensation. In line with the intent of regulation, we also
find that UK banks defer more bonuses and reduce risk. However, when compared to their US
counterparts and other UK firms, UK banks also experience higher CEO turnover. Finally, we find that
UK banks' compensation contracts become more complex after the regulation. Therefore, while
regulation may have had the desired effect in terms of risk-taking, it may also have given rise to some
unintended costs.
JEL classification: G28, G34, G38 Keywords: Executive compensation; financial institutions; regulation; EU bonus caps; UK Remuneration
Code.
*Anya Kleymenova gratefully acknowledges the financial support of the Accounting Research Center at Chicago Booth, the
Harry W. Kirchheimer and the Centel Foundation/Robert P. Reuss Faculty Research Funds at the University of Chicago Booth
School of Business, the University of Chicago Booth School of Business, the Economic & Social Research Council, and London
Business School. İrem Tuna gratefully acknowledges the financial support of the European Research Council (Grant ERC-2010-
263525). We thank Christopher Armstrong, Eli Bartov, Phil Berger, Thomas Bourveau (discussant), Francois Brochet
(discussant), Robert Bushman, Hans Christensen, Johanna Cowan, Yiwei Dou, Fabrizio Ferri, Pingyang Gao, Bjorn Jorgensen,
Kinda Hachem, Sam Harrington, Mirko Heinle, Lord King of Lothbury, Ningzhong Li, Xiumin Martin, Mehir Mehta, Joshua
Ronen, Stephen Ryan, Haresh Sapra, Ron Shalev, Douglas Skinner, Eddie Riedl, Tjomme Rusticus, Marshall Vance (discussant),
Martin Walker (discussant), Christopher Williams, Anastasia Zakolyukina, Luigi Zingales, and the participants at the Manchester
Business School Executive Compensation conference, the 2013 EAA conference, the 2015 GW Cherry Blossom Conference, the
2015 UCLA Accounting Conference, the 2015 GIA Conference, NYU Stern accounting seminar, the 2016 FARS Midyear
meeting, the 2016 AAA annual meeting, the 2016 CMU Accounting Mini-conference, Rochester, Simon School accounting
seminar, 2016 HKUST Accounting Symposium and HBS A&M seminar for their helpful comments and suggestions. We are
grateful to Claudia Imperatore, Jacky Jiang, Marcel Tuijin and Yina Yang for excellent research assistance.
1
1 Introduction
The level and structure of executive compensation has been a frequently debated topic among
politicians, CEOs, and academics since the financial crisis of 2007-2009. Critiques of
compensation practices at financial service companies often attribute the crisis at least in part to
incentive pay that purportedly encourages excessive risk taking. Regulators in Europe and the
US have proposed, and in some cases even implemented, regulation that monitors or modifies
the level and the structure of executive compensation in the financial services industry. Our
objective in this paper is to examine the consequences of regulating executive compensation
using the evidence from its first implementation in the UK and comparing the results to samples
of financial institutions in Europe and the United States.
We first study the capital market reaction to the UK Remuneration Code and the
European Union’s cap on bonuses using equity and the credit default swap (CDS) markets. The
UK was the first to implement its Remuneration Code, which required deferral of bonus
compensation initially for large banks and then for all financial institutions.1 In August 2009,
after a consultation period and having considered the recommendations of the Turner Review
(Turner [2009]) and Walker Review (Walker [2009a and 2009b], the then FSA published the UK
Remuneration Code.2 This was widely seen as a response to the financial crisis and an effort to
curtail the pay practices that allegedly contributed to the crisis. The objective of the
1 Throughout the paper, we use the terms “banks” , “BIPRU firms” and “firms in the financial services sector (or
industry)” interchangeably. All references are used to refer to firms that fall under the FSA/FCA Remuneration
Code in the UK. BIPRU refers to firms covered by chapter 12 of the Prudential Sourcebook for Banks, Building
Societies and Investment Firms. 2 On April 1, 2013, FSA was abolished and became two separate regulatory authorities: the Prudential Regulation
Authority (PRA) as part of the Bank of England and the Financial Conduct Authority (FCA). The Remuneration
Code remains in effect. FCA oversees most of the Remuneration Code for BIPRU firms; PRA jointly oversees the
implementation of the Remuneration Code together with FCA for financial institutions subject to the EU Capital
Directive IV. For consistency we continue referring to the FSA as the main regulator behind the Remunera tion Code
throughout this paper.
2
Remuneration Code was to decrease short-termism among executives by requiring them to defer
a larger portion of their bonus compensation and introducing performance vesting conditions for
these bonuses to increase pay-performance sensitivity and curb risk-taking behavior. In February
2013, the UK Remuneration Code was followed by the EU-wide bonus caps for all banks in the
European Union. In the United States, the US Security and Exchange Commission together with
the Federal Reserve and other financial sector regulators jointly issued a proposal in April of
2011 to require large and systemically important financial institutions (with total assets of more
than $50bln) to defer at least 50% of compensation as incentives-based pay as part of the Dodd-
Frank Act requirements (section 956). The agencies re-issued a revised regulation for
consultation in April 2016 with applicability to a wider set of financial institutions.
Implementation of the revised regulation is expected to take place later in 2017. The revised
proposed regulation for US financial institutions is similar in spirit to the UK’s: it requires
deferral and vesting of performance-based compensation to decrease risk-taking incentives. This
section of the Dodd-Frank Act has not been implemented yet, however.3
Interestingly, the announcement of the UK Remuneration Code was accompanied by a
positive market reaction, with BIPRU firms experiencing cumulative abnormal returns of 5.04%
in a three-day window around the announcement date. Our more extensive study of 18 event
dates shows that while the UK Remuneration Code was perceived positively by the capital
market, potentially because the restrictions were not as strict as prior expectations, EU bonus cap
proposal was perceived negatively. Our analysis indicates that BIPRU firms experienced
3 We find that the banks subject to section 956 experience positive, statistically significant cumulative abnormal
returns of 0.28% over a three-day window around the announcement of the initial regulation and positive and
statistically significant cumulative abnormal returns of 4.12% around the announcement of the updated regulation in
April 2016. The updated rules were largely unexpected, although the Obama administration foreshadowed an
increase in regulatory oversight during President Obama’s speech on March 7, 2016. The market reaction to that
event was negative and statistically significant cumulative abnormal returns of 1.08% around the date of the speech
(see “Obama Defends His Record on Financial Regulation,” The Wall Street Journal, March 7, 2016).
3
statistically significant 2.25% negative cumulative abnormal returns around the three-day
window surrounding the announcement of the European Union proposal to cap executive
bonuses. Similarly, other EU banks also experienced negative cumulative abnormal returns of -
0.26% over the same period. Our findings therefore are indicative of potential endogenous costs
of regulation.
We next study the economic consequences of the new regulation. We use the UK as our
setting for these analyses since the UK was the first to pass regulation on compensation
following the financial crisis therefore providing a longer time period to observe its effects. In
our tests, we compare UK banks to other UK firms, other banks of similar size in the European
Union and the US in order to isolate the effect of the regulation from that of the macroeconomic
environment at the time. Consistent with the mandate of the Remuneration Code, we observe that
changes in BIPRU firms’ compensation are concentrated in deferred bonuses. We find that these
firms defer more bonuses after the regulation became effective. However, we also observe that
pay-performance sensitivity of deferred bonus compensation decreases after the regulation. In
line with the intent of regulation, we also find some evidence of a decrease in measures of firm
risk in response to increased bonus deferrals by BIPRU firms post-2010. To evaluate the
concerns raised about potential unintended consequences of regulation, we examine whether
there is a change in executive turnover behavior after the regulation. We find that the likelihood
of unforced CEO turnover for BIPRU firms increases in the post-2010 period. This also holds
true when we compare BIPRU firms to US banks, suggesting that compared to US bank CEOs,
BIPRU CEOs in the UK are more likely to switch jobs following the introduction of regulation.
Finally, we find that following the introduction of the Remuneration Code, BIPRU CEOs’
4
compensation contracts become more complex in comparison to compensation contracts of
CEOs of other large UK firms.
Our paper contributes to the literature on the economic consequences of regulatory
changes in corporate governance and, in particular, executive compensation. Although there is
extensive literature on the consequences of regulation in other domains, studies of compensation
regulation have tended to focus on say-on-pay regulation (e.g., Cai and Walking [2011], Larcker,
Ormazabal, and Taylor [2011], Ferri and Maber [2013]). We complement these studies by
documenting some of the consequences of specific and mandatory constraints placed on
executive compensation contracts. However, we acknowledge that a potential limitation of our
study is that the consequences we document may not be generalizable to other firms, given the
economic and regulatory idiosyncrasies of the financial services sector. Although the regulation
we examine is unique to the financial services sector, this sector is a crucial part of the global
economy and regulatory intervention in this sector can have repercussions in the wider economy.
Despite focusing on the UK as the setting, our analyses provide an opportunity to inform the
ongoing regulation process, given international interest and effort in regulating compensation,
and our comparative results from the US and Europe.
The remainder of this paper is organized as follows. Section 2 discusses the evolution of
the Remuneration Code. Section 3 discusses our motivation and related research. Section 4
describes our data and research design. Section 5 presents our findings and section 6 concludes.
2 The UK Remuneration Code and EU Bonus Cap
Towards the end of the financial crisis, the UK government requested a review of banks’
corporate governance and the causes of the financial crisis with recommendations for changes in
5
regulation in order to improve the quality of the banking system. Lord Adair Turner, the
chairman of the FSA at the time, conducted the review of the causes of the financial crisis, while
Sir David Walker reviewed banks’ corporate governance. In addition to raising various other
concerns, both reviews recommended that compensation practices be changed.
Concurrent with the Turner and Walker reviews, the FSA issued a proposed remuneration
code on February 26, 2009. The code, which came into effect on January 1, 2010,4 requires
remuneration policies to be consistent with “effective risk management.”5 As anticipated at the
time of the adoption, the code was revised and the new version became effective on January 1,
2011. In 2010, the Remuneration Code applied to the 26 largest banks, building societies and
broker-dealers. Starting January 1, 2011, the Remuneration Code applies to all banks, building
societies, investment banks and firms covered by the Capital Adequacy Directive (UCITs, fund
managers, broker-dealers, asset management firms and some firms that engage in corporate
finance, venture capital, the provision of financial advice and stockbrokers). The FSA refers to
these firms collectively as “BIPRU” in the Remuneration Code. Approximately 2,750 firms fall
within the wider scope of the regulation.6
Not all BIPRU firms are affected in the same way as there are proportionality criteria. In
particular, the FSA defines three proportionality tiers: Tiers 1 and 2 contain credit institutions
and broker-dealers that engage in significant propriety trading and investment banking activities
and have assets in excess of £50 billion for Tier 1 and between £15 billion and £50 billion for
Tier 2. Tier 3 mainly consists of smaller banks and building societies (with total assets not
4 Chapter 19, entitled “The Remuneration Code”, was included in the Senior Management Arrangement, Systems
and Controls (SYSC) sourcebook. 5 Section 2.1 of the code states that “A firm must establish, implement, and maintain remuneration policies,
procedures and practices that are consistent with and promote effective risk management .” 6 For the full definition of the applicability of the code, see for example the FSA/FCA revised website at
https://www.fca.org.uk/firms/being-regulated/remuneration-codes.
6
exceeding £15 billion) and firms that may occasionally take overnight or short-term risks with
their balance sheet. Tier 3 also contains firms that generate income from agency business without
putting their balance sheets at risk. The proportionate approach implies that the requirements for
Tier 3 are less onerous than for Tiers 1 and 2. Firms in Tier 3, for instance, are not expected to
have a remuneration committee or apply some of the more prescriptive elements of the
remuneration rules.7
The restrictions implemented by the Remuneration Code mainly focused on bonus
compensation. In particular, at least 50% of bonuses have to be deferred for at least three years,
and have to have performance vesting conditions attached. These rules apply to executives
receiving more than 33% of total remuneration in variable pay, and whose total remuneration
exceeds £500,000. In other words, the Remuneration Code affects pay of a wider range of
employees and not just the top-tier executives.
In contrast to the Remuneration Code, in February 2013 the European Parliament, the
European Commission and the European Council proposed a new rule that capped the ratio of
variable and fixed pay at the one-to-one level, with some flexibility to increase the ratio to one-
to-two if there was a supermajority shareholder approval (“EU bonus cap”). The cap applies to
all EU banks operating in the EU (including their employees based outside of the EU) and to
non-EU banks operating in the EU. It became effective on January 1, 2014 and applied to
bonuses paid in 2014 that related to performance in 2014. All financial institutions subject to the
European Capital Directive (CRD4) are covered by the EU bonus cap rules. Furthermore, this
7 See for example, http://www.fsa.gov.uk/about/what/international/remuneration/application;
http://fsahandbook.info/FSA/html/handbook/SYSC/19A/1 and http://www.fsa.gov.uk/static/pubs/guidance/fg12-
19.pdf. Since we compare UK banks to other large UK companies and large US and EU banks, in our analysis, we
focus on Tier 1 credit institutions that were subject to the Remuneration Code starting from its first introduction in
2009 (affecting pay in 2010).
7
cap applies to a wide set of employees identified as being in a position to have “a material impact
on the institution’s risk profile.” In reality, all employees with total annual remuneration
exceeding EUR 500,000 or in senior positions are affected by these caps. The United Kingdom
was the only EU member state that voted against this proposal and appealed the European
Parliament’s decision at the European Court of Justice. However, in November 2014 the UK
government decided to drop the appeal and bonus caps came into effect in the UK starting from
January 1, 2015.8
3 Related research and motivation
Regulation might be warranted in response to market failure (Demsetz [1968], Stigler [1971],
Posner [1974]). The need for regulation arises because there is a mismatch between private
benefits and costs and social benefits and costs. In particular, even if compensation contracts
overcome the principal-agent problem and are optimally aligned with the interest of
shareholders, they might still lead to socially excessive risk-taking because they do not
internalize the costs of resolving potential bank failures. At least part of the cost of bank failures
is borne by depositors and tax payers and not executives or shareholders; therefore, there is little
incentive on the part of bank executives to take this cost fully into account. In addition,
compensation contracts may also be subject to rent-extraction in a typical principal-agent setting,
where managers “capture the board,” bid up their compensation above the optimal level, extract
surplus from shareholders, and take excessive risk (Thanassoulis [2012]). The introduction of the
Remuneration Code in the UK potentially addresses both of these concerns. Introducing
regulation is not costless, however. While regulation that addresses a market failure could
8 See for example, “Bank of England deputy governor condemns EU bonus cap,” The Guardian, October 16, 2014
and “George Osborne backs down over EU cap on bankers’ bonuses ,” The Guardian, November 20, 2014.
8
potentially lead to a socially preferable outcome, it can also have unintended consequences for
banks and the economy at large, such as loss of talent and inefficient changes to banks’ asset
portfolios (Core and Guay [2010], Murphy [2013]).
The Remuneration Code mentions the market failures it seeks to address, focusing
specifically on curbing risk-taking behavior by banks’ managers by changing the horizon of
compensation, introducing risk-adjusted incentive-based pay and reducing cash-based
compensation:
“The interests of shareholders with a focus on short-term profits are not aligned with those of long-term
shareholders. They are further not aligned with the interests of society as a whole, as they do not take into
account the wider consequences of excessive risk-taking. Shareholders with a focus on short term profits
can include employees of the firm who are participants in share incentive schemes, which often mature in a
relatively short period of time. Pressure from shareholders with short -term perspectives is one factor why
remuneration packages geared towards the short-term and leading to excessive risk-taking are offered to
employees in the banking industry.” (Financial Services Authority [2009a])
Our objective is to study the economic consequences of executive pay regulation. Prior
event studies on compensation regulation find differing results. Ferri and Maber [2013] find that
the introduction of say-on-pay regulation in the UK resulted in a positive stock price reaction
among firms with excess pay and poor performance. Cai and Walking [2011] find a positive
market reaction for firms with the highest level of excess cash compensation for the first day the
Shareholder Vote on Executive Compensation Act was introduced and mentioned in the press in
the US, whereas Larcker et al. [2011] find that the market reacts negatively on average on that
event day, but not on the day the Act was passed by the House. Out of the list of executive pay
events identified in Larcker et al. [2011], the introduction of the Excessive Pay Capped
Deduction Act and Shareholder Approval Act is the closest in nature to our setting as it specifies
mandatory requirements. On this event day, the authors find no significant average market
reaction and no significant association between the market reaction and excess pay.
9
In addition to the regulation of pay in the UK and US, the European Union also
implemented compensation regulation effective as of January 1, 2014. Thanassoulis [2012]
provides the closest analytical setting that resembles the implementation of bonus caps by the
EU. In his analytical model, competition for a limited set of bankers drives up the demand and
compensation and results in bankers’ being compensated in variable bonuses rather than fixed
wages. While he finds that the optimal banking regulation does indeed involve some limit to the
proportion of banks’ balance sheets that could be used towards paying bankers’ bonuses, his
model also predicts that stringent bonus caps are value destroying, default risk enhancing and
suboptimal for regulators overseeing a small number of banks. Similarly, Murphy [2013]
provides an ex ante economic analysis of the proposed bonus caps and concludes that bonus caps
would increase risk-taking incentives and result in a loss of value in the EU banking sector.
Ferrarini and Ungureanu [2011] examine the changes in pay practices as a result of the Principles
and Standards for Sound Compensation Practices that were accepted by the G20 countries. The
authors state that the resulting regulation of bankers’ pay is more detailed in Europe than in the
US, but remuneration practices at large US banks are converging to those of EU banks.
