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)***