University of Amsterdam
Faculty of Economics and Business (ABS) Master thesis Accountancy & Control
Master thesis Control:
A research into the effects of formula-based bonuses and discretionary bonuses for executives prior to the crisis on the financial performance during the financial crisis: empirical evidence
from the U.S. banking industry
Student: Meiqi de Vries Student Number: 10025049 Date of first version: 22 June 2015 Thesis supervisor: Dr. ir. S.P. van Triest Word count: 29.570
2
Acknowledgements
This thesis is written as completion to the Master Programme of Accountancy & Control at the University of Amsterdam. Although many see the thesis process as an individual task, the
current version of this master thesis would not have existed without the help and suggestions of others. During the final completion of this thesis, I realized that the suggestions I received
from these people have possibly been more important than the individual research itself. Therefore, I would like to thank the persons who made the current version of this thesis
possible.
First, I would like to thank my thesis supervisor Sander van Triest for his constructive feedback and patience during the thesis process. Especially the latter would not have always
been easy, as the perspective I used throughout this study differs considerably from the mainstream accounting and control literature.
Second, I would like to thank my friend Tarah Huisden for her assistance on the extensive revision work of this thesis. Perhaps more important, I owe her for the insight to view an
discretionary bonus as an unexpected reward. This insight made it possible to apply psychology studies in the formulation of the hypotheses and it simplified the compensation
data collection process.
Third, I would like to thank my former classmate Marianne Speckmann, who made me realize that not all bonus incentives are strictly formula-based, as some bonuses are paid out despite not achieving any pre-set performance targets. This led to the separate consideration of both
formula-based bonus and discretionary bonus in this study.
Lastly, I would like to thank my parents, HuiSheng and Klaas, for their continuous mental and financial support during my studies at the University of Amsterdam. Also, my gratitude goes
out to friends and colleagues for always being flexible and patient.
Thanks to all these people, I have been able to stay truly intrinsically motivated during the complete research process.
Meiqi de Vries
Alphen aan den Rijn, 22 juni 2015
3
Statement of originality
This document is written by student Meiqi de Vries who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This study investigates, using a regression-based approach, whether financial performance during the recent financial crisis is related to executive bonus incentives before the crisis. The focus of this study is on the performance-contingent and discretionary bonuses that are paid to the top executives of 422 U.S. banks in 2007. In addition, five measures of banking
performance in 2008 are considered. I found evidence that the amount of performance-contingent cash compensation earned by the highest-paid executive prior to the crisis is negatively related to the banking performance during the crisis. Furthermore, I found some evidence that the amount of discretionary bonus compensation prior to the crisis is positively related to banking performance during the crisis. The first-mentioned negative effect is considerably larger than the second-mentioned positive effect, resulting in an overall negative effect of total cash compensation prior to the crisis on banking performance during the crisis. Finally, managerial power, as measured by the CEO Pay Slice, does not influence the
incentives-performance relationship for the sample banks. The findings are explained using the optimal incentive contracting theory and experimental studies from cognitive psychology.