The recent financial crisis drew additional attention to CEO pay in the financial sector
internationally. Gregg, Jewell, and Tonks [2012] find that compensation in the financial services
industry in the UK is high. However, pay-performance sensitivity of compensation in this
industry is not higher than in other industries, which the authors interpret as being inconsistent
with the argument that excessive incentives caused the financial crisis. Using a sample of 53
systemically significant banks in the EU over the 1999-2009 period, Ayadi, Arbak, and De
Groen [2012] find that the use of long-term incentive plans increases risk-taking while option
10
plans do not, which is also inconsistent with the reasoning underlying the EU regulation to curb
option pay but encourage long-term incentive plans closely linked to performance.
Using an international sample of banks over the 2000-2008 period, Suntheim [2011]
finds that banks, whose CEOs had high risk-taking incentives, performed worse after the
collapse of Lehman Brothers, whereas banks that granted more stocks performed better. He also
finds that CEOs with high vega and low delta undertake riskier, fee-based activities and their
firms have higher leverage.
Studying the consequences of compensation regulation in the recent UK setting is
different. Although there has been regulation requiring companies to obtain shareholders’ vote
on executives’ pay both in the UK and US, this vote is non-binding. Our setting allows us to
examine the economic consequences when regulation requires compensation practices to comply
with specific rules. Since compliance with the regulation we study is mandatory, our study
differs from the previous work on say-on-pay, where firms’ compliance with shareholders’ vote
is optional. Furthermore, unlike papers such as Bebchuk and Spamann [2010] and Bhagat and
Romano [2010] that criticize existing pay practices and propose a reform in executive
compensation, or Core and Guay [2010] and Murphy [2013] that provide an ex ante analysis of
proposed compensation regulation, our paper undertakes an analysis of the consequences of
compensation regulation using as a setting its actual implementation in the UK (Remuneration
Code) and EU (bonus caps).
11
4 Data and research design
4.1 Sample selection and description
Our database collates information from a number of sources. We obtain CEO compensation data
for the UK FTSE350 firms trading on the London Stock Exchange Main Market from Thomson
Reuters IDS and Compustat Capital IQ. For a subset of firms for which data is available we also
collect credit default swap (CDS) data from Markit. Although the scope of the regulation we
study extends beyond the CEOs, we constrain our analysis to CEOs to facilitate comparability
across our test sample and control samples. IDS collect compensation and incentives data from
companies’ annual remuneration reports and the resulting data contains information on
components of executive compensation and their contractual features. The dataset spans the
period of 2006 to 2012 and contains information for 2,470 unique executives from 532 firms. We
supplement this data with market and accounting information from Thomson Reuters
Datastream, Capital IQ, and Compustat Global. Bank-specific data for UK, other EU, and US
banks is from Bankscope.
As the recent regulation of executive compensation focuses on BIPRU firms, we split our
sample into firms that fall under the FSA Remuneration Code regulation (BIPRU) and other
firms. We have 110 firm-year observations for BIPRU firms and 1,429 firm-year observations
for the non-BIPRU firms in the UK sample.9 Our final UK sample consists of 1,539 CEO-year
observations from 2006 to 2012. We remove the firm-year observations where a CEO spent less
than a year at the firm to prevent our results from being influenced by unusual circumstances
9 Remuneration Code is in effect for all BIPRU firms that are FSA regulated. For evaluating the sensitivity of our
results to our sample composition, we collected data on the BIPRU firms that do not belong to FTSE350 via Capital
IQ, where available. Not surprisingly, total compensation and its components for those BIPRU firms are lower than
the respective value for the group of FTSE350 BIPRU firms. Although inclusion of the non -FTSE350 BIPRU firms
results in reduced differences between BIPRU and non-BIPRU firms, our inferences from multivariate analyses
remain unchanged.
12
which may generate atypical compensation outcomes.10 Since regulation is typically introduced
in response to an event that is taken to indicate a market failure, it might not be considered a
truly exogenous shock. In order to address this concern, we compare UK BIPRU firms to other
similar banks in the European Union and the US.11 Using propensity score matching, we
construct a matched sample of EU and US banks using total assets, profitability (return on assets)
and bank leverage as our observable characteristics. We use a one-to-one nearest neighbor
matching without replacement. The final matched sample consists of 66 EU bank-year
observations and 116 US bank-year observations. We convert all currency-denominated amounts
into real UK pound sterling using 2012 as our index year.12
Table 1 presents descriptive statistics for the sample, where Panel A contains the
subsample of BIPRU firms, Panel B contains all other UK firms and Panels C and D present the
information for US and EU banks from the matched sample respectively. In this table, we
tabulate the main compensation components and their determinants. Panel A also shows whether
average values of tabulated variables are significantly different between the two UK subsamples
while Panel C and D show whether US and EU banks are different from their matched UK
counterparts. Detailed definitions of all variables used in our analysis are in Appendix A. To
mitigate the effects of extreme observations, all continuous variables are winsorized at the 1st and
99th percentile of their respective distributions in each year.
10
In our analysis of CEO turnover, however, we retain the partial year observations to identify instances of turnover. 11
Since US and EU banks are larger than the UK financial institutions that are required to follow the Remuneration
code in the second wave (i.e. starting January 1, 2011), we have to use the financial institutions that fall under the
first wave of the Remuneration Code due to our matched-sample design. Therefore, although we are able to compare
UK financial institutions that fall under the first wave to banks in the US and EU, we cannot implement tests based
on the staggered adoption of the Remuneration Code. 12
We match banks annually on these observable characteristics. In un tabulated robustness results, we also redo our
sample with a one-off match from the start of the sample and obtain similar results.
13
As can be seen from Panel A of Table 1, relative to non-BIPRU firms, CEOs at BIPRU
firms on average receive higher total compensation (at the time of the grant or at the time of
vesting, the latter of which is labelled as “Take Home Pay”). BIPRU firms have higher bonus
and deferred bonus, whereas non-BIPRU firms have higher salary. CEOs of BIPRU firms have
tenure of similar length to other CEOs (9 years on average for both) but they tend to oversee
larger institutions as the average size (measured as the market value of the firm) and revenues
(measured as sales or the top line of revenue) of BIPRU firms are significantly larger than those
of other firms in our sample. BIPRU firms also have higher book-to-market ratios, higher
leverage, and higher shareholder return, measured as the total return to shareholders for a one-
year holding period, starting from the beginning of the corresponding fiscal year. Descriptive
statistics tabulated in Table 1, Panel A also suggest that salary constitutes a larger proportion of
total executive compensation than cash bonus and percentage of incentives pay in total
compensation at UK banks (and UK firms in general), consistent with Conyon and Murphy’s
[2000] findings for UK companies.
Panel C of Table 1 shows the main components of CEO compensation and its
determinants for US banks in the matched sample as well as provides a comparison between UK
BIPRU and US banks. On average, US banks’ CEOs have higher total compensation but at least
during our sample period appear to have smaller take-home pay than UK BIPRU CEOs. US
banks also have higher incentives-based pay in their compensation contracts and lower salary
than UK BIPRUs. Since banks were matched on size, they do not appear to differ significantly
on their size and profitability. US banks, however, appear to have higher leverage and higher
idiosyncratic risk. The finding that total compensation is higher in the US than in the UK is
consistent with the findings of Conyon, Core, and Guay [2011], who report that UK CEOs are
14
paid less than US CEOs on an unadjusted basis. However, they also document that UK CEOs
have less stock and option incentives. When they adjust executive compensation estimates of risk
premiums, risk-adjusted pay of UK and US CEOs are no longer significantly different.
Panel D shows the summary statistics for the EU banks in the matched sample. EU banks
on average have lower total compensation and take-home compensation than UK BIPRUs, have
a higher component of pay based on salary (on average 59%) and lower proportion of incentives-
based pay.
5 Results
5.1 Event study analysis of the Remuneration Code implementation and EU
regulation to cap bankers’ pay
We first analyze the market reaction to a number of events that affected the regulation of
compensation for UK and EU banks. We are interested in studying the market reaction to these
events because a priori it is not clear whether regulation enacted to address market failure is also
perceived positively by shareholders as it addresses a misalignment of incentives between
shareholders and the management. The first event is FSA’s proposal of the Remuneration code
on February 26, 2009 in the UK. The most recent change was proposed on February 27, 2013 by
the European Union to cap bonuses in the financial sector at the level of salary, with a possibility
of increasing it to two times the base salary if the firm has the supermajority backing of its
shareholders. The bonus caps went into effect on January 1, 2014, despite a pending appeal from
the UK government. After much discussion on the direction the appeal might take, the UK
15
government finally announced on November 20, 2014 their decision to drop the appeal.13 The
full set of regulatory changes affecting EU and UK financial institutions as well as the
corresponding market reaction tests results’ are summarized in Appendix B. We present
multivariate analyses of stock returns to a subset of these events in Table 2. As in all event
studies, we jointly test whether the market revised its priors about the likelihood of regulation
going forward and shareholders assessment that the regulation is going to have an effect on
shareholders’ wealth.
Table 2, Panel A presents our findings for cumulative abnormal returns (CARs)
computed relative to the UK FTSE All-share value-weighted index around the event date using a
three-day window centered on the event date (-1,+1).14 Panel B also includes the market reaction
of comparable EU and US banks for the EU bonus cap announcement and Dodd-Frank Act
Section 956 proposal announcements. We start our analysis with event date 1, which is the initial
proposal of the Remuneration Code on February 26, 2009.15 Event 2 is the publication date of the
final version of the Remuneration Code on August 12, 2009. Event 11 is the EU decision on
February 27, 2013 to cap bonuses at the maximum of two-times the salary. Finally, event 17 is
the date when the UK dropped its appeal of the EU decision to cap executive bonuses at financial
institutions.
On average, firms in our UK sample experienced significant positive abnormal returns
(2.3%) on the day of the announcement of the Remuneration Code. However, this effect was
more positive and statistically significant for BIPRU firms (cumulative abnormal returns of
13
See for example, “George Osborne withdraws challenge to bonus cap after legal blow,” The Telegraph, November
20, 2014. 14
Our results are insensitive to the choice of the reference market index measure and constituents weights. Since our
events are aligned in time, we also conduct placebo tests and estimate bootstrapped standard errors to ensure that the
documented effects are not driven by some other effects or spurious correlation. 15
This is the first public announcement of the Remuneration Code. UK was the first of the G20 countries to
announce and implement changes to remuneration of financial sector executives.
16
5.0%). Following the final publication of the Remuneration Code, BIPRU firms on average
experienced negative abnormal returns of 1.1%. There are two potential interpretations of the
initial positive reaction to the announcement of the Remuneration Code: either that the
Remuneration Code was perceived as beneficial regulation for the shareholders or that it was
perceived as having fewer restrictions on financial institutions than originally anticipated.16 The
subsequent negative reaction reflects the final code was more restrictive than the initial proposal.
On average, firms in our sample experienced significant negative abnormal returns with
average CARs of -0.17% around the announcement of the EU bonus cap proposals (event 11).
However, non-BIPRU firms experienced on average negative and significant abnormal returns
with CARs of -0.03% while BIPRU firms saw on average negative CARs of -2.25%. Affected
BIPRU firms had average cumulative abnormal returns of -2.2%.17 We define Affected firms as
those that already paid bonuses more than two times the salary. Since this is the absolute
maximum level of bonus allowed under the EU bonus cap, this group of firms are required to
make some changes.18 Our prior was that the market reaction would be more negative for the
firms that already paid bonuses higher than the maximum and required to change their CEO
compensation right away while those paying more than one times the salary could put bonuses to
shareholders for an approval. However, our results seem to suggest that the consequences of the
bonus cap for this group of firms have not been perceived as being worse than the consequences
for all BIPRU firms. Finally, on the day when the UK dropped its appeal against the EU bonus
16
Unlike in other countries, marginal investors in the UK tend to be pension funds and other institutional investors
whose investment horizon is longer. Deferral of compensation (the main component of regulation) shifts the horizon
of executives’ compensation and potentially aligns their incentives with those of longer-horizon investors. This
could be another potential explanation for the positive market reaction. 17
We conducted an extensive news search using Factiva and Lexis-Nexis to identify whether any other material
events have occurred on any of our event dates. To the best of our knowledge, no other material events occurred on
these dates affecting UK financial institutions. 18
Please recall that the EU regulation caps bonuses at one times the salary. The cap can be increased to two times
the salary provided there is supermajority support by the shareholders.
17
cap regulation (event 17), all UK firms had negative abnormal returns with BIPRU firms
experiencing on average significant negative abnormal returns of -0.49%, consistent with
shareholders not welcoming the restriction.
In this particular setting, all events are clustered in time and apply to all UK or EU banks
in the sample. As a result, cross-sectional correlation of returns could create serious inference
problems—not addressing this will tend to overstate the significance of results (Bernard [1987],
Lo [2003]). Therefore, following Lo [2003] and Larcker et al. [2011], we compute bootstrapped
p-values for all coefficients using simulated non-event days in the sample. Inferences for all
analyses remain unchanged if p-values are computed in this manner. We also test that the results
are not driven by some spurious relationship arising on the day of a given event, by using a
random set of placebo dates and find insignificant results for these dates. Since US banks were
not directly affected by the UK and EU compensation regulation, we could potentially use their
three-day abnormal returns around the event dates in placebo tests as those returns should not
have been affected by the events in Europe. However, we found that on all but two event dates
there were confounding factors affecting US bank returns, which precludes us from using the US
banks in the placebo tests. Therefore, we only present the results for the main event dates and
note if a confounding event took place at the same time in the European or US market (see Table
2, Panel B).19
Panel C of Table 2 presents more informative results, by showing the coefficients of
regressing the event-day abnormal returns on the variables of interest and controlling for firm
characteristics. Bonus > 2 x Salary indicator (Cash Bonus > 2 x Salary indicator) identifies the
firms that already pay bonus (cash bonus) amounts that exceed two times the salary, therefore for
19
A list of confounding events for US banks is available from the authors upon request. We make sure that none of
these events are directly related to UK or EU regulation. There were no confounding events for the EU banks.
18
whom the EU proposal of bonus caps will be binding. Bonus > 2 x Salary indicator (Cash Bonus
> 2 x Salary indicator) is equal to 1 for all companies that have bonuses (cash bonuses) that are
at least two times the corresponding salaries. Although the caps on bonuses had not been
introduced yet as of events 1 and 2, we include these indicators in all models to control for
bonuses that may have been viewed as excessive. We control for firm size (Size), which is the
natural logarithm of market value, Book to market and Momentum, defined as the market
adjusted return for a given stock in the sample over the prior sixty days. Based on the results
tabulated in column (3), the negative reaction to the announcement of bonus cap by the BIPRU
firms whose cash bonus was greater than two times the salary is significantly larger with the
cumulative abnormal return of 1.98%. However, when we focus on total bonus instead of cash
bonus in column (7), there is no incremental negative reaction to the news by BIPRU firms,
consistent with investors perceiving only the cap on cash bonus to be bad news.
While the EU regulation applies to all bonuses, by the time of the announcement, UK had
already implemented the Remuneration Code and as a result banks were deferring at least half of
their bonuses. The EU regulation, would further affect the levels of bonuses available to BIPRU
CEOs. Compared to UK banks, EU banks had a smaller negative abnormal return of 0.26% and
US banks had positive abnormal returns of 2.06% on the day of the EU cap announcement
(however, in the case of the US there is a confounding event). In the untabulated results, we have
also investigated the full set of market reaction tests for EU and US banks. Affected EU banks
with cash bonus balances greater than two-times of salary experienced negative cumulative
abnormal returns of 0.86%. The overall reaction in the multivariate tests was in the order of
negative CARs of 1.05%. Furthermore, the announcement by the UK government that they
would drop the appeal against bonus cap actually lead to a further negative reaction in the UK
19
market, with negative abnormal returns for affected UK banks of 5.41% and 2.28%, where
whether a bank is affected was identified based on cash bonus or total bonus, respectively.
In panel D and E of Table 2, we investigate whether the positive reaction to event 1
comes from the fact that some bank CEOs voluntarily deferred bonuses prior to the
announcement of the Remuneration Code (Panel D) or were required to do so by their boards and
shareholders (Panel E). As can be seen from the results of Panel E, most of the positive reaction
to the regulation comes from the banks that have already required some form of mandatory
deferral of bonuses.20 Banks with mandatory deferrals of bonuses (measured as an indicator
variable taking the value of one for banks that have mandatory deferrals), had CARs of 2.97%
higher than banks that did not have mandatory deferrals, as shown in model (1). Model (2) shows
that a one percentage increase in the proportion of bonuses mandatorily deferred leads to an
increase in the market reaction by 4.7%. We also see that a one percentage increase in the
proportion of total compensation consisting of the mandatorily deferred bonus results in an
average increase in the market reaction of 6.6% at the time of the announcement of the
Remuneration Code. Overall, it appears that on average, the Remuneration Code was perceived
as good news for firms with some mandatory deferral arrangements already in place.21,22
It is interesting to note that while the announcement of the Remuneration Code generated
a positive reaction, the announcement of bonus caps by the EU and the subsequent decision by
20
We investigate whether CEOs were already deferring more than fifty percent of their bonuses in line with the
proposed regulation. We find that prior to regulation, even banks with mandatory deferrals, deferred less than what
was ultimately proposed by the regulation. Therefore, the regulatory change applied to compensation packages of all
UK BIPRU CEOs in our sample. 21
In subsequent (untabulated) tests, we find that the magnitude of the deferral in place is not associated with the
market reaction we observe on event dates after the initial announcement of the Remuneration Code. 22
In an untabulated robustness analysis, we also run the same set of market reaction tests for all event dates for the
EU and US banks using the percentage of deferrals as the interaction term. We find no significa nt association
between the interaction term and the announcement returns on these event dates for the EU and US banks.
20
the UK to accept this regulation was negative.23 This negative reaction to the announcement of
bonus cap implies that shareholders expect a loss of company value due to the pay restrictions
imposed, consistent with the view in Demsetz and Lehn [1985] that corporate governance is
chosen optimally. Therefore, constraints imposed by regulation can move companies out of
equilibrium. The observed loss of value could be due to the expectation that firms would increase
their fixed pay to comply with the regulation, therefore increasing their fixed costs, or due to a
larger proportion of total pay coming from fixed pay, generating the incentive for the executives
to take bad risks and avoid good risks, or shareholders rationally expecting that firms would
increase other components of pay not subject to the cap, thereby not effectively decreasing
excess pay, or the anticipation that the specialized talent at regulated financial services
companies would migrate to unregulated companies within the sector or to unregulated
countries.
5.2 Effects of the Remuneration Code on executive pay
Next we examine whether CEO compensation at BIPRU firms differs from that of other firms in
the UK and banks outside of the UK, and whether BIPRU firms behave differently after the
Remuneration Code has come into effect.24 Fernandes, Ferreira, Matos, and Murphy [2013] find
that UK executive compensation structure has changed from 1997 to 2006. Given these changes
may continue in the absence of the compensation regulation, our analyses adopt a difference-in-
23
One potential issue with our tests is that we cannot test or observe what happens in the period leading up to the
announcement of the Code. We use the very first date when the Remuneration Code is publicly disclosed as our first
event date as any other dates prior to it would be spurious and speculative since it was not publicly known whether
any regulation would take place. 24
Since EU bonus caps only came into effect for UK banks at the start of 2015, we only focus on the impact of the
UK Remuneration Code. Our sample stops in 2012, which preempts the announcement of the bonus caps and
therefore we do not expect banks in our sample to change their compensation packages in anticipation of this
regulation.
21
difference approach.25 We focus on several components of CEO’s compensation package as
changes in one component of compensation (bonuses) could be accompanied by other changes in
the overall compensation package of an executive. If initial compensation was agreed upon by
CEOs, boards and shareholders, changing one component of the compensation package as a
result of regulation is likely to result in re-contracting to make CEOs “whole” again. However, if
regulation is indeed addressing a board capture and rent extraction by CEOs, new regulation is
less likely to result in changes in other components of compensation.
Following the extant literature on executive compensation (see for example Conyon,
Core and Guay [2011]), we estimate a set of multivariate regressions using ordinary least squares
and controlling for the determinants of CEO pay, namely size (Log(Sales)t-1), growth
opportunities (Book to markett-1), idiosyncratic risk (Log(Idio. Risk)t-1), leverage (Leveraget-1),
firm’s overall performance (Shareholder returnt), CEO tenure (Log(Tenure)t-1), and industry
fixed effects.26 We define Sales for financial institutions as a top-level revenue number
(consisting of gross interest and other income), which is equivalent to sales for a non-financial
firm.
Table 3 presents our results for CEO’s total pay (measured as the natural logarithm of
CEO total compensation, Log CEO Total Compensation) and its main components; namely
salary (Log CEO Salary) and grants of annual incentive-based pay (Log of CEO Incentive Grants
25
One of the requirements for the difference-in-differences analysis is ensuring that control and treatment firms are
not different from each other in the pre-period sample. Compared to other UK firms, BIPRU firms have higher total
compensation, incentive grants, and deferred bonuses in the pre-regulation period. However, we do not observe any
significant difference between UK and non-UK financial institutions in our matched sample during the same period.
This should not be surprising as banks were matched to the non-UK financial institutions using propensity-score
matching in the pre-period, are from the same sector and have similar business models. 26
Note that for our bank-specific analyses we do not use industry fixed effects as we restrict our sample to financial
institutions only.
22
and Log of CEO Deferred Bonus).27 Although all our analyses, apart from bank-specific
comparisons, include industry fixed effects, we suppress the coefficients on all the industry
indicators with the exception of that on BIPRU firms, as they are the focus of the regulation the
effects of which we study in this paper. BIPRU is the indicator variable for financial firms
subject to the Remuneration Code, which takes the value of one if a financial firm falls under the
FSA Remuneration Code guidelines and is subject to the FSA regulation. These are the largest
BIPRU firms mostly consisting of universal BIPRU firms, investment BIPRU firms and
investment management companies. In addition to the industry fixed effects, the first model
under each dependent variable contains the economic determinants. The second model adds an
indicator variable, Post 2010, which takes the value of one for years 2010 to 2012, i.e. the years
in which the Remuneration Code is in effect, and its interaction with BIPRU.
Models (1) and (2) have Log CEO Total Compensation as the dependent variable. Model
(2) shows that, consistent with prior research (e.g. Conyon, Core, and Guay [2011]), total
compensation increases with size and decreases with idiosyncratic risk. Book-to-Market,
Log(Tenure), Shareholder Return and Leverage are insignificant. BIPRU firms receive higher
total pay, pre- 2010, and this comparison does not change post-2010.
Models (3) and (4) use Log CEO Salary as the dependent variable. We find in Model (4)
that CEO’s cash compensation increases with size and Book-to-Market, and decreases with
idiosyncratic risk and Leverage. Shareholder Return and Tenure are insignificant. BIPRU firms
seem to pay lower salaries on average and following the regulation their salaries continue to be
lower than non-BIPRU firms. In the post-2010 period all UK firms appear to reduce salaries on
average.
27
Total Compensation is the sum of salary, cash bonus , and cash benefits (CEO Cash Compensation), deferred
bonus and equity incentives , such as the value of options and LTIPs granted (CEO Incentives).
23
In Models (5) and (6), we use Log CEO Incentive Grants as the dependent variable. The
evidence for the control variables in Model (6) is consistent with that presented for Model (4),
with the exception of Book-to-Market and idiosyncratic risk, which become insignificant. BIPRU
firms grant more incentives on average than non-BIPRU firms and both BIPRU and non-BIPRU
firms grant lower incentive pay on average in the post-2010 period. BIPRU firms do not change
their incentives-granting behavior significantly in the post-2010 period.
Models (7) and (8) utilize Log of Deferred Bonus as the dependent variable. Since the
Remuneration Code specifically aims for more deferrals of pay, the evidence from these models
is potentially more relevant. The evidence for the control variables in Model (8) is consistent
with that presented for Model (4), with the exception of Leverage, which is insignificant. Model
(8) shows that BIPRU firms grant more deferred bonus in general, and more so in the post-2010
period, as indicated by a positive and significant interaction term between BIPRU and Post 2010.
These findings are consistent with the Remuneration Code being effectively implemented. We
also note that the lack of significant coefficients for BIPRU and Post 2010 in models (4) and (6)
suggests that the increase in deferred bonus in this time period was not accompanied by re-
contracting in the other compensation components we study.
We next compare BIPRU firms with US and EU banks in Panels B and C. These tests
enable us to assess whether the observed changes in compensation in the post-2010 period can be
attributed to regulation or the pressures in the macroeconomic environment. Some information
was available in Capital IQ on the composition of CEOs cash and non-cash bonuses, which
allowed us to compute an equivalent of a deferred bonus for US and EU banks.28 Panel B shows
28
Unlike UK financial institutions that require a proportion of bonus to be deferred, US banks’ CEO have an option
to voluntarily defer a portion of their bonus compensation by transferring designated amounts to pension accounts.
24
that total compensation and some of its components do not significantly differ between US banks
and UK BIPRUs. The only models showing significant differences are models (7) and (8), where
the dependent variable is deferred bonus. Given that compulsory bonus deferral was the subject
of the Remuneration Code, it is not surprising to find that UK BIPRU firms defer significantly
more bonuses. The deferrals also increase following the implementation of the Remuneration
Code and hence the interaction effect has a positive and statistically significant coefficient of
2.16.
Panel C compares BIPRU banks with other EU banks. Model (6) for CEO Incentive
grants shows that on average BIPRU firms grant more incentive-based pay than other EU banks.
While both EU and BIPRU banks increase their incentive grants in the post-2010 period,
following the introduction of the Remuneration Code, there is weakly statistically significant
evidence that BIPRU banks reduce their incentive grants compared to the levels issued by
comparable other EU banks. Model (8) shows that while on average BIPRU banks do not defer
more bonuses than their EU counterparts, following the introductions of the Remuneration Code,
they do indeed defer significantly more bonuses than other comparable EU banks. This is not
surprising as the Remuneration Code specifically required UK banks to implement mandatory
bonus deferrals of bonuses. Overall, our results in Table 3 show that the UK BIPRU firms
complied with the requirement to defer bonuses, and these deferrals are not accompanied by
changes in other components of pay. Comparisons with the US and EU banks suggest that the
changes in the compensation at UK BIPRU firms were due to the regulation, rather than the
macroeconomic conditions faced by financial institutions during that time period.
From our reading of US banks’ proxy statements and the data available through ISS, we do not observe mandatory
bonus deferrals that we find for UK and EU banks’ CEOs.
25
5.3 Analysis of the consequences of the Remuneration Code
Given that the Remuneration Code was aimed at changing pay practices, in particular bonus
deferrals, it is not surprising that we observe the aforementioned changes in compensation. We
next turn to the effects of the Remuneration Code on pay-performance sensitivity, risk-taking
behavior and CEO turnover.
Table 4 examines whether the Remuneration Code brought any changes to CEO’s pay-
performance sensitivity at BIPRU firms. Pay-performance sensitivity in the UK is deemed low,
similar to the evidence from the US (e.g. Conyon, Gregg, and Machin [1995], Conyon [1997],
Benito and Conyon [1999]). In panel A of Table 4, we present two columns for each of our four
dependent variables, Log of CEO Total Compensation, Log of CEO Cash Bonus, Log of CEO
Incentive Grants and Log of Deferred Bonus. Similar to prior studies, we measure performance
as total shareholder return (e.g., Crawford, Ezzell and Miles [1995], John, Mehran and Qian
[2010]). Given that one of the aims of the regulation was to change the horizon of CEO
incentives, we study one- and two-year horizons of shareholder returns in the first and second
column respectively for each of our dependent variables. The first column presents the pay-
performance sensitivity and the change in sensitivity of pay to performance post-implementation
of the Remuneration Code using a one-year horizon and the second column evaluates the pay-
performance sensitivity of executive pay to a two-year horizon. The regressions tabulated in
Columns (1) and (2) use Log CEO Total Compensation as the dependent variable. We observe
that total compensation of BIRPU firms is positively associated with the one-year performance;
however, there has not been a change subsequent to the regulation for a one-year horizon. Over a
two-year horizon, BIPRU firms show a positive and statistically significant association between
shareholder return and total compensation, which becomes stronger following the regulation.
26
Columns (3) and (4) show the results for Log Cash Bonus, referring to the cash
component of CEO bonus-pay. We do not find any evidence of CEO pay-performance sensitivity
using cash-based bonuses over a one-year horizon and positive sensitivity for BIPRU’s CEO
cash bonus compensation over a two-year horizon. There is also no significant difference
between pre- and post-periods in pay-performance sensitivity over both horizons. Columns (5)
and (6) tabulate the results for the regression where Log CEO Incentive Grants is the dependent
variable. We fail to find any evidence of pay-performance sensitivity using incentive grants for
BIPRU firms and non-BIPRU firms, pre- or post-2010 over a one-year horizon but document a
positive and statistically significant association when we consider the two-year horizon after the
regulation.
The final two columns of Panel A present the regressions where the dependent variable is
the natural logarithm of deferred bonus. While we observe that deferred bonus at BIPRU firms is
positively associated with performance, this association turns negative post-2010, suggesting
when performance is worse during this period; CEOs defer more bonuses. This finding is
consistent across one- and two-year horizons of shareholder returns. One potential interpretation
of this finding is also that this is an example of the cost of regulation: CEOs’ deferred bonus
compensation becomes less sensitive to performance.29
Panels B and C of Table 4 present the results for pay-performance sensitivity over one-
and two-year horizons for BIRPU banks compared to US and EU banks, respectively. Panel B
29
One shortcoming of our dependent variables in this table is that we measure them as values on the grant date.
Essentially, these variables do not capture performance incentives as appropriately as deltas. Unfortunately, we do
not have sufficient information about existing option and stock holding as well as grant expiration dates to compute
Core and Guay [1999 and 2002] sensitivities of CEO equity incentives to changes in stock price (delta) and
volatility (vega). IDS does not collect this information for their dataset and the coverage in Capital IQ is insufficient
to compute meaningful statistics for our UK sample. Furthermore, although the percentage of firms with live option
plans is 32 (11) in BIPRU (non-BIPRU) firms, the percentage of option grant values in total pay is only 2 percent
for both groups of firms. We hand-collect ownership data for our sample using FactSet and firms’ disclosures and
present these results as supplemental analysis.
27
shows that similar to our findings earlier with the determinants analysis, US banks pay-
performance sensitivity is not significantly different from that of UK BIPRU financial
institutions before the Remuneration Code came into effect. One significant difference is that in
the post-2010 period UK BIPRU’s cash bonus is more highly correlated with performance. We
also note that in the post-2010 period, both UK and US banks’ deferred bonus became more
sensitive to performance. Panel C shows that CEO incentive pay for both BIPRU and EU banks
is negatively associated with performance in the post-2010 period, while deferred bonus is
positively associated with performance in the same period over both one- and two-year horizons.
The lack of difference between UK BIPRU firms and the US and EU banks in terms of pay-
performance sensitivity in the post-2010 period could be due to (i) UK firms using a horizon
other than one year or two years over which they assess performance in their compensation
decisions (in line with changes in the Remuneration Code), and/or (ii) US and EU banks
potentially making other unobservable changes in their compensation design due to pressure.
In Table 5, we examine firms’ risk-taking behavior. In Panel A, where we compare UK
banks with other UK firms, we explore the effects of changes in compensation on risk and
default risk using three proxies for general risk measures: (i) idiosyncratic volatility, which
measures firm-specific risk, (ii) total volatility, which measures the overall risk exposure, (iii)
and leverage. We also use two proxies for default risk: (i) Z-Score, which measures closeness of
a firm and in particular a financial institution to its default barrier as it considers both
performance (ROA) and the share of book equity in total assets (Equity/Assets or capital ratio for
banks). For example, this measure is used in a cross-country setting by Laeven and Levine
[2009] who study the impact of corporate governance on banks’ risk-taking. This measure
represents an inverse of a probability of insolvency, where a higher z-score, implies a lower risk
28
of default and therefore greater stability. The original measure uses the standard deviation of
ROA as a scaler; however, given our smaller sample and horizon we substitute it by the total
volatility of returns. For a subsample of firms and banks, for which we have CDS spread data
from Markit, we also study changes in average 5-year CDS spreads following changes in
regulation. We use the 5-year, senior secured CDS spreads, which tend to correspond to the most
liquid credit default swap contracts. For parsimony, we also add additional control variables
when we study CDS spreads to take into account the determinants that have been documented in
prior studies. We see that the first three measures of risk are lower post-2010, which is consistent
with all firms taking less risk in this period. BIPRU firms have somewhat higher idiosyncratic
risk and total volatility than other firms in the UK post-2010 as evidenced by the positive and
significant coefficient on the interaction between BIPRU and Post 2010 in columns (2) and (4).
We see that higher percentage of bonus deferral in the post-2010 period for UK BIPRU firms,
however, is associated with a reduction of risk across these three measures, consistent with bonus
deferrals prescribed by the regulation resulting in less risk-taking or these firms having fewer
opportunities to do so in the aftermath of the financial crisis.
When we consider proxies of default risk, we find more mixed results. While in general,
UK BIPRU firms tend to have a lower risk of default (as measured by a positive and significant
z-score), the default scores are actually negative and significant after 2010. The interaction with
the deferred bonus is positive, however it is insignificant. For CDS spreads, we find that UK
BIPRU firms have higher spreads after 2010; however higher percentage of bonus deferral after
2010 appears to mitigate some of this risk and reduce CDS spreads by 12 to 18 basis points.
To assess whether the change in risk-taking behavior observed in panel A can be
attributed to regulation, we compare UK BIPRU firms with the US and EU banks. For these
29
comparisons, we use bank-specific data from Bankscope and test changes in bank-specific
measures of risk following the implementation of the Remuneration Code. The measures of
bank-specific risk are: the Leverage ratio, which is computed as Tier 1 bank regulatory capital
divided by tangible assets adjusted by derivative liabilities (Leverage Ratio); the proportion of
risk-weighted assets in total assets (RWA as a Proportion of Total Assets), which measures the
overall riskiness of banks’ assets; Tier1 capital ratio (Proportion of Tier1 Regulatory Capital in
Total Risk Weighted Assets), which measures how well-capitalized a bank is; provisions as a
share of total loans (Provisions as a Proportion of Total Loans), which measures the quality of
the loan portfolio; the proportion of riskier commercial and industrial (C&I) loans in the overall
portfolio (C&I Loans as a Proportion of Total Loans), which measures the relative riskiness of
the loan book; and the proportion of non-performing loans in the total loan portfolio (NPL as a
Proportion of Total Loans), as an additional measure of quality of current loans. We also study
z-scores and CDS spreads as additional measures of default risk. Panels B and C of Table 5
present bank-specific measures of risk for the subsample of BIPRU firms, matched to US and
EU banks, respectively. In Panel B, we find that UK BIPRU firms have significantly increased
their regulatory capital (Tier 1 ratio) in the post-2010 period compared to US banks, suggesting
they decreased their risk-taking. However, this effect is reduced by the percentage of bonuses
deferred, indicated by a negative and significant coefficient on the interaction term in the post-
2010 period. We also find that BIPRU firms increased their C&I lending after the regulation
(although this effect is not significant when taking into account bonus deferrals) but improved
their overall quality of the loan portfolios as we observer a significant decrease in the share of
non-performing loans. The z-scores for BIPRU firms are not significantly different from those of
US banks. We do observer however, that BIPRU firms that defer a higher percentage of bonuses
30
in the post-2010 period also experience a reduction in CDS spreads, indicating some reduction in
their risk of default.