Keywords: Financial crisis, executive variable compensation, CEO incentives, performance-contingent
4 Table of contents Acknowledgements ... 2 Statement of originality ………...3 Abstract ………...3 1. Introduction …...………...6 1.1 Relevance ...………... 6 1.2 Contributions ..….………….………... 7 1.3 Research question …...………... 8 1.4 Findings ……….. 8
2. Literature review and hypotheses ...………... 10
2.1 Introduction ………... 10
2.2 Executive compensation ……….. 10
2.2.1 Composition of executive compensation ……… 10
2.2.2 Composition of variable executive compensation ……….. 11
2.2.3 Discretionary bonus payments ……… 12
2.3 Formula-based incentives …..….………..13
2.3.1 Initial conceptual framework ……….. 13
2.3.2 Bonner et al. (2000) ……… 14
2.3.3 The economic view of formula-based incentives ………... 15
2.4 Discretionary incentives ……….. 16
2.4.1 Economic Theory …...………...……….. 16
2.4.2 Managerial power theory ……...………. 17
2.4.3 Cognitive psychology ……….. 19
2.5 Incentives in the recent financial crisis ……… 23
2.5.1 Motivation for the research setting ………. 23
2.5.2 Prior literature ……….……… 23
2.5.4 Current study ……….. 24
2.6 Hypotheses development ……...………..……… 25
2.6.1 Formula-based bonus payments ………. 25
2.6.2 Discretionary bonus payments ……… 27
3. Research methodology ……...……… 29
3.1 Introduction ………...………... 29
3.2 Sample and data ………..………. 29
3.3 Measurement of formula-based and discretionary bonus ……… 31
3.4 Measurement of financial performance .………... 34
3.5 Measurement of executive power ……….………... 35
3.6 Empirical models and control variables ………... 37
4. Empirical results ……… 40
5
4.2 Descriptive examples ……….……….. 40
4.3 Descriptive statistics ……….……… 43
4.4 Main findings ………... 48
4.4.1 Executive incentives and bank performance during the crisis………...….………. 48
4.4.2 Formula-based bonus and managerial power ...………... 51
4.4.3 Discretionary bonus and managerial power ...……….. 51
4.5 Robust analyses …………...……… 53
4.5.1 First robustness check ………. 53
4.5.2. Second robustness check ……… 54
5. Conclusion and discussion …..………... 58
5.1 Introduction ……… 58
5.2 Answer to the research question ……… 58
5.3 Discussion of the results ……… 59
5.4 Implications ……… 61 5.5 Limitations ………. 62 References ……… 63 Appendix A ……….………. 66 Appendix B ……….………. 68 Appendix C ……….………. 69 Appendix D ……….……….… 73
6
1. Introduction
1.1 Relevance
―Perhaps nowhere is this clearer than in the economic calamity that gripped the world economy in 2008 and 2009. Each player in the system focused only on the short-term reward – the buyer who wanted a house, the mortgage broker who wanted a commission, the Wall Street trader who wanted new securities to sell, the politician who wanted a buoyant economy during reelection – and ignored the long-term effects of their actions on themselves or others. When the music stopped, the entire system nearly collapsed. This is the nature of economic bubbles: What seems to be irrational exuberance is ultimately a bad case of extrinsically motivated myopia.‖ (Pink, 2009, p. 21)
The purpose of this research is to examine the effect of executive bonus incentives prior to the recent financial crisis on banking performance during the crisis in the United States. Specifically – besides the effect of the overall bonus incentives – the effects of the two separate components of total bonus incentives will be studied. For the first component, formula-based bonus incentives, the actual payment is contingent on the achievement of certain pre-determined performance goals. Based on this characteristic, the monthly salary is not a formula-based bonus incentive as payment does not vary with the performance level. Both the terms formula-based bonus and performance-contingent bonus will be used interchangeably in this paper. The second component, discretionary bonus incentives, is not dependent on the achievement of specific previously set performance goals. Indeed, the payment of a discretionary bonus occurs according to the subjective judgements of the principal, thus without the fulfilment of any objective performance criteria. The current remuneration practices usually have some element of contingency, whether we are considering monetary, non-monetary, tangible or non-tangible incentives. Furthermore, this tendency seems to have increased since the 1990s.