Panel C shows that compared to EU banks, BIPRU firm improved the quality of their
asset portfolio as the proportion of risk-weighted asset to total assets has decreased (albeit, this
result is weakly significant). Changes in the percentage of deferred bonuses in the post-2010
period are negatively and significantly associated with the provisions as a proportion of total
loans proxy. However, there is no difference between BIPRU firms and EU banks in terms of the
association between deferred bonuses and provisions as a proportion of total loans in the post-
2010 period. We do find some statistically significant difference for non-performing loans,
suggesting that BIPRU firms see an improvement in their loan portfolio relative to their EU
counterparts. We do not observer any significant differences in our measures of default risk.
Overall, our evidence suggests that UK BIPRU firms with higher bonus deferrals post-2010 have
been taking less risk relative to other UK firms and comparisons with EU and US banks provide
mixed evidence as discussed above.
We next turn to the effects of changes in compensation on CEO turnover. The potential
loss of talent due to constrained compensation was one of the main concerns brought up during
the consultation stage for the Remuneration Code. For example, during the consultation period of
the Walker Review, more than 180 submissions of feedback were received. In one of these
submissions dated September 30, 2009, Alan Brener, the Director of Regulatory Risk
Management at Abbey National plc, stated the following:
“…Secondly, as you are aware, we are part of a major European banking group which operates
globally. Many of our senior management have international careers with Santander Group
(Santander) and some, as part of their career, will spend a few years in the UK and are effectively
“on loan” to us. Their remuneration is determined by a Remuneration Committee in Madrid and
clearly this will comply with both EU and Spanish law and regulations. The pay arrangements
will, of course, comply with the UK requirements while they are working in this country.
31
These individuals are talented and their services are in demand whether it be in the Group’s offices
in Norway or Chile or Hong Kong or New York. However, if the proposed measures governing
remuneration in the UK exceed those of the EU and/or other member states, it is highly likely that
a posting to the UK will be unattractive.
Santander is one of the few successful banking groups which has successfully avoided the
problems of the current financial crisis. This is due to the mixture of good regulation and good
management. There is a danger that inappropriate regulation of remuneration, particularly if it
goes further than that required by the EU, may turn the UK into a financial services backwater
with, in this case, Santander senior management reluctant to be posted here…”
If, as argued in the above example, changes to compensation brought on by the
Remuneration Code have constrained the ability of firms to achieve the optimal level or structure
of pay to retain talent, we would expect to see a higher likelihood of turnover in the post-2010
period. In Table 6, Panel A, we find that the likelihood of CEO turnover for BIPRU firms is
higher in the post-2010 period. However, this difference in the likelihood is not significantly
affected by the proportion of deferred bonuses. Panels B and C present the comparison with US
and EU banks, respectively. Panel B, shows that while the likelihood of turnover has decreased
in the post-2010 period for US banks; UK BIPRU CEOs are more likely to change jobs than US
bank CEOs in the post-2010 period. Panel C, however, shows, that the likelihood of turnover by
EU bank CEOs is not significantly different from that of UK BIPRUs.
We collect information on what happens to the CEOs that leave the UK BIPRU firms in
our sample and find that they retire, start their own business, or leave for jobs in the unregulated
part of the financial services sector. Since our sample includes only CEOs, we do not capture the
effect of the Remuneration Code on turnover of other executives whose compensation was
affected by the Remuneration Code and who may have more job mobility than top-level
executives.30 While the Remuneration Code has a wider application to employees who earn more
30
See for example the discussion in the popular press: Christopher Langner, “Why Bankers Are Leaving Finance for
No-Salary Tech Jobs ,” Bloomberg, available online at http://www.bloomberg.com/news/articles/2015-03-
15/bankers-embracing-zero-salary-in-tech-may-make-peers-obsolete.
32
than £500,000 per year in total compensation or participate in executive decision-making, data
limitations and UK disclosure requirements for remuneration reporting only allow us to capture
the information for CEOs or CFOs.
5.4 Additional analysis of compensation contract changes
Our final piece of analysis presents additional descriptive evidence for the structure of the
contractual changes of UK BIPRU firms after the introduction of the Remuneration Code
together with a discussion of untabulated multivariate results using contractual changes data for
the UK sample. Since we can only rely on the UK evidence for this analysis, we limit the
presentation of our results to simple univariate difference-in-difference comparisons of
individual contractual features using UK data for UK BIPRU and other FTSE350 firms.31 Table
7 presents these comparisons. Consistent with our results discussed earlier in the paper, salaries
and take-home pay for UK BIPRU are significantly higher in the pre- and post-period compared
to other UK firms; however the proportion of salary in total compensation decreases in the post-
2010 period while shares of incentive pay increase significantly. As mandated by regulation, the
percentage of deferred bonus in total compensation increases significantly for BIPRU firms.
Prior to the regulation, all BIPRU firms had a bonus scheme in place; however, it appears that
some firms have stopped paying bonuses after the Remuneration Code was put into place.32 All
BIPRU firms have a mandatory deferral policy in place in the post-2010 period in line with the
regulation, while only 72% of other large UK firms do. BIPRU firms decrease their live LTIP
31
We review financial and proxy statements for US and EU banks in our sample as well as ISS data for the US to
see if we can compare compensation packages across these samples as well. From our review of financial statements
and the ISS data for US banks, we find that the presentation of data for targets and vesting periods is different
enough that comparisons would be difficult. Furthermore, US banks defer bonus compensation entirely voluntarily
without any targets or vesting period requirements. We therefore, focus on the UK for this comparison where
presentation of compensation packages is comparable across firms. 32
We recognize that this might be an outcome of other confounding events such as the ongoing LIBOR scandal,
however.
33
schemes but not significantly differently from other UK firms. BIPRU firms increase the number
of option schemes and the shareholding requirements more than other UK firms post-2010.
An important aspect of the Remuneration Code is that deferred bonus compensation has
to be tied to future performance. Interestingly, Table 7 shows that the total number of unique
performance targets for BIPRU firms decreases on average compared to other firms and that this
decrease comes from the number of bonus targets. This is consistent with the argument that in
order to be able to evaluate performance and meet targets, the number of targets actually needs to
be manageable and consistent (Kole [1997]).
Finally, we compute Number of Contract Changes (weighted), which is defined as the
number of contractual features changing in a given component of compensation scaled by the
number of contractual features present in that component at the beginning of the year, summed
across the components with weights applied as the proportion that the corresponding
compensation component represents in total pay. This measure should capture the overall
changes (weighed by their importance in the total compensation) and therefore also provide an
indication of the change in the complexity of the compensation. We observe that prior to the
introduction of the Remuneration Code, BIPRU firms had fewer contractual changes than other
UK firms; however, this changes following the regulation. This suggests that the regulation
added a certain aspect of complexity and uncertainty to the contractual design of UK BIPRU
CEO’s compensation contracts.
In untabulated analyses, we investigate the impact of Number of Contract Changes
(weighted). Although for parsimony we do not tabulate these results, we discuss them in this
section in reference to all the models presented in the main analysis. Similar to the models
presented in Table 3, Panel A, we investigate the impact of the number of contractual changes on
34
levels of total compensation and find that although the extent of changes to the contract features
in the post-2010 period is associated with lower total compensation, these changes are not
statistically significant. However, the number of changes made to the contract features at BIPRU
firms seems to be associated with lower cash compensation. We also find that contract changes
in the post-2010 period are associated with lower CEO incentive grants and higher bonus
deferrals for BIPRU firms, which are consistent with changes being made subsequent to the
regulation.
In our analysis of pay-performance sensitivity changes observed in UK CEOs’
compensation, we observe that the sensitivity of compensation to performance at BIPRU firms
post-2010 is positively associated with the Number of Contract Changes (weighted), suggesting
that the changes implemented at these firms enhance pay-performance sensitivity.
We find that contractual changes are negatively and statistically significantly associated
with our risk measures, suggesting that changes in contracts potentially decrease BIPRU CEOs
risk-taking incentives. This finding is consistent with BIPRU firms making changes to their
contract features post-2010 to comply with the Remuneration Code, and that these changes are
affecting risk taking behavior.
Last but not least, we also find that the likelihood of executive turnover also increases
with increases in the number of contract changes in the post-2010 period for BIPRU firms. This
suggests that the Remuneration Code is associated with higher executive turnover for the
regulated firms that make changes to contractual features of compensation compared to other
large UK firms after the regulation has come into force.
35
6 Conclusion
The 2007-2009 financial crisis brought further attention to executive compensation in the
financial services industry. In the UK, the new Remuneration Code came into effect in 2010 as a
result of the allegation that misalignment of shareholders’ and executives’ incentives was at least
partly to blame for excessive risk-taking leading up to the financial crisis.
In this paper, we examine changes in compensation practices of UK financial institutions
(BIPRU firms) following changes in regulation and the resulting consequences. While the new
regulation specifically targeted bonus and their deferrals, we review all main aspects of
compensation to study whether compensation contracts overall were changed following
regulation. In order to estimate the effects of regulation, we construct three different sets of
comparison groups not directly affected by this regulatory change: other large UK firms, and US
and EU banks of similar size, profitability and business models. Comparing BIPRU firms with
EU and US banks allows us to create a group of control firms that are similar in size and their
operations, which experienced the financial crisis but were not exposed to the same regulatory
changes. This also allows us to disentangle the effects of the change due to the crisis from the
effect of the enactment of the Remuneration Code.
We begin our analysis with a series of event studies starting from the announcement of
the Remuneration Code and leading up to the introduction of the EU bonus caps, which allows
us to investigate the capital market perception of these changes. While we find that the initial
reaction to the Remuneration Code was perceived positively, the subsequent introduction of
bonus caps by the new set of EU rules was perceived negatively by the capital market investors.
We then study the changes in compensation contracts and find that following the
introduction of the Remuneration Code, BIPRU firms have significantly higher deferred bonuses
36
compared to US banks, as required by the new regulation. UK BIPRU also defer higher bonuses
than their EU counterparts and have lower other incentives-based pay. Furthermore, we find that
BIPRU firms become less risky and exhibit higher pay-performance sensitivity, in line with the
intended purpose of the regulation. However, we also find higher turnover among BIPRU firms
post-2010 compared to both other UK firms and US banks, potentially due to changes in
regulation or changes in job opportunities in the financial services sector. Finally, we also find
some evidence of increased complexity of BIPRU CEO compensation contracts following the
introduction of the Remuneration Code.
Our paper contributes to the literature on executive compensation, and in particular its
regulation. Given the potential similarities of CEO pay in the UK, EU, and US (e.g., Conyon,
Core, and Guay [2011]), and to the extent that the existing institutions and enforcement are
similar across the countries currently in the process of proposing regulation of compensation, our
findings from the UK as well as our comparative results from the US and EU should be of
international interest to the parties engaged in the compensation debate. As we have shown in
this paper, while some of the regulatory changes might have had positive consequences, there are
also potential endogenous costs to regulating executive pay. As the European Banking Authority
(EBA [2015]) recently stated, mutual funds under the scope of the Capital Requirements
Directive will also be subject to the same executive compensation bonus cap and deferral
requirements starting in 2017. Therefore, our evidence can be informative about the
consequences of regulating pay at a wider set of firms.
37
Appendix A. Variable Definitions
Variable Definition Source
% Cash bonus in total compensation Share of cash bonus in total compensation. Computed by
authors using IDS
data and Capital IQ
% Deferred bonus in total compensation Share of deferred bonus in total compensation. Computed by
authors using IDS
data and Capital IQ
% of Incentives in total compensation Share of options and LTIPs in total compensation. Computed by authors using IDS
data and Capital IQ
%LTIP in total plan Share of long-term incentive performance plans (LTIP) to total
compensation.
Computed by
authors using IDS
data and Capital IQ
% of Salary in total compensation Share of base salary in the CEO's total compensation. Computed by
authors using IDS
data and Capital IQ
% Options granted in total compensation Share of the value of options granted during the fiscal year in total compensation.
Computed by authors using IDS
data and Capital IQ
% Total bonus in total compensation Share of total bonus (cash and deferred) in total compensation. Computed by
authors using IDS
data and Capital IQ
Abnormal return Market-adjusted event day return. Computed by the
authors using
Datastream data and
Compustat Global data
Annual bonus deferral scheme Indicator variable which takes the value of 1 if a company has a deferral
policy in place for bonuses and 0 otherwise.
IDS, CIQ
Annual bonus scheme Indicator variable which takes the value of 1 if a company has an annual bonus scheme in place.
IDS, CIQ
Bonus > 2 x Salary indicator Indicator variable which takes the value of 1 for all companies that have
bonuses at least two times greater than the corresponding base salaries.
Computed by
authors using IDS
and CIQ data
Book to market Ratio of book value of assets to the sum of book value of liabilities plus market value of equity.
Datastream
C&I Loans as a proportion of total loans Ratio of commercial and industrial loans to average total loans for the
period
Bankscope
Cash bonus Cash bonus received in a given year (£thousands). IDS, CIQ
38
Variable Definition Source
Cash Bonus > 2 x Salary indicator Indicator variable which takes the value of 1 for all companies that have
bonuses equal
Computed by
authors using IDS
and CIQ data
CDS spread 5-year average annual credit default spread for senior secured debt Markit
CEO Cash compensation Cash component of total compensation computed as the sum of salary,
cash bonus and other annual cash compensation in a given year
(£thousands).
IDS, CIQ
CEO Deferred bonus Deferred portion of a bonus received in a given year. Typically subject to
performance conditions (£thousands).
IDS, CIQ
CEO Incentive grants Incentive pay computed as the sum of values of deferred bonus, options
granted and LTIPs granted.
IDS, CIQ
CEO Total compensation Total amount received by the CEO in a given year; consists of base salary, total bonus, option and LTIP incentives grants valued at the point
of grants and miscellaneous payments.
Computed by authors using IDS
and CIQ data
Deferral policy compulsory Indicator variable which takes the value of 1 if a company has a
compulsory bonus deferral policy in place.
IDS
Deferral policy voluntary Indicator variable which takes the value of 1 if a company has a voluntary bonus deferral policy in place.
IDS
Idiosyncratic risk Standard deviation of the residuals from a market model estimated daily
over the previous year (t-1).
Datastream
Leverage Ratio of book value of liabilities to market value of assets. Datastream
Leverage ratio Computed as Tier 1 regulatory capital divided by tangible assets adjusted
by derivative liabilities
Bankscope
Momentum Market adjusted return over 60 days prior to the event date Computed by the
authors using Datastream data
MV Market value at the start of the year (£thousands). Datastream,
Bankscope, Compustat Global
NPL as a proportion of total loans Ratio of non-performing loans (loans overdue by more than 90-days) to
average total loans for the period
Bankscope
Number of contract changes (weighted) Number of changes to a given compensation component weighted by the
importance of this component in total compensation and the number of other aspects of the contract that change at the same time.
Computed by the
authors using IDS data
Number of live LTIP schemes Number of live LTIP schemes. IDS
Number of live Option schemes Number of live Option schemes. IDS
Post 2010 Indicator Indicator variable which takes the value of 1 starting from the fiscal year
of 2010 to indicate changes in compensation policy.
Provisions as a proportion of Total
Loans
Provisions for non-performing loans as a proportion of average total loans
(banks only).
Bankscope
ROA Return on assets computed as the ratio of net income before extraordinary
items and average total assets.
Datastream
RWA as a proportion of total assets Ratio of risk-weighted assets to total assets (banks only). Bankscope
Salary CEO reported annual salary IDS, CIQ
Sales Sales or gross interest and other income for financial institutions
(measured in £thousands).
Datastream
39
Variable Definition Source
Shareholder return 1-year (2-year) One-year (two-year) total return to shareholders. Datastream,
Bankscope,
Compustat Global
Shareholding requirement Indicator variable which takes the value of one if a company has a shareholding requirement policy in place.
IDS, CIQ
Size Natural logarithm of market value. Datastream,
Bankscope,
Compustat Global
Tenure CEO's tenure measured as the number of year spent with a given firm. Computed by the authors using IDS,
CIQ data
Tier 1 Ratio Ratio of Tier 1 regulatory capital to risk-weighted assets (banks only). Bankscope
Total number of bonus targets per year Total number of performance targets relating to annual bonus payments. IDS
Total number of changes in a contract
per year
Total number of changes to the overall incentives contract. IDS
Total number of changes to bonus
targets
Total number of changes to the number of annual bonus performance
targets.
IDS
Total number of changes to LTIP targets Total number of changes to the number of annual LTIP performance targets.
IDS
Total number of changes to Option
targets
Total number of changes to the number of annual Options performance
targets.
IDS
Total number of LTIP targets per year Total number of performance targets relating to annual LTIP grants. IDS
Total number of Option targets per year Total number of performance targets relating to annual Option grants. IDS
Total number of unique performance
targets
Number of performance targets for each individual contract. Overlapping
targets, such as for example TSR for bonus targets and TSR for LTIP
performance targets were counted as one, indicating that an individual
needs to meet only one performance target to satisfy performance
conditions.
Computed by the
authors using IDS
Total volatility Standard deviation of returns estimated daily over the previous year (t-1). Datastream
Turnover Indicator variable which takes the value of 1 if there is a change of CEO
in a given year.
Computed by the
authors using IDS and CIQ data
UK BIPRU Indicator variable which takes the value of 1 if the company is a UK-
based financial institution subject to the FSA Remuneration Code
regulation and 0 otherwise.
FSA
UK BIPRU x Post 2010 interaction Interaction variable between the indicator variable for UK financial
institutions and post-2010 period.
Z-score Proxy for default risk measuring proximity of firm default based on the
value of its liabilities and the volatility of its returns. Computed as
ln((ROA+Leverage)/std(Returns)), where Leverage is defined as a ratio of book equity to total assets (based on Laeven and Levine [2009].
Computed by the
authors using
Datastream, CIQ, Bankscope,
Compustat Global
data
40
Appendix B. Summary of main compensation regulation events in the EU and UK affecting financial institutions
(2009-2014)
Event Date Legislative or regulatory
event Application
Country of
application
Expected
market
reaction
Multivariate regression
results:
BIPRU
Multivariate
regression results:
BIPRU x
Affected by
EU bonus
cap
EU Banks
1 2/26/2009 FSA proposes its
Remuneration Code
Potentially applicable to 40
BIPRUs UK +/-
positive,
statistically
significant
negative,
insignificant
2 8/12/2009 FSA publishes its final version
of the Remuneration Code
Appears to apply to the top 26
BIPRUs UK +/-
negative,
statistically
significant
for the wider definition of
BIPRU
positive,
insignificant
3 1/1/2010
Remuneration Code becomes
effective (retroactively applies
to all compensation granted in
2010 that relates to 2009
performance)
Appears to apply to the top 26
BIPRUs UK +/-
positive,
statistically
significant
positive,
insignificant
4 6/30/2010
EU proposes to introduce tougher regulations for
financial institutions'
employees compensation
Appears more restrictive than the UK Remuneration Code;
has wider applicability to all
firms subject to the CRD
EU - negative,
statistically
significant
negative,
insignificant
5 7/29/2010
FSA publishes revised version
of the Remuneration Code
with wider application to more than 2,500 financial
institutions
Applies to more than 2,500 financial institutions
UK -
negative
reaction for
non UK
Banks firms and positive
for UK
BIPRU firms
positive, insignificant
6 10/8/2010
CEBS introduces guidelines
that are tougher than the
Remuneration Code and require deferral of up to 60%
of variable pay
Appears more restrictive than
the UK Remuneration Code; has wider applicability
EU -
positive,
statistically significant
positive,
insignificant
7 12/10/2010 EU proposes to introduce
tougher regulations
Appears more restrictive than
the UK Remuneration Code; EU -
positive,
statistically
negative,
significant
41
Event Date Legislative or regulatory
event Application
Country of
application
Expected
market
reaction
Multivariate
regression
results: BIPRU
Multivariate
regression
results:
BIPRU x
Affected by EU bonus
cap
EU Banks
has wider applicability significant
8 12/17/2010
FSA publishes revised version
of the Remuneration Code
with wider application to more than 2,500 financial
institutions
Much wider applicability than
the 2010 version of the Code UK +/-
negative,
statistically significant
positive,
insignificant
9 1/1/2011
Revised Remuneration Code
becomes effective (applies to
compensation relating to 2010
performance)
Applies to more than 2,500
financial institutions UK +/-
positive,
statistically
significant
positive,
insignificant
10 5/15/2012 EU proposals for bonus caps Applies to all EAA financial
institutions subject to the CRD EU -
negative,
insignificant
positive,
insignificant
11 2/27/2013
EU announces the decision to
cap bonuses at 1x salary (with 2x max variable component if
approved by the supermajority
of shareholders)
Applies to all EAA financial institutions subject to the CRD
IV
EU - negative, statically
significant
positive, insignificant
negative, statistically
significant
12 9/25/2013 UK appeals the bonus cap
decision
Applies to all EAA financial
institutions subject to the CRD
IV
UK +
no
significant
reaction
no
significant
reaction
13 6/12/2013
UK Parliamentary
Commission on Banking Supervision standards
proposes stricter rules
Applies to all FSA regulated banks
UK -
negative,
statically significant
positive,
statistically significant
14 10/24/2013
FSA announcement of
proposals to implement the
UK Parliamentary
Commission recommendations
Applies to all FSA regulated
banks UK +/-
no
significant
reaction
no
significant
reaction
15 1/1/2014 Bonus caps are in effect
Applies to all EAA financial
institutions subject to the CRD
IV
EU +/-
positive,
statistically
significant
negative,
insignificant
16 3/25/2014 UK proposals for bonus clawbacks (more restrictive
than EU)
Applies to all FSA regulated banks
UK - positive,
statistically
significant
negative, significant
17 11/20/2014 UK drops appeal against
bonus caps
Applies to all EAA financial
institutions subject to the CRD UK -
negative,
statistically
negative,
statistically
42
Event Date Legislative or regulatory
event Application
Country of
application
Expected
market
reaction
Multivariate
regression
results: BIPRU
Multivariate
regression
results:
BIPRU x
Affected by EU bonus
cap
EU Banks
IV significant significant
18 03/02/2011 SEC publishes proposed rules for Dodd-Frank Section 956
implementation
Applies to large and
systemically important
financial institutions in the United States (joined proposal
by the SEC, Federal Reserve
Board, US Department of the
Treasury, FDIC, NCUA and
FHFA)
US +/- insignificant insignificant negative,
statistically
significant
19 04/21/2016
Regulatory agencies publish
new proposed rules for Dodd-
Frank Section 956
implementation
Applies to large and
systemically important financial institutions in the
United States (joined proposal
by the SEC, Federal Reserve
Board, US Department of the
Treasury, FDIC, NCUA and FHFA). Replaces previous
proposal from 2011
US +/- insignificant insignificant insignificant
43
References
AYADI, R., E. ARBAK, and W. P. DE GROEN. “Executive Compensation and Risk Taking in European Banking” in Research Handbook on International Banking and Governance, edited by J. R. Barth, C. Lin, and C. G. Wihlborg. Cheltenham, UK: Edward Elgar Publishing, 2012.
BEBCHUK, L., and H. SPAMANN. “Regulating Bankers’ Pay.” Georgetown Law Journal 98
(2010): 247-287.
BENITO, A., and M. CONYON. “The Governance of Directors’ Pay: Evidence from UK Companies.” Journal of Management and Governance 3 (1999): 117–136.
BERNARD, V. “Cross-Sectional Dependence and Problems in Inference in Market-Based
Accounting Research.” Journal of Accounting Research 25 (1987): 1-48.
BHAGAT, S., and R. ROMANO. “Reforming Executive Compensation: Simplicity, Transparency and Committing to the Long-term.” European Company and Financial Law
Review 7 (2010): 273-296.
CAI, J., and R. WALKING. “Shareholders’ Say on Pay: Does It Create Value?” Journal of Financial and Quantitative Analysis 46 (2011): 299-339.
CONYON, M. “Corporate Governance and Executive Compensation.” International Journal of
Industrial Organisation 15 (1997): 493–509.
CONYON, M., J. CORE, and W. GUAY. “Are U.S. CEOs Paid More Than U.K. CEOs? Inferences from Risk-adjusted Pay.” Review of Financial Studies 24 (2011): 403-438.
CONYON, M., P. GREGG, and S. MACHIN. “Taking Care of Business: Executive
Compensation in the United Kingdom.” Economic Journal 105 (1995): 704–714.
CONYON, M., and K. MURPHY. “The Prince and The Pauper? CEO Pay in the United States and United Kingdom.” Economic Journal 110 (2000): 640–671.
CORE, J., and W. GUAY. “The Use of Equity Grants to Manage Optimal Equity Incentive Levels.” Journal of Accounting and Economics 28 (1999): 151-184.
CORE, J., and W. GUAY. “Estimating the Value of Employee Stock Option Portfolios and Their
Sensitivities to Price and Volatility.” Journal of Accounting Research 40 (2002): 613-630.
CORE, J., and W. GUAY. “Is There a Case for Regulating Pay in the Financial Services Industry?” in After the Crash: The Future of Finance, edited by Y. Fuchita, R. J. Herring, and R.
E. Litan. Washington, DC: Brookings Institution Press, 2010; 115-140.
CRAWFORD, A., J. EZZELL, and J. MILES. “Bank CEO Pay-Performance Relations and the Effects of Deregulation.” Journal of Business 68 (1995): 231-256.
44
DEMSETZ, H. “Why Regulate Utilities?” The Journal of Law and Economics 11 (1968): 55-65.
DEMSETZ, H., and K. LEHN. “The Structure of Corporate Ownership: Causes and Consequences.” Journal of Political Economy 93 (1985): 1155-1177.
EUROPEAN BANKING AUTHORITY. “Guidelines on sound remuneration policies under Articles 74(3) and 75(2) of Directive 2013/36/EU and disclosures under Article 450 of Regulation (EU) No 575/2013”, December 21, 2015.
FERNANDES, N., M. A. FERREIRA, P. MATOS, and K. J. MURPHY. “Are US CEOs Paid
More? New International Evidence.” Review of Financial Studies 26 (2013): 323-367.
FERRARINI, G., and M. UNGUREANU. “Bankers’ Pay after the 2008 Crisis: Regulatory Reforms in the US and the EU.” Zeitschrift fur Bankrecht und Bankwirtschaft 6 (2011): 418-430.
FERRI, F., and D. MABER. “Say on Pay Votes and CEO Compensation: Evidence from the
UK.” Review of Finance 17 (2013): 527-563.
FINANCIAL SERVICES AUTHORITY. “Reforming Remuneration Practices in Financial Services.” Consultation Paper 09/10, March, 2009a.
FINANCIAL SERVICES AUTHORITY. “Reforming Remuneration Practices in Financial
Services: Feedback on CP09/10 and Final Rules.” Policy Statement 09/15, August, 2009b.
FINANCIAL SERVICES AUTHORITY. “Revising the Remuneration Code: Feedback on CP10/19 and Final Rules.” Policy Statement 10/20, December, 2010.
GREGG, P., S. JEWELL, and I. TONKS. “Executive Pay and Performance: Did Bankers’
Bonuses Cause the Crisis?” International Review of Finance 12 (2012): 89-122.
JOHN, K., H. MEHRAN, and Y. QIAN. “Outside Monitoring and CEO Compensation in the Banking Industry.” Journal of Corporate Finance 16 (2010): 383-399.
KOLE, S. “The Complexity of Compensation Contracts.” Journal of Financial Economics 43 (1997): 79-104.
LARCKER, D., G. ORMAZABAL, and D. TAYLOR. “The Market Reaction to Corporate Governance Regulations.” Journal of Financial Economics 101 (2011): 431-448.
LAEVEN, L., and R. LEVINE. “Bank Governance, Regulation and Risk Taking.” Journal of Financial Economics 93 (2009): 259-275.
LO, K. “Economic Consequences of Regulated Changes in Disclosure: The Case of Executive
Compensation.” Journal of Accounting and Economics 35 (2003): 285-314.
45
MURPHY, K. “Regulating Banking Bonuses in the European Union: A Case Study in Unintended Consequences.” European Financial Management 19 (2013): 631-657.
POSNER, R. “Theories of Economic Regulation.” The Bell Journal of Economics and
Management Science 5 (1974): 335-358.
STIGLER, G. “The Theory of Economic Regulation.” The Bell Journal of Economics and Management Science 2 (1971): 3-21.
SUNTHEIM, F. “Managerial Compensation in the Financial Service Industry,” Working Paper,
Bocconi University, 2011. Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1592163.
THANASSOULIS, J. “The Case for Intervening in Bankers’ Pay.” The Journal of Finance 67
(2012): 849-895.
TURNER, A. “A Regulatory Response to the Global Banking Crisis.” Financial Services Authority, March, 2009.
WALKER, D. “A Review of Corporate Governance in UK Banks and Other Financial Industry
Entities.” Financial Services Authority, July, 2009a.
WALKER, D. “A Review of Corporate Governance in UK Banks and Other Financial Industry Entities: Final Recommendations.” Financial Services Authority, November, 2009b.
46
Table 1: Descriptive statistics
This table presents the descriptive statistics for CEO compensation and its determinants for the BIPRU firms subject to the
Remuneration Code (BIPRU) and the largest UK companies (constituents of FTSE350, excluding BIPRU), and matched UK and
US financial institutions. All values are in real, 2012 thousands of pound sterling (£thousands) unless otherwise indicated. Panel
A presents the summary statistics data for UK BIPRU firms and Panel B shows summary statistics for other UK institutions.
Panel A also shows whether means of the variables are significantly different from the corresponding variables of Panel B using the difference in means test. Panel C and Panel D show the descriptive statistics for the matched sample of US and EU banks.
Similar to Panel A, both panels also show whether these banks are significantly different from each other using a difference in
means test. To mitigate the effects of extreme observations all continuous variables are winsorized at the 1% and 99% tails of
their respective distributions in each sample year. All variables are defined in Appendix A. ***, **, * designate significance at
1%, 5% and 10% levels respectively.
Panel A: UK BIPRU firms
Variable N Mean Std dev Min Median Max
Total compensation 110 3,949.58*** 4,151 49.95 3,086 36,530
Take home pay 110 2,078.99*** 1,610 95.63 1,607 8,650
% Salary in total comp. 110 0.39*** 0.32 0.01 0.34 1.00
% Cash bonus in total comp. 110 0.21 0.21 0.00 0.18 0.92
% Incentives pay in total comp. 110 0.04 0.14 0.00 0.00 0.78
% Deferred bonus in total comp. 110 0.23*** 0.24 0.00 0.17 0.92
Tenure 110 8.87 7 0 7 29
Revenues 110 13,000,000*** 21,600,000 0 774,898 69,300,000
MV 110 6,250,984*** 10,700,000 47,766 1,283,012 43,700,000
Book to Market 110 0.94*** 1.01 0.00 0.92 8.64
Leverage 110 0.61*** 0.33 0.00 0.68 0.99
Shareholder Return 110 0.01* 0.54 -0.98 -0.03 1.78
Idiosyncratic Risk 110 0.32 0.20 0.09 0.26 1.33
Panel B: Other UK firms
Variable N Mean Std dev Min Median Max
Total compensation 1,429 2,101 2,252 85.85 1,376 23,578
Take home pay 1,429 1,178 844 0.00 906 5,860
% Salary in total comp. 1,429 0.54 0.21 0.01 0.51 1.46
% Cash bonus in total comp. 1,429 0.22 0.17 0.00 0.18 1.03
% Incentives pay in total comp. 1,429 0.04 0.13 0.00 0.00 0.93
% Deferred bonus in total comp. 1,429 0.10 0.27 0.00 0.00 6.13
Tenure 1,429 9 7 0 7 38
Revenues 1,429 3,750,737 10,200,000 0 869,303 218,000,000
MV 1,429 3,333,205 6,821,322 4,416 975,277 46,700,000
Book to Market 1,429 0.69 0.27 0.18 0.67 1.44
Leverage 1,429 0.41 0.22 0.03 0.39 0.98
Shareholder Return 1,429 0.10 0.60 -0.98 0.00 2.73
Idiosyncratic Risk 1,429 0.30 0.14 0.09 0.27 1.46
47
Panel C: US banks (matched)
Variable N Mean Std dev Min Median Max
Total compensation 116 5,350.06* 6,309 0.00 1,821 23,578
Take home pay 116 1,229.31*** 1,327 0.00 666 5,791
% Salary in total comp. 116 0.31* 0.30 0.01 0.21 1.00
% Total bonus in total comp. 116 0.18*** 0.22 0.00 0.08 0.91
% Incentives pay in total comp. 116 0.23*** 0.29 0.00 0.05 1.00
Tenure 116 7.21* 8 0 4 34
Revenues 116 9,796,481 15,800,000 5,379 1,413,185 53,600,000
MV 116 5,914,419 7,955,247 647 565,121 24,600,000
Book to Market 116 0.97 0.11 0.41 1.00 1.24
Leverage Ratio 116 0.85*** 0.17 0.08 0.90 1.00
Shareholder Return 116 0.03 0.61 -0.98 0.00 1.98
Idiosyncratic Risk 116 0.46*** 0.33 0.10 0.32 1.70
Panel D: EU banks (matched)
Variable N Mean Std dev Min Median Max
Total compensation 66 1,802*** 2,049 0 1,104 9,177
Take home pay 66 1,505.89*** 1,536 38 1,073 6,993
% Salary in total comp. 66 0.59*** 0.28 0.06 0.58 1.00
% Total bonus in total comp. 66 0.3*** 0.24 0.00 0.26 0.94
% Incentives pay in total comp. 15 0.04 0.07 0.00 0.00 0.19
Tenure 66 7.23 7 1 5 30
Revenues 66 21,300,000** 25,100,000 43,050 5,335,621 66,600,000
MV 66 10,600,000** 12,700,000 47,812 6,524,534 45,400,000
Book to Market 66 0.99 0.15 0.60 1.00 1.42
Leverage Ratio 66 0.78*** 0.32 0.04 0.94 0.99
Shareholder Return 66 -0.06 0.45 -0.84 -0.06 1.60
Idiosyncratic Risk 66 0.29 0.15 0.12 0.25 1.03
48
Table 2: Market reaction tests to UK and EU compensation regulation
This table presents the results from estimating the market reaction to the 17 regulatory events concerning executive compensation
for financial institutions in the UK and the EU and events 18 and 19 showing the reaction to the announcement of the proposal of
the implementation of section 956 of the Dodd-Frank Act by the US regulators. Details of each event date as well as a summary
of multivariate results for each could be found in Appendix B. Panel A presents cumulative abnormal returns, in which the
abnormal return is computed relative to the UK FTSE All-share value-weighted market index (UK sample). Panel B also includes the market reaction of EU and US banks relative to the value-weighted MSCI Europe excluding UK index (EU sample) and
value-weighted MSCI US index (US sample). The results are insensitive to the choice of the reference market index measure as
well as to the usage of a global index. EU and US banks market reaction is estimated using corresponding market indices . Panel
C presents results from a regression of cumulative abnormal returns surrounding the main regulatory events on the variables of
interest and control variables. Panel D presents results from a regression of cumulative abnormal returns surrounding event 1 controlling for voluntary deferral, its percentage and amount. Panel E presents results from a regression of cumulative abnormal
returns surrounding event 1 controlling for compulsory deferral, its percentage and amount. Event 1 is the proposal of the
Remuneration Code by the FSA, Event 2 is the publication of the Remuneration Code, Event 11 is the proposal of the EU bonus
caps and Event 17 is the date when the UK government dropped the appeal against EU bonus caps. Total Bonus refers to the
aggregate bonus a CEO received in the corresponding year with the indicator variable capturing whether this amount was greater than two-times the corresponding base salary. Cash bonus refers to the cash proportion of the total bonus received by the CEO
compared to the base salary. Affected is equal to one when either Bonus is greater than two times the salary or Cash Bonus is
greater than two times the salary. Size is the natural logarithm of market value, Book to market is the ratio of book value to
market value and Momentum is the market adjusted return for a given stock in the sample over the prior sixty days. All variables
are defined in Appendix A. The events are summarized in Appendix B. Values of t-statistics (reported in parentheses) are computed based on robust standard errors. ***, **, * designate significance at 1%, 5% and 10% levels respectively.