The statement above from the bestseller author Daniel Pink clarifies how the incentive systems within the American financial sector ultimately led to the destructive effects of the recent financial crisis, which still has a non-negligible impact on most economies globally. When reading the financial section of any newspaper, it seems that the Dutch economy struggles with the aftermath of the financial crisis on a daily basis. In 2009, the Dutch Central Bank has conducted an independent study that investigates the executive compensation policies at 27 major Dutch banks, insurance companies and pension funds. The introduction of the report states that the short-term and asymmetric incentive systems in the financial industry are major causes of the recent financial crisis, because the systems encouraged excessive risk-taking behaviour from executives and employees of the financial sector (DNB, 2009, p. 3). This implies that the U.S. situation to some extent applies for the Dutch context as well (see quotation above). The Dutch Central Bank concludes that financial firms need to set a maximum on the variable compensation for executives, that firms need to use consistent policies in the assignment of variable compensation, and that the criteria of performance-contingent incentives need to be communicated clearly and ex-ante (DNB, 2009, p. 11). In response to the recent financial crisis, various national regulations have been set in North America and the European Union that aim to limit the amount of variable incentives in
7
the financial sector. In the case of the European Union, the fourth Capital Requirements Directive has been formally approved by the European Parliament in 2013, including a bonus cap on bankers’ pay. An important principle in this Directive entails that the variable component may not exceed 100% of the fixed component in the total remuneration for each individual working in a European financial institution (Freshfields Bruckhaus Deringer, 2013, p. 2). Because the cap is a EU maximum, national member states may determine a lower national maximum percentage for the variable component of the remuneration. The United Kingdom has even challenged the European bonus cap at the Court of Justice in Luxembourg. The country argued that the requirements were a disproportionate way of achieving EU legislators’ aims and that bankers’ pay was an issue for shareholders rather than the EU to regulate (The Telegraph, 2014). Eventually, the United Kingdom dropped its challenge to the European bonus cap after an adviser to the Court of Justice rejected all of the country’s claims (Anderson, 2014).
The Netherlands have had the highest ambitions among all EU member states concerning the bonus cap on bankers’ compensation. A proposal has been approved by the House of Representatives and Dutch Senate in February 2015, which sets the maximum variable component of the bankers’ pay on 20% of the fixed component. In addition, the variable compensation payment to each employee working in financial firms with registered offices in the Netherlands, must be largely (>50%) based on non-financial performance measures (Staatsblad, 2015). Due to its strict nature, the initial proposal has been received with a considerable amount of criticism by some authors. They see this cap as symbol policy, as it could be easily circumvented by increasing other components of the bankers’ compensation (Witteveen & Van Tuyll van Serooskerken, 2014, p. 3; Luyendijk, 2015).
1.2 Contributions
The main contributions of the current study to the existing body of accounting literature on executive incentives are threefold and interrelated. First, the incentive-performance relationship will be examined through the separation of bonus incentives into its discretionary and formula-based cash bonuses. To my knowledge, only one prior study has employed this method in the regression of future financial performance on current variable compensation, while controlling for current financial performance (Ederhof, 2010). Second, besides the optimal incentive contracting and managerial power theory, further explained in paragraph 2.4, findings from the psychology field are used to develop the hypotheses for this study. These latter findings are originally obtained and replicated in experimental and field studies from cognitive psychology. Third, the hypotheses are tested using a sample that consists solely of U.S. banks, as the findings from the psychology literature require the use of extremely high incentives to reward specific tasks that are carried out in a highly competitive environment in order to be applicable. Therefore, I chose the specific research setting of the U.S. banking industry during the recent financial crisis, because I expect this setting is mostly compliant with the requirements from the psychology literature used. No prior literature has come to my attention that studies the effect of formula-based and discretionary bonuses on the future financial performance, while employing the context of the banking sector during the recent financial crisis.
8
In addition, I contribute to the existing literature by manually collecting all compensation data from the original proxy statements, which U.S. banks are required to disclose to the Securities and Exchange Commission (SEC). The hand collection of the compensation data is a major contribution of this study, as the consistency and reliability of the data may possibly be the highest – in comparison to studies which retrieve data from the popular database Execucomp. As I will explain later in this paper, Execucomp highly underestimates overall executive bonus incentives. In addition, this compensation database does not contain data concerning the two separate components of overall bonus incentives. Finally, the sample used for this study is larger than the possible sample size when using a public compensation database.