Panel A: Cumulative abnormal returns (-1,+1) for UK regulatory events
Events Date Legislative or regulatory event All UK BIPRU All UK without BIPRU
1 26/02/2009 FSA proposes its Remuneration Code 0.0230*** 0.0504*** 0.0211***
2 12/08/2009 FSA publishes its final version of the Remuneration Code
0.0006 -0.011*** 0.0011
3 01/01/2010
Remuneration Code becomes effective
(retroactively applies to all compensation granted in 2010 that relates to 2009 performance)
0.0039*** 0.0138*** 0.0031***
4 30/06/2010
EU proposes to introduce tougher
regulations for financial institutions' employees compensation
-0.0155 -0.0041 -0.0167
5 29/07/2010
FSA publishes revised version of the Remuneration Code with wider
application to more than 2,500 financial institutions
-0.0216* 0.0096*** -0.0243*
6 08/10/2010
CEBS introduces guidelines that are
tougher than the Remuneration Code and require deferral of up to 60% of variable pay
0.0009 0.0083*** 0.0000
7 10/12/2010 EU proposes to introduce tougher
regulations 0.0036*** -0.0051 0.0037***
8 17/12/2010
FSA publishes revised version of the Remuneration Code with wider application to more than 2,500
financial institutions
0.0037*** -0.0092*** 0.0043***
9 01/01/2011 Revised Remuneration Code becomes effective (applies to compensation relating to 2010 performance)
0.0023*** 0.0037*** 0.0025***
10 15/05/2012 EU proposals for bonus caps 0.0008 -0.0009 0.0013*
11 27/02/2013
EU announces the decision to cap bonuses at 1x salary (with 2x max
variable component if approved by the supermajority of shareholders)
-0.0017*** -0.0225*** -0.0003
12 25/09/2013 UK appeals the bonus cap decision 0.0001 -0.0009 0.0000
13 12/06/2013
UK Parliamentary Commission on
Banking Supervision standards proposes stricter rules
-0.0046*** -0.0181*** -0.0034***
14 24/10/2013 FSA announcement of proposals to implement the UK Parliamentary
Commission recommendations
-0.0032*** -0.0006 -0.0032***
49
15 01/01/2014 Bonus caps are in effect 0.0012** 0.0096*** 0.0006
16 25/03/2014 UK proposals for bonus clawbacks
(more restrictive than EU) 0.009 -0.0008 0.0096
17 20/11/2014 UK drops appeal against bonus caps -0.0063*** -0.0049** -0.0060***
Observations 1,602 91 1,469
50
Panel B: Cumulative abnormal returns (-1,+1) for EU Bonus Caps and Dodd-Frank Act Section 956 rules implementation
Events Date Legislative or regulatory event All UK BIPRU BIPRU Affected by EU
Bonus Cap
EU Banks (without UK
BIPRU) US Banks
11 27/02/2013
EU announces the decision to cap bonuses at 1x salary (with 2x max variable component if approved by the supermajority
of shareholders)
-0.0017*** -0.0225*** -0.0220*** -0.0026*** 0.0123***
18 02/03/2011 SEC publishes proposed rules for Dodd-Frank Section 956
implementation
0.0009 -0.0018*** -0.0032*** 0.0025**
19 4/21/2016 Regulatory agencies publish new proposed rules for Dodd-Frank
Section 956 implementation
0.0509 0.0549 0.0412***
Observations 1,602 91 15 1,309 4,542
51
Panel C: Cross-sectional variation on major regulatory event dates (UK BIPRU vs. UK firms)
Affected = 1 if Affected = 1 if
Cash Bonus > 2 x salary Total Bonus > 2 x salary
Event 1 2 11 17 1 2 11 17
(1) (2) (3) (4) (5) (6) (7) (8)
Intercept 0.0082 0.0039 0.0116* -0.0211*** 0.0097 0.0024 0.0083 -0.0243***
(0.501) (0.497) (1.755) (-3.236) (0.595) (0.297) (1.259) (-3.517)
UK BIRPU 0.0123* -0.00489* -0.0046 0.0002 0.0051 -0.00636* -0.0100 -0.0011
(1.678) (-1.838) (-0.7687) (0.066) (0.5903) (-1.8817) (-1.1834) (-0.362)
Affected (Bonus > 2 x Salary) -0.0093 -0.0044 -0.0122*** -0.0030 -0.0018 0.0003 -0.0023 0.0017
(-1.228) (1.247) (-4.604) (-1.222) (-0.437) (0.129) (-1.5541) (1.0000)
UK BIPRU x Affected (Bonus > 2 x Salary) 0.0285** -0.0180* -0.0198* -0.0541*** 0.0250** -0.0086 -0.0015 -0.0228***
(2.375) (-1.852) (2.336) (-5.893) (2.174) (3.943) (-0.1398) (-2.9495)
Size 0.0004 -0.0001 -0.000911** 0.00112** 0.0003 0.0000 -0.0007 0.00131***
(0.388) (-0.152) (-1.998) (2.534) (0.304) (0.033) (-1.4869) (2.741)
Book to market 0.00251*** 0.0000 0.0000 0.0001 0.00248*** 0.0000 0.0000 -0.0001
(2.585) (-0.148) (0.356) (-0.317) (2.611) (-0.269) (0.1572) (-0.409)
Momentum -7.508*** -3.685*** -2.635*** -3.436*** -7.526*** -3.687*** -2.711*** -3.625***
(-10.71) (-29.959) (-6.604) (-7.502) (-10.767) (-29.496) (-6.7775) (-7.88)
Observations 1,512 1,518 1,553 1,571 1,512 1,518 1,553 1,571
R-squared 0.225 0.278 0.064 0.081 0.225 0.274 0.048 0.066
52
Panel D: Cross-sectional variation of the reaction to event 1 based on voluntary bonus deferrals already in place
(UK BIPRU vs. UK Firms)
(1) (2) (3)
Intercept 0.0182 0.0004 0.0197
(0.958) (0.023) (1.015)
UK BIPRU 0.0195*** 0.0192*** 0.0197***
(2.786) (3.2) (2.814)
Voluntary bonus deferral -0.0174***
(-2.9)
UK BIPRU x Voluntary deferral -0.006
(0.415)
% Voluntary bonus deferral -0.034
(-1.489)
UK BIPRU x % Voluntary deferral 0.031
(0.962)
Amount of voluntary deferral
-0.000129*
(-1.675)
UK BIPRU x Amount of voluntary deferral
0.000
(-0.017)
Size 0.000 0.001 0.000
(-0.231) (0.029) (-3.415)
Book to market 0.00268*** 0.00249*** 0.00274***
(3.109) (2.621) (3.171)
Momentum -8.196*** -7.811*** -8.171***
(-13.706) (-11.353) (-13.618)
Observations 1,512 1,512 1,512
R-squared 0.269 0.256 0.267
53
Panel E: Cross-sectional variation of the reaction to event 1 based on mandatory bonus deferrals already in place
(UK BIPRU vs. UK Firms)
(1) (2) (3)
Intercept -0.0022 -0.0025 0.0181
(-0.123) (-0.14) (0.938)
UK BIPRU 0.0158** 0.0171*** 0.0160**
(2.3795) (2.668) (2.292)
Mandatory deferral (0.001)
(0.234)
UK BIPRU x Mandatory deferral 0.0297**
(2.106)
% Mandatory bonus deferral 0.0084
(0.532)
UK BIPRU x % Mandatory deferral 0.0470**
(2.166)
Amount of mandatory deferral 0.0000
(-0.396)
UK BIPRU x Amount of mandatory deferral 0.0660***
(3.385)
Size 0.0011 0.0011 -0.0003
(0.883) (0.892) (-0.258)
Book to market 0.00251*** 0.00252*** 0.00277***
(2.639) (2.647) (3.199)
Momentum -7.805*** -7.825*** -8.174***
(-11.378) (-11.357) (-13.692)
Observations 1,512 1,512 1,512
R-squared 0.256 0.256 0.267
54
Table 3: Determinants of CEO compensation
This table presents the results of OLS regression estimation of the determinants of CEO compensation for CEOs of the largest UK BIPRU and non-BIPRU firms in Panel A. Panel
B and Panel C present the same estimation using the propensity -score matched samples of US and EU banks respectively. Log CEO Total Compensation is the natural logarithm of
CEOs’ total compensation computed as the sum of basic salary, total bonus, other benefits, options granted valued using the Black Scholes method and LTIPs valued as options,
restricted stock or cash depending on their respective category. Log CEO Salary is the natural logarithm of CEOs’ base salary. Log CEO Incentives Grants is the natural logarithm
of CEOs’ incentive pay computed as the sum of values of deferred bonus, options granted and LTIPs granted. Log CEO Deferred Bonus is the natural logarithm of CEO’s total deferred bonus in a given year. UK BIPRU is equal to one for BIPRU firms subject to the FSA Remuneration Code. Post 2010 takes the value of 1 for years starting from 2010. All
other variables are defined in Appendix A. The sample period is 2006-2012. To mitigate the effects of extreme observations all continuous variables are winsorized at the 1% and
99% tails of their respective distributions in each sample year. Values of t-statistics (reported in parentheses) are computed based on robust standard errors clustered at the industry
level (Panel A) and robust standard errors (Panels B and C). ***, **, * designate significance at 1%, 5% and 10% levels respectively.
Panel A: UK BIPRU vs. UK Firms
Log CEO Total Compensation Log CEO Salary Log CEO Incentive Grants Log CEO Deferred Bonus
(1) (2) (3) (4) (5) (6) (7) (8)
UK BIPRU 0.52*** 0.49*** -0.15*** -0.16*** 0.40*** 0.42*** 0.91*** 0.23*** (20.811) (11.226) (-18.849) (-15.517) (7.014) (6.656) (11.236) (3.477)
Post 2010 indicator -0.12 -0.14*** -0.32** 0.37
(-1.436) (-5.008) (-2.230) (1.588)
UK BIPRU x Post 2010 interaction 0.09 0.03 -0.04 1.64***
(1.048) (1.574) (-0.298) (10.867)
Log(Sales)t-1 0.26*** 0.28*** 0.19*** 0.21*** 0.17** 0.21** 0.40*** 0.34***
(13.129) (11.467) (11.169) (11.097) (2.546) (2.833) (6.702) (3.971)
Book to market t-1 -0.04 -0.03 0.08** 0.10*** 0.48 0.54 0.65** 0.52**
(-0.206) (-0.125) (2.426) (3.747) (1.381) (1.697) (2.786) (2.368)
Log(Idio. Risk)t-1 -0.30** -0.32** -0.09 -0.11* -0.05 -0.11 -1.11*** -1.03***
(-2.640) (-2.890) (-1.548) (-2.057) (-0.414) (-0.903) (-4.426) (-3.541)
Log(Tenure)t-1 0.11 0.12 0.01 0.01 0.05 0.07 0.10 0.07
(1.638) (1.756) (0.198) (0.543) (0.629) (0.864) (0.893) (0.606)
Leveraget-1 0.20 0.15 -0.14*** -0.21*** -0.50 -0.67** -0.19 0.14
(1.709) (1.241) (-3.078) (-4.419) (-1.676) (-2.172) (-0.590) (0.351)
Shareholder returnt 0.01 0.01 -0.01 -0.01 -0.03 -0.03 0.07 0.06
(0.183) (0.171) (-0.547) (-0.667) (-0.386) (-0.411) (0.557) (0.510)
Industry Indicators Yes Yes Yes Yes Yes Yes Yes Yes
Observations 1,539 1,539 1,539 1,539 1,539 1,539 1,539 1,539
Adjusted R-squared 0.167 0.167 0.118 0.118 0.055 0.059 0.180 0.192
55
Panel B: UK BIPRU vs. US Banks
Log CEO Total Compensation Log CEO Salary Log CEO Incentive Grants Log CEO Deferred Bonus
(1) (2) (3) (4) (5) (6) (7) (8)
UK BIPRU 0.18 0.10 -0.09 -0.10 -0.43 -0.35 0.83** -0.09
(1.046) (0.481) (-1.555) (-1.156) (-1.603) (-0.866) (1.997) (-0.167)
Post 2010 indicator -0.32 -0.08 -0.01 0.02 (-1.535) (-0.664) (-0.020) (0.053)
UK BIPRU x Post 2010 interaction 0.16 0.01 -0.19 2.16***
(0.442) (0.073) (-0.311) (2.882)
Log(Sales)t-1 0.28*** 0.30*** 0.19*** 0.20*** 0.10 0.11 0.19** 0.12
(6.805) (7.064) (7.508) (7.975) (1.141) (1.143) (2.032) (1.274) Book to market t-1 0.08 0.08 0.08** 0.09*** 0.47*** 0.48*** 0.76*** 0.68***
(1.121) (1.248) (2.557) (2.711) (2.635) (2.709) (5.740) (4.119)
Log(Idio. Risk)t-1 0.01 0.01 -0.08 -0.08 1.08*** 1.09*** -0.35 -0.40
(0.074) (0.155) (-1.142) (-1.112) (4.781) (4.777) (-1.322) (-1.483)
Log(Tenure)t-1 0.12 0.14 -0.04 -0.03 0.27 0.27 0.11 0.06 (1.313) (1.571) (-0.593) (-0.482) (1.376) (1.380) (0.547) (0.274)
Leveraget-1 0.13 0.08 0.03 0.01 0.36 0.32 0.67 1.16
(0.449) (0.271) (0.252) (0.104) (0.544) (0.484) (0.915) (1.548)
Shareholder returnt 0.31* 0.31** 0.10 0.11 -0.29 -0.29 -0.54* -0.57*
(1.946) (1.999) (1.381) (1.407) (-0.896) (-0.887) (-1.778) (-1.801) Intercept 3.68*** 3.55*** 3.54*** 3.49*** 0.52 0.46 -1.87* -1.17
(8.073) (7.697) (11.761) (11.622) (0.496) (0.427) (-1.669) (-1.060)
Observations 307 307 307 307 307 307 287 287
Adjusted R-squared 0.170 0.171 0.269 0.266 0.133 0.127 0.076 0.112
56
Panel C: UK BIPRU vs. EU Banks
Log CEO Total Compensation Log CEO Salary Log CEO Incentive Grants Log CEO Deferred Bonus
(1) (2) (3) (4) (5) (6) (7) (8)
UK BIPRU 0.41* 0.33 -0.06 -0.15 0.38 0.81** 0.13 -0.69
(1.943) (1.410) (-0.722) (-1.202) (1.271) (2.329) (0.282) (-1.201)
Post 2010 indicator -0.34 -0.32* 0.75** 0.42 (-1.355) (-1.832) (1.979) (0.678)
UK BIPRU x Post 2010 interaction 0.16 0.18 -0.96* 1.86**
(0.414) (0.876) (-1.905) (2.265)
Log(Sales)t-1 0.24*** 0.27*** 0.22*** 0.25*** 0.05 0.01 0.25** 0.13
(4.536) (5.022) (6.696) (7.905) (0.912) (0.193) (2.263) (1.004) Book to market t-1 0.21*** 0.23*** 0.13** 0.14** 0.78*** 0.77*** 0.65*** 0.54***
(3.247) (3.433) (2.550) (2.550) (4.223) (4.087) (4.698) (3.607)
Log(Idio. Risk)t-1 -0.48*** -0.49*** -0.11 -0.12 -0.19 -0.16 0.40 0.40
(-4.028) (-3.930) (-1.528) (-1.614) (-0.971) (-0.803) (1.337) (1.362)
Log(Tenure)t-1 0.05 0.05 0.04 0.05 -0.06 -0.06 0.20 0.15 (0.481) (0.504) (0.797) (0.842) (-0.382) (-0.334) (0.857) (0.695)
Leveraget-1 -0.15 -0.25 -0.08 -0.17 -0.36 -0.28 1.23* 1.91**
(-0.420) (-0.700) (-0.324) (-0.687) (-0.808) (-0.642) (1.677) (2.501)
Shareholder returnt 0.03 0.05 -0.20* -0.18* -0.12 -0.15 -0.35 -0.41
(0.116) (0.194) (-1.818) (-1.668) (-0.410) (-0.523) (-0.845) (-0.928) Intercept 3.55*** 3.37*** 3.02*** 2.84*** -0.34 -0.17 -1.46 -0.26
(6.056) (5.613) (10.257) (9.182) (-0.381) (-0.183) (-1.125) (-0.189)
Observations 245 245 245 245 245 245 226 226
Adjusted R-squared 0.217 0.217 0.335 0.346 0.060 0.070 0.101 0.152
57
Table 4: Effect of new regulation on pay-performance sensitivity for UK CEOs
This table presents the results of OLS regressions to estimate CEO’s pay sensitivity to performance. Panel A presents the results for UK BIPRU firm and the largest UK non-
financial firms. Panel B and Panel C present the same estimation using the propensity -score matched samples of US and EU banks respectively. Log CEO Total Compensation is
defined as the sum of salary, bonus, other benefits, value of options and LTIPs. Log CEO Incentive Grants is the natural logarithm of CEOs’ incentive pay computed as the sum of
values of deferred bonus, options granted and LTIPs granted. Log CEO Deferred Bonus is the natural logarithm of CEO’s total deferred bonus in a given year. UK BIPRU is equal
to one for BIPRU firms subject to the FSA Remuneration Code. Shareholder return 1-year (2-year) is the one-year (two-year) total return to shareholders. All other variables are defined in Appendix A. To mitigate the effects of extreme observations all continuous variables are winsorized at the 1% and 99% tails of their respective distributions in each
sample year. Values of t-statistics (reported in parentheses) are computed based on robust standard errors clustered at the industry level (Panel A) and robust standard errors
(Panels B and C). ***, **, * designate significance at 1%, 5% and 10% levels respectively.