1.3 Research question
This research examines the levels of formula-based and discretionary bonuses in 2006 and their influence on the financial performance of banks during the financial crisis years. The following research question will be answered in this study:
Are formula-based bonuses and discretionary bonuses for banking executives prior to the financial crisis associated with the financial performance of the banking sector during the crisis?
The compensation data, collected from the proxy statements, were paid to the banking executives in 2007. However, these data are related to the performance of the executives in 2006, thus the cash compensation is earned in 2006. Therefore, 2006 is considered as the base year for this study. Although, we should keep in mind that the bonus incentives are paid in 2007. Considering many view the autumn of 2007 as the start of the recent financial crisis (The Economist, 2013), the effects on the banking performance in both 2007 and 2008 are investigated in this study.
1.4 Findings
The results of both the main regression analyses and the robust analyses provide support for the research question, when we consider the U.S. banking performance in 2008. In 2007, the results are less significant. Based on the empirical results, total bonus incentives in 2006 are negatively related to banking performance in 2008. Specifically, the results support the prediction that higher formula-based bonuses earned by banking executives in 2006 led to worse financial performance in 2008 – under both the main regression analyses and the robust analyses. Furthermore, based on the robust analyses, I found some support for the prediction that higher discretionary bonuses earned by banking executives in 2006 led to better financial performance of the sample banks in 2008. This prediction is not supported based on the main regression analyses. It should be noted that the negative effect of formula-based bonus incentives is considerably more persistent and significant than the positive effect of discretionary bonus incentives. This leads overall to a negative effect of total bonus incentives prior to the financial crisis on banking performance during the crisis. Finally, managerial power, as measured by the CEO Pay Slice, does not negatively affect the bonus incentives– banking performance relationship for the sample banks. Thus, the predictions concerning the
9
CEO Pay Slice are not supported based on the empirical findings. I argue that this lack of significant findings might be explained by the imprecise nature of the CEO Pay Slice as a measure of managerial power.
10
2. Literature Review and Hypotheses
2.1 Introduction
This chapter outlines prior literature on variable executive compensation, based on three fundamental different perspectives. These perspectives are useful to interpret the empirical results of this study, given that prior empirical results on the subject have been interpreted using economic theories only. Paragraph 2.2 describes the different components of a typical executive compensation scheme. Paragraph 2.3 outlines the formula-based bonus incentives using a conceptual framework, which has been used and extended by different authors in their research on variable compensation/incentives. In paragraph 2.4, three different theories are used to explain discretionary bonus incentives: the optimal incentive contracting theory, managerial power theory and cognitive psychology theories. The latter theories have not yet been used in accounting research on the subject of variable executive compensation. In paragraph 2.5, the choice for the financial crisis setting is motivated and relevant studies are reviewed. In paragraph 2.6, four hypotheses are developed, which will be tested in the empirical analysis. Paragraph 2.7 summarizes and concludes.
2.2 Executive compensation
2.2.1 Composition of executive compensation
The recent rise in executive compensation has resulted in a large amount of public attention globally, as evidenced from paragraph 1.1. Although empirical evidence exists which links the patterns in executive pay to the size of large firms across firms, over time and between countries, executive pay has indeed increased dramatically relative to the average worker’s remuneration in the United States (Gabaix & Landier, 2006, p. 93; Teather, 2005). In particular, Gabaix and Landier found that the sixfold increase of CEO pay in the U.S. between 1980 and 2003 can be fully attributed to the sixfold increase in market capitalizations of the companies during that period (2006, p. 93). In addition, Frydman and Jenter provide a graphical representation of the development of CEO compensation relative to the compensation of other best-paid executives in U.S. firms (2010, p. 41). From this graph, I conclude that the median level of executive compensation has increased substantially for Chief Executive Officers, especially between 1990 and 2000. This conclusion could be achieved by comparing the CEO compensation in time, and by comparing the CEO compensation with the median compensation level of other top executives (based on the three highest-paid individuals who are not the CEO).