Panel A: UK BIPRU vs. UK Firms Log CEO Total Compensation Log CEO Cash Bonus Log CEO Incentives Log CEO Deferred Bonus
(1) (2) (3) (4) (5) (6) (7) (8)
UK BIPRU 0.45*** 0.40*** 0.14 -0.35*** 0.45*** 0.46*** 0.12 -0.80***
(11.010) (9.242) (1.624) (-3.054) (6.980) (6.426) (1.676) (-10.573)
Post 2010 indicator -0.13 -0.14 -0.27 -0.30 -0.32* -0.43* 0.14 0.67***
(-1.482) (-1.710) (-1.475) (-1.746) (-1.989) (-2.074) (0.616) (3.618)
UK BIPRU x Post 2010 interaction 0.16 0.07 0.29 -0.48** 0.03 -0.26 0.61** 2.60***
(1.060) (1.037) (1.143) (-2.722) (0.071) (-1.381) (2.531) (17.971)
Shareholder return 1-year -0.07 0.30 0.21 -0.01
(-0.528) (1.356) (0.944) (-0.084)
Shareholder Return 1-year x Post 2010 0.03 -0.29 -0.13 2.04***
(0.265) (-1.629) (-0.757) (10.983)
UK BIPRU x Shareholder return 1-year 0.48*** -0.04 -0.05 1.39***
(3.452) (-0.184) (-0.286) (7.778)
UK BIPRU x Shareholder return 1-year x Post 2010 -0.09 0.62 0.17 -1.58***
(-0.339) (1.336) (0.446) (-5.113)
Shareholder return 2-year -0.06 0.04 -0.09 0.13
(-0.498) (0.233) (-0.885) (0.638)
Shareholder Return 2-year x Post 2010 0.05 0.20 0.23 -0.19
(0.767) (1.390) (1.284) (-0.862)
UK BIPRU x Shareholder return 2-year 0.30** 1.62*** -0.11 0.31**
(2.956) (9.846) (-0.711) (2.376)
UK BIPRU x Shareholder return 2-year x Post 2010 0.27** 0.25 0.81*** -1.43***
(2.256) (1.339) (5.248) (-7.075)
Log(Sales)t-1 0.28*** 0.28*** 0.12 0.12 0.24** 0.23** 0.33*** 0.43***
(10.494) (10.140) (1.277) (1.458) (2.700) (2.522) (4.030) (4.083)
Book to market t-1 0.00 0.06 -0.51** -0.39* 0.54 0.64* 0.44* -0.96**
(0.013) (0.308) (-2.345) (-1.847) (1.444) (1.981) (1.884) (-2.509)
Log(Idio. Risk)t-1 -0.32** -0.34*** -1.47*** -1.44*** -0.07 -0.18 -0.94*** -0.54*
(-2.764) (-3.124) (-4.705) (-4.839) (-0.457) (-1.059) (-3.203) (-2.006)
Log(Tenure)t-1 0.11 0.11 0.11 0.10 0.08 0.07 0.05 0.09
(1.605) (1.682) (0.889) (0.830) (0.954) (0.834) (0.386) (0.608)
Leveraget-1 0.18 0.21 0.45 0.88* -0.81** -0.80** 0.36 1.13
(1.479) (1.739) (1.436) (2.043) (-2.483) (-2.547) (0.822) (1.740)
58
Industry Indicators Yes Yes Yes Yes Yes Yes Yes Yes
Observations 1,491 1,442 1,491 1,442 1,491 1,442 1,491 1,442
Adjusted R-squared 0.169 0.177 0.122 0.138 0.062 0.065 0.186 0.237
59
Panel B: UK BIPRU vs. US Banks
Log CEO Total Compensation Log CEO Cash Bonus Log CEO Incentives Log CEO Deferred Bonus (1) (2) (3) (4) (5) (6) (7) (8)
UK BIPRU 0.07 0.02 -0.58 -0.97* -0.28 -0.01 -0.04 -0.20
(0.323) (0.108) (-1.251) (-1.941) (-0.688) (-0.020) (-0.075) (-0.352)
Post 2010 indicator -0.26 -0.48** 0.27 0.11 -0.15 -0.36 -0.25 -0.29
(-1.315) (-2.183) (0.637) (0.265) (-0.307) (-0.757) (-0.521) (-0.602)
UK BIPRU x Post 2010 interaction -0.07 0.12 -0.77 -0.75 -0.19 -0.35 2.39*** 2.64***
(-0.145) (0.270) (-1.070) (-0.993) (-0.320) (-0.562) (3.017) (3.208)
Shareholder return 1-year -0.79 1.48* 1.38 0.96
(-1.249) (1.925) (1.475) (1.029)
Shareholder Return 1-year x Post 2010 0.20 -3.55*** -0.41 0.21
(0.355) (-3.070) (-0.288) (0.153)
UK BIPRU x Shareholder return 1-year 0.24 -1.54 -0.97 0.42
(0.760) (-1.269) (-0.882) (0.300)
UK BIPRU x Shareholder return 1-year x Post 2010 1.51 4.84*** 0.90 -0.58
(1.361) (2.794) (0.538) (-0.307)
Shareholder return 2-year 0.56*** 0.05 1.22** 1.18*
(3.053) (0.104) (2.146) (1.949)
Shareholder Return 2-year x Post 2010 -0.02 -0.66 0.06 -0.34
(-0.056) (-0.912) (0.070) (-0.406)
UK BIPRU x Shareholder return 2-year 0.21 1.27** -0.19 -0.79
(0.559) (2.023) (-0.248) (-0.854)
UK BIPRU x Shareholder return 2-year x Post 2010 0.37 1.30 0.33 -0.56
(0.578) (1.290) (0.262) (-0.419)
Log(Sales)t-1 0.30*** 0.36*** -0.07 -0.03 0.15 0.26** 0.19** 0.20*
(6.600) (8.288) (-0.804) (-0.290) (1.447) (2.425) (1.983) (1.940)
Book to market t-1 0.11 0.17** -0.30** -0.13 0.47*** 0.52*** 0.57*** 0.65***
(1.300) (2.347) (-2.098) (-1.387) (2.611) (3.890) (3.630) (3.370)
Log(Idio. Risk)t-1 0.04 0.18* -1.75*** -1.75*** 1.20*** 1.51*** -0.20 -0.25
(0.416) (1.842) (-8.414) (-7.472) (4.851) (5.634) (-0.726) (-0.817)
Log(Tenure)t-1 0.15 0.12 -0.15 -0.17 0.25 0.26 0.05 0.01
(1.643) (1.300) (-0.813) (-0.949) (1.261) (1.327) (0.262) (0.036)
Leveraget-1 0.01 0.16 0.03 0.48 0.26 -0.05 1.18 1.48*
(0.043) (0.510) (0.050) (0.789) (0.385) (-0.074) (1.560) (1.854)
Intercept 3.61*** 2.73*** 4.80*** 3.81*** 0.13 -1.02 -1.88* -2.23*
(7.446) (5.641) (4.623) (3.497) (0.116) (-0.886) (-1.738) (-1.887)
Observations 304 295 304 295 304 295 284 276
Adjusted R-squared 0.174 0.213 0.233 0.237 0.131 0.151 0.115 0.113
60
Panel C: UK BIPRU vs. EU Banks
Log CEO Total Compensation Log CEO Cash Bonus Log CEO Incentives Log CEO Deferred Bonus (1) (2) (3) (4) (5) (6) (7) (8)
UK BIPRU 0.45* 0.30 -1.44*** -1.58*** 0.97*** 1.04** -0.63 -0.65
(1.747) (1.097) (-2.729) (-2.838) (2.729) (2.570) (-1.070) (-1.046)
Post 2010 indicator -0.30 -0.45* 0.24 -0.08 0.76** 0.70* 0.42 0.19
(-1.144) (-1.694) (0.558) (-0.175) (2.095) (1.870) (0.640) (0.282)
UK BIPRU x Post 2010 interaction -0.06 0.11 -0.66 -0.64 -1.12** -1.21** 1.81** 2.13**
(-0.121) (0.220) (-0.913) (-0.841) (-2.194) (-2.263) (2.035) (2.327)
Shareholder return 1-year 0.50 -0.02 0.64 0.54
(1.607) (-0.023) (0.908) (0.761)
Shareholder Return 1-year x Post 2010 0.03 -1.45 0.60 -1.33
(0.054) (-1.419) (0.572) (-0.892)
UK BIPRU x Shareholder return 1-year -0.26 0.05 -0.79 1.10
(-0.492) (0.044) (-0.789) (0.862)
UK BIPRU x Shareholder return 1-year x Post 2010 0.79 2.70 0.14 0.91
(0.773) (1.575) (0.096) (0.458)
Shareholder return 2-year 0.30 0.60 0.18 0.96**
(1.295) (1.616) (0.651) (2.265)
Shareholder Return 2-year x Post 2010 0.37 -0.10 0.46 0.09
(1.262) (-0.207) (0.763) (0.105)
UK BIPRU x Shareholder return 2-year 0.04 1.32* -0.46 0.03
(0.112) (1.931) (-0.800) (0.031)
UK BIPRU x Shareholder return 2-year x Post 2010 0.22 0.28 0.67 -1.34
(0.367) (0.306) (0.626) (-0.949)
Log(Sales)t-1 0.28*** 0.32*** -0.09 0.05 0.02 0.04 0.19 0.25*
(4.855) (5.411) (-0.880) (0.540) (0.333) (0.523) (1.544) (1.949)
Book to market t-1 0.24*** 0.30*** -0.62*** -0.50*** 0.77*** 0.79*** 0.44*** 0.55***
(3.034) (3.610) (-3.029) (-3.802) (4.114) (4.522) (3.006) (3.037)
Log(Idio. Risk)t-1 -0.39*** -0.38*** -1.15*** -0.80*** -0.04 -0.04 0.55* 0.56*
(-2.892) (-2.714) (-4.993) (-3.782) (-0.199) (-0.177) (1.787) (1.854)
Log(Tenure)t-1 0.05 0.02 0.48*** 0.44** -0.05 -0.10 0.13 0.19
(0.507) (0.204) (2.673) (2.474) (-0.318) (-0.577) (0.608) (0.833)
Leveraget-1 -0.32 -0.19 1.06* 1.27** -0.34 -0.52 1.80** 1.87**
(-0.898) (-0.474) (1.858) (2.086) (-0.743) (-1.085) (2.294) (2.288)
Intercept 3.29*** 2.69*** 5.12*** 3.22*** -0.28 -0.26 -0.87 -1.94
(5.132) (3.886) (4.508) (2.703) (-0.302) (-0.266) (-0.651) (-1.351)
Observations 242 234 242 234 242 234 223 215
Adjusted R-squared 0.232 0.259 0.188 0.261 0.079 0.085 0.170 0.161
61
Table 5: Effect of new regulation on risk-taking
This table presents the results of OLS regressions to estimate the effect of new regulation on risk. Panel A presents results using the idiosyncratic measure of volatility
(Idiosyncratic risk) computed from the market model, Total Volatility, Leverage, Z-score (as a measure of default risk) and CDS spread (5-year average annual credit default
spread) for the UK sample. Panel B and Panel C presents bank-specific measures of risk and compare UK BIPRU financial institutions to US and EU banks respectively . Leverage
ratio (for banks only) is Tier 1 regulatory capital divided by tangible assets adjusted by derivative liabilities. Tier 1 Ratio is the ratio of Tier 1 regulatory capital to risk-weighted
assets. Provisions as a Proportion of Total Loans is a ratio of annual provisions to average total loans for the period. C&I Loans as Proportion of Total Loans is a ratio of commercial and industrial loans to average total loans for the period. NPL as a Proportion of Total Loans is a ratio of non-performing loans (loans overdue by more than 90-days)
to average total loans for the period. UK BIPRU is equal to one for BIPRU firms subject to the FSA Remuneration Code. Post 2010 takes the value of 1 for years following 2010.
Deferred bonus is the share of deferred bonus in total compensation in given year. All other variables are defined in Appendix A. The sample period is 2006-2012. To mitigate the
effects of extreme observations all continuous variables are winsorized at the 1% and 99% tails of their respective distribut ions in each sample year. Values of t-statistics (reported
in parentheses) are computed based on robust standard errors clustered at the industry level (Panel A) and robust standard errors (Panels B and C). ***, **, * designate significance
at 1%, 5% and 10% levels respectively.
Panel A: UK BIPRU vs. UK Firms (General risk)
Idiosyncratic Risk Total Volatility Leverage Z-Score CDS spread
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
UK BIPRU -0.00 -0.02*** -0.01** -0.04*** 0.12*** 0.12*** 0.33*** 0.46*** 0.00 -0.00** -0.00 -0.00 -0.00 -0.00
(-0.119) (-3.262) (-2.477) (-5.881) (22.257) (17.458) (13.205) (11.991) (0.345) (-2.967) (-0.135) (-0.872) (-0.654) (-0.371)
Post 2010 -0.07*** -0.08*** -0.09*** -0.10*** -0.05*** -0.07*** 0.04 0.01 0.01*** 0.01*** 0.02***
(-5.025) (-5.598) (-4.916) (-5.167) (-4.914) (-4.408) (0.800) (0.230) (4.300) (6.758) (12.988)
UK BIPRU x Post 2010 0.01 0.07*** -0.00 0.07*** -0.02** 0.00 -0.41*** -0.46*** -0.01** 0.09*** 0.05 -0.00 0.08*** 0.06***
(1.284) (7.404) (-0.002) (6.209) (-2.945) (0.383) (-5.943) (-10.331) (-2.537) (3.566) (1.588) (-1.415) (4.066) (4.272)
% Deferred bonus -0.14*** -0.16*** -0.12** -0.06 -0.01** 0.00 -0.00 -0.00
(-5.047) (-4.716) (-2.336) (-0.215) (-2.746) (0.101) (-1.103) (-0.464)
% Deferred bonus x Post 2010 0.12*** 0.13*** 0.15** 0.23 -0.02** -0.01** 0.01*** 0.01**
(4.193) (3.595) (2.550) (0.858) (-2.761) (-2.686) (3.946) (2.766)
UK BIPRU x % Deferred bonus 0.16*** 0.24*** 0.03 -0.63** 0.02*** 0.00 0.00 0.00
(5.420) (6.177) (0.760) (-2.615) (4.520) (1.135) (0.813) (0.626)
UK BIPRU x % Deferred bonus x Post 2010 -0.27*** -0.35*** -0.15** 0.29 -0.18*** -0.10 -0.15*** -0.12***
(-10.856) (-10.990) (-2.582) (1.032) (-4.022) (-1.695) (-4.145) (-4.396)
Log(Sales)t-1 -0.01** -0.01*** -0.00 0.00 0.01 0.01 0.07*** 0.07*** 0.00*** 0.00*** 0.00*** -0.00 -0.00 0.00
(-2.958) (-3.125) (-0.120) (0.036) (0.768) (0.885) (3.259) (3.104) (3.902) (4.137) (4.623) (-1.685) (-1.453) (0.649)
Book to market t-1 0.03** 0.03** 0.04** 0.04** 0.08*** 0.08*** -0.06 -0.05 0.01 0.01* 0.00 0.00 0.00 0.00
(2.210) (2.403) (2.183) (2.267) (3.284) (3.196) (-0.274) (-0.223) (1.732) (2.136) (1.043) (0.723) (1.175) (1.190)
Leverage t-1 0.08 0.09* 0.05 0.06 0.48*** 0.48*** 0.71*** 0.69*** -0.01** -0.01*** -0.01** -0.00 -0.00 -0.00
(1.682) (1.816) (0.889) (1.006) (8.063) (8.143) (4.032) (3.877) (-2.907) (-3.071) (-2.589) (-0.212) (-0.387) (-1.210)
ROA t-1 0.01 0.00
(1.226) (0.551)
Size t-1 -0.00** -0.00
(-2.698) (-1.231)
Total volatility t-1 0.05*** 0.02***
(7.939) (3.631)
62
Industry Indicators Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Year Indicators No No No No No No No No No No No Yes Yes Yes
Observations 1,538 1,538 1,537 1,537 1,537 1,537 1,429 1,429 499 499 498 499 499 498
Adjusted R-squared 0.153 0.164 0.149 0.166 0.521 0.524 0.268 0.271 0.110 0.125 0.572 0.979 0.980 0.980
63
Panel B: Banks-specific measures of risk (UK BIPRU vs. US Banks)
Leverage Ratio
RWA as a
Proportion of
Total Assets Tier1 Ratio
Provisions as a
Proportion of
Total Loans
C&I Loans as
Proportion of
Total Loans
NPL as a
Proportion of
Total Loans Z-Score CDS-Spread
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
UK BIPRU 7.37*** 8.05*** 0.05 0.05 -0.04 -0.05 0.03 0.02 0.10** 0.11* 0.00*** 0.00** -0.09 -0.12 0.00 -0.00
(3.936) (3.186) (0.628) (0.494) (-1.495) (-1.411) (1.054) (0.628) (2.304) (1.677) (3.296) (2.280) (-0.582) (-0.613) (0.108) (-0.689)
Post 2010 0.59 1.17* -0.03 0.02 -0.01 -0.05 0.06 0.07 -0.09** -0.12** 0.00* 0.00 0.25** 0.37*** 0.01 0.01
(1.513) (1.894) (-0.531) (0.288) (-0.488) (-1.116) (0.876) (0.777) (-2.452) (-2.030) (1.870) (0.636) (2.256) (2.745) (1.006) (0.854)
UK BIPRU x Post 2010 -1.72 -4.00 -0.06 -0.00 0.07** 0.10** -0.05 -0.01 0.09 0.20** -0.00 0.00 -0.46* -0.57 -0.00 0.17***
(-0.603) (-0.844) (-0.633) (-0.034) (2.256) (2.094) (-0.565) (-0.212) (1.335) (2.245) (-0.193) (0.429) (-1.791) (-1.366) (-0.191) (2.670)
% Deferred bonus 4.34 -0.20 -0.67** -0.31 -0.17 -0.00*** -0.70 -0.01
(1.586) (-0.883) (-2.163) (-1.521) (-0.288) (-2.879) (-0.966) (-1.042)
% Deferred bonus x Post 2010 -10.67** -0.03 -0.72** 0.06 0.37 0.00* 0.10 -0.00
(-2.413) (-0.091) (2.079) (0.343) (0.549) (1.662) (0.118) (-0.001)
UK BIPRU x % Deferred bonus 0.59 0.49* 0.70** 0.26 0.19 0.00** -0.18 0.01
(0.040) (1.856) (2.241) (1.274) (0.319) (2.450) (-0.208) (1.064)
UK BIPRU x % Deferred bonus x
Post 2010 11.86 -0.52 -0.76** -0.28 -0.68 -0.00* 0.79 -0.33***
(0.709) (-1.180) (-2.178) (-1.095) (-0.958) (-1.749) (0.607) (-2.730)
Log(Sales)t-1 0.22 0.21 -0.03*** -0.03*** 0.00 -0.00 -0.00 -0.00 0.01 -0.00 -0.00*** -0.00*** -0.02 0.01 0.00 0.00
(1.581) (1.175) (-3.214) (-3.248) (0.286) (-0.311) (-0.716) (-0.626) (1.077) (-0.060) (-3.961) (-3.560) (-0.755) (0.272) (1.036) (1.201)
Book to market t-1 0.55** 0.37 -0.26* -0.24** -0.09 0.23 0.07 0.08 0.07 -0.02 0.00 0.00* 0.15* 0.17*** 0.01 0.01*
(2.111) (0.717) (-1.952) (-2.076) (-0.285) (0.679) (0.312) (0.326) (0.251) (-0.068) (1.337) (1.714) (1.941) (3.417) (0.979) (1.819)
Leverage t-1 5.04*** 7.44*** 0.12 0.10 -0.56** -0.75*** -0.26 -0.38 0.02 -0.09 -0.00** -0.00** 1.25*** 0.74*** -0.01 -0.01*
(3.624) (3.258) (0.835) (0.658) (-2.148) (-2.878) (-1.451) (-1.571) (0.123) (-0.305) (-2.500) (-2.434) (6.791) (3.660) (-1.221) (-1.832)
Intercept -7.30*** -9.28*** 1.19*** 1.10*** 0.71*** 0.63*** 0.22** 0.34*** 0.15 0.44 0.00*** 0.00*** -0.69** -0.71** 0.00 -0.00
(-3.817) (-3.581) (6.694) (5.917) (4.196) (3.187) (2.383) (2.632) (0.756) (1.315) (4.203) (4.618) (-2.411) (-2.155) (0.087) (-0.344)
Observations 173 173 98 98 132 132 146 146 114 114 156 118 267 267 97 86
Adjusted R-squared 0.270 0.249 0.059 0.060 0.436 0.428 0.012 0.026 0.064 0.117 0.263 0.288 0.196 0.174 0.021 0.032
64
Panel C: Banks-specific measures of risk (UK BIPRU vs. EU Banks)
Leverage Ratio
RWA as a
Proportion of
Total Assets Tier1 Ratio
Provisions as a
Proportion of
Total Loans
C&I Loans as
Proportion of
Total Loans
NPL as a
Proportion of
Total Loans Z-Score CDS-Spread
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
UK BIPRU 7.60*** 5.84*** 0.26*** 0.26** -0.01 -0.02 -0.01 0.01 0.04 0.26 -0.00 -0.00 -0.33** -0.05 -0.00 -0.01
(4.101) (3.107) (3.535) (2.328) (-1.388) (-1.488) (-0.749) (0.508) (0.602) (1.642) (-0.430) (-0.788) (-2.238) (-0.264) (-0.715) (-1.285)
Post 2010 0.06 -0.45 -0.02 -0.12 0.02** 0.02 -0.01 0.01 -0.13 0.11 -0.00 0.00 -0.07 0.01 -0.01 -0.01
(0.113) (-0.492) (-0.286) (-1.012) (2.143) (1.339) (-1.152) (0.524) (-1.622) (0.617) (-0.213) (0.086) (-0.511) (0.044) (-0.698) (-0.623)
UK BIPRU x Post 2010 -1.16 -0.67 -0.05 0.06 0.01 0.01 0.03 -0.00 0.20* -0.01 0.00 0.00 -0.13 -0.42 0.01 0.12
(-0.389) (-0.145) (-0.460) (0.418) (0.528) (0.605) (1.426) (-0.030) (1.822) (-0.055) (0.008) (1.098) (-0.480) (-0.891) (0.849) (1.289)
% Deferred bonus -2.21 0.07 -0.04 0.05 0.41 -0.00 0.44 -0.02
(-1.612) (0.378) (-1.109) (1.281) (0.899) (-0.952) (0.727) (-1.626)
% Deferred bonus x Post 2010 0.41 0.22 0.00 -0.10* -1.12* 0.00 -0.90 -0.07
(0.226) (0.764) (-0.040) (-1.849) (-1.967) (0.994) (-1.106) (-1.170)
UK BIPRU x % Deferred bonus 8.29 0.17 0.08* -0.09 -0.37 0.00 -1.17 0.02
(0.713) (0.739) (1.837) (-1.553) (-0.781) (1.007) (-1.471) (1.510)
UK BIPRU x % Deferred bonus x
Post 2010 -1.76 -0.69* -0.03 0.11 0.82 -0.00* 1.74 -0.13
(-0.121) (-1.947) (-0.547) (1.240) (1.348) (-1.789) (1.250) (-0.848)
Log(Sales)t-1 0.04 0.05 -0.04** -0.01 0.01* 0.00 0.01*** 0.01*** -0.02 -0.02 -0.00 -0.00 0.01 0.04 0.00 0.01
(0.147) (0.190) (-1.994) (-0.293) (1.784) (0.519) (3.381) (3.057) (-0.890) (-0.519) (-0.357) (-0.715) (0.177) (0.891) (1.624) (1.547)
Book to market t-1 0.49* 0.31 -0.43** -0.20 0.33** 0.41*** -0.14*** -0.14*** -0.21 0.63 -0.00 0.00 0.16** 0.17** 0.00 0.01
(1.978) (0.820) (-2.411) (-1.128) (2.300) (3.665) (-8.445) (-7.065) (-0.205) (0.497) (-0.036) (0.439) (1.977) (2.467) (0.104) (0.921)
Leverage t-1 3.47** 4.30** -0.06 -0.30 -0.67*** -0.69*** -0.28*** -0.31*** -0.85 -0.82 -0.00*** -0.00*** 1.45*** 1.24*** -0.00 -0.01
(2.415) (2.379) (-0.185) (-0.914) (-5.058) (-6.599) (-7.518) (-8.043) (-0.669) (-0.518) (-6.628) (-6.205) (6.830) (4.905) (-0.618) (-0.663)
Intercept -3.58 -3.65 1.43*** 0.96*** 0.32*** 0.32*** 0.29*** 0.27*** 1.76*** 0.56 0.00*** 0.00*** -0.87** -1.27*** -0.03 -0.04
(-1.423) (-1.192) (4.357) (2.694) (7.245) (5.529) (6.476) (4.791) (2.714) (0.707) (5.608) (4.765) (-2.261) (-2.831) (-1.060) (-1.125)
Observations 127 127 88 88 90 90 115 115 115 115 90 90 223 223 63 58
Adjusted R-squared 0.256 0.215 0.320 0.327 0.742 0.807 0.612 0.619 0.177 0.210 0.733 0.751 0.267 0.238 0.026 0.064
65
Table 6: Effect of new regulation on CEO turnover
This table presents conditional logistic (Panel A) and logistic regressions (Panels B and C) to estimate the effect of new
regulation on the likelihood of CEO turnover. Deferred bonus is the share of deferred bonus in total compensation in given year.
All variables are defined in Appendix A. The sample period is 2007-2012. To mitigate the effects of extreme observations all
continuous variables are winsorized at the 1% and 99% tails of their respective distributions in each sample year. Values of z-
statistics (reported in parentheses) are computed based on robust standard errors. ***, **, * designate significance at 1%, 5% and 10% levels respectively.
Panel A: UK BIPRU vs. UK Firms
(1) (2)
UK BIPRU 0.25*** 0.47***
(3.479) (6.062)
Post 2010 -0.18 -0.09
(-1.044) (-0.498)
UK BIPRU x Post 2010 0.40** 0.52***
(2.405) (2.873)
% Deferred bonus 1.26**
(2.044)
% Deferred bonus x Post 2010 -1.10
(-1.438)
UK BIPRU x % Deferred bonus -1.73***
(-2.744)
UK BIPRU x % Deferred bonus x Post 2010 0.56
(0.751)
Shareholder return t 0.12 0.12
(1.170) (1.081)
ROA t-1 -2.01*** -1.96***
(-3.766) (-3.574)
Book to market t-1 0.29*** 0.30***
(3.482) (3.651)
Log(Tenure) t-1 -0.36** -0.35**
(-2.315) (-2.263)
Industry Indicators Yes Yes
Observations 1,535 1,535
Pseudo R-squared 0.0249 0.0284
66
Panel B: UK BIPRU vs. US Banks
(1) (2)
UK BIPRU 0.00 0.35
(0.006) (0.765)
Post 2010 -0.82** -0.98**
(-2.184) (-2.030)
UK BIPRU x Post 2010 1.07* 1.46*
(1.891) (1.927)
% Deferred bonus 0.77
(0.783)
% Deferred bonus x Post 2010 1.66
(0.888)
UK BIPRU x % Deferred bonus -2.01
(-1.166)
UK BIPRU x % Deferred bonus x Post 2010 -2.17
(-0.799)
Shareholder return t 0.20 0.30
(0.851) (1.188)
ROA t-1 -2.81 -2.93
(-1.427) (-1.480)
Book to market t-1 0.18 0.16
(1.050) (0.939)
Log(Tenure) t-1 0.12 0.16
(0.857) (1.063)
Intercept -1.44*** -1.62***
(-4.213) (-3.974)
Observations 335 306
Pseudo R-squared 0.0356 0.0511
67
Panel C: UK BIPRU vs. EU Banks
(1) (2)
UK BIPRU 0.48 0.17
(1.126) (0.308)
Post 2010 0.51 -0.41
(1.204) (-0.648)
UK BIPRU x Post 2010 -0.26 0.92
(-0.436) (1.078)
% Deferred bonus -3.36
(-1.197)
% Deferred bonus x Post 2010 4.75
(1.410)
UK BIPRU x % Deferred bonus 2.16
(0.687)
UK BIPRU x % Deferred bonus x Post 2010 -5.46
(-1.409)
Shareholder return t 0.18 0.14
(0.641) (0.466)
ROA t-1 0.79 -0.32
(0.414) (-0.166)
Book to market t-1 0.14 0.18
(0.821) (1.054)
Log(Tenure) t-1 0.12 0.22
(0.739) (1.263)
Intercept -1.99*** -1.66***
(-4.697) (-2.934)
Observations 260 239
Pseudo R-squared 0.0185 0.0448
68
Table 7: Compensation contract changes
This table presents the univariate differences of CEO compensation contracts in the UK and the complexity of the underlying compensation contracts for UK firms subject to the
Remuneration Code (BIPRU) and the largest UK companies (constituents of FTSE350, excluding BIPRU). All values are in real, 2012 thousands of pound sterling (£thousands)
unless otherwise indicated. UK BIPRU firms are compared to other UK FTISE350 constituents before and after the introduction of the Remuneration Code. This table also shows
the complexity of UK CEO compensation contracts in relation to the number of targets and corresponding changes. To mitigate the effects of extreme observations all continuous
variables are winsorized at the 1% and 99% tails of their respective distributions in each sample year. All variables are defined in Appendix A. ***, **, * designate significance at 1%, 5% and 10% levels respectively .
Before FSA Regulation in 2010 After FSA Regulation in 2010 Difference - in - Differences
Variables Control
(UK
O ther)
Treated (UK
BIPRU)
Difference T-statistic Control
(UK
O ther)
Treated (UK
BIPRU)
Difference T-statistic T-statistic
(1) (2) (3) = (2) - (1) (4) (5) (6) (7) = (6) - (5) (8) (9) = (7) - (3) (10)
Total compensation 1,540.28 3,571.68 2,031.39 (10.31)*** 2,055.48 4,759.78 2,704.30 (13.27)*** 672.91 (14.445)***
Take home pay 776.52 1,950.45 1,173.93 (7.735)*** 1,067.46 2,287.07 1,219.61 (11.413)*** 45.678 (3.821)***
% Salary in total compensation 0.545 0.472 -0.073 (-4.475)*** 0.528 0.360 -0.169 (-40.732)*** -0.096 (-6.111)***
% Incentive pay in total compensation 0.076 0.022 -0.055 (-2.76)*** 0.064 0.042 -0.022 (-1.671)* 0.033 (4.121)***
% Deferred bonus in total compensation 0.086 0.241 0.155 (15.127)*** 0.113 0.306 0.194 (31.177)*** 0.039 (7.208)***
Annual bonus scheme 0.984 1.000 0.016 (1.448) 0.987 0.974 -0.013 (-1.787)* -0.029 (-6.446)***
Annual bonus deferral scheme 0.601 0.929 0.328 (13.321)*** 0.724 1.000 0.276 (7.015)*** -0.051 (-1.922)*
Deferral policy compulsory 0.592 0.857 0.265 (10.079)*** 0.717 1.000 0.283 (7.266)*** 0.018 (0.703)
Deferral policy voluntary 0.087 0.036 -0.051 (-2.943)*** 0.095 0.077 -0.018 (-0.918) 0.033 (2.099)**
Number of live LTIP schemes 0.060 0.036 -0.024 (-1.69)* 0.037 0.000 -0.037 (-3.68)*** -0.013 (-0.776)
Number of live Option schemes 0.163 0.143 -0.020 (-1.065) 0.154 0.256 0.102 (6.493)*** 0.122 (5.14)***
Shareholding requirement 0.100 0.143 0.043 (4.441)*** 0.086 0.192 0.107 (5.774)*** 0.064 (3.895)***
Total number of unique perf. targets per year 6.417 6.929 0.511 (3.246)*** 5.494 5.949 0.454 (3.325)*** -0.057 (-0.43)
Total number of performance targets per year 2.412 5.387 2.975 (2.244)** 2.368 5.773 3.404 (2.583)*** 0.429 (3.751)***
Total number of bonus targets per year 3.250 3.964 0.714 (6.663)*** 3.980 4.385 0.405 (3.335)*** -0.309 (-3.458)***
Total number of LTIP targets per year 1.195 1.036 -0.159 (-2.012)** 1.429 1.487 0.058 (1.359) 0.217 (2.594)***
Total number of Option targets per year 0.849 0.964 0.116 (1.656)* 0.559 0.641 0.082 (1.773)* -0.034 (-0.509)
Total number of changes in a contract per year 3.573 3.429 -0.145 (-0.795) 3.591 3.205 -0.386 (-2.389)*** -0.241 (-1.069)
Total number of changes to bonus targets 0.974 0.643 -0.331 (-4.178)*** 0.537 0.692 0.156 (2.239)** 0.487 (4.043)***
Total number of changes to LTIP targets 0.429 0.321 -0.107 (-2.546)*** 0.372 0.359 -0.013 (-0.396) 0.095 (1.528)
Total number of changes to Option targets 0.266 0.143 -0.123 (-3.194)*** 0.214 0.179 -0.034 (-1.14) 0.089 (2.151)**
Number of contract changes (weighted) 0.974 0.643 -0.331 (-4.178)*** 0.537 0.692 0.156 (2.239)** 0.487 (4.043)***
top related