Despite substantial differences in pay practices across firms, most CEO compensation packages are composed of the following five basic elements: salary, annual bonus, payouts from long-term incentive plans, restricted bonus grants and restricted stock grants (Frydman & Jenter, 2010, p. 6). In addition, CEOs often incidentally receive one-time payments: severance payments (‘golden parachute’), ‘golden hellos’, retention bonus payments, guaranteed bonus payments, various perquisites and eventually, periodic contributions to defined-benefit pension plans (Frydman & Jenter, 2010, p. 6; Teather, 2005; DNB, 2009, p. 25). The following quote illustrates that these incidental payments could involve large dollar amounts:
11
―Hewlett Packard has had a costly few months. Chief executive Carly Fiorina took $42m with her when she was forced out of the computer and printer maker after failing to deliver on profits. Her replacement, Mark Hurd, was then given a $20m golden hello. That, of course, was in addition to his annual salary, which could reach $22m a year.‖ (Teather, 2005)
Figure 1 shows that the composition of CEO compensation packages has changed considerably over time. From 1936 to 1950, CEO compensation was almost composed solely of salaries and annual bonuses. From 2000 to 2005, less than half of the CEO compensation package consists of salaries and annual bonuses. In addition, the annual bonuses were typically formula-based, so that payment was tied to measures of annual accounting performance, and paid in either cash or stock (Frydman & Jenter, 2010, p. 6). Around 60% of CEO pay consists of payout from long-term incentive plans, which are bonus plans based on multi-year performance, stock grants and stock option compensation.
Figure 1. The structure of CEO compensation from 1936 to 2005 (Frydman & Jenter, 2010, p. 42).
2.2.2 Composition of variable executive compensation
Although strong incentives can be provided by an executive’s holdings of stock and stock options, this study focuses solely on the incentive effects provided by bonus plans. In the United States, the board of directors and compensation committee usually determine the annual bonus payments to reward CEOs based on both publicly observable and private signals, known only to the parties involved (Hayes & Schaefer, 2000, p. 278). This line of argument supports the oversight function of the board of directors by using incentive instruments over which the board has direct control (Hayes & Schaefer, 2000, p. 274; p. 278). Hall and Liebman report that the vast majority of variation in executive wealth associated with changes in firm value stems from executives’ holdings of stock and stock options (1998). This view suggests that the boards of directors have completely delegated the tasks of
12
monitoring and rewarding top executives to capital markets as a result of the extensive use of equity-based incentives. Indeed, the purpose of stock and stock option compensation is to tie remuneration directly to share prices and incentivize executives to increase shareholder value (Frydman & Jenter, 2010, p. 6). I support the first view, stating that the board of directors has an oversight function on the firm’s top managers. Therefore, this study examines whether the board of directors perform a governance role by collecting and evaluating performance-related information that is both publicly available and private information that is available only to contract parties (Hayes & Schaefer, 2000, p. 274).
The remainder of this paragraph clarifies the possible elements of variable compensation, which is defined by Hay Group as ‗conditional and thus not guaranteed reward for individual or collective performances as an incentive instrument‘ (Voeten, 2011, p. 1). On an individual level, there are roughly two forms of variable compensation, namely formula-based bonus and discretionary bonus. Formula-based bonus has a high ‘line of sight’, meaning there is a clear relationship between delivering certain pre-specified performances and the accompanying reward. The Dutch Central Bank defines this bonus plan as the actual received amount of bonus contingent on ex-ante performance targets, so that the bonus payment is dependent on the performance of the individual or firm (2009, p. 25).
Discretionary bonus has a low ‘line of sight’, meaning that the relationship between the performances delivered and the reward received is somewhat ambiguous and not transparent to non-contract parties (Voeten, 2011, p. 1). The Dutch Central Bank defines this bonus plan as bonus payment based on the manager’s discretion, without the fulfilment of any pre-determined performance targets (2009, p. 25). For the Dutch sample, at least 11,4% of the executives has received a discretionary bonus in 2007, while this percentage increased slightly to 12,6% in 2008 (DNB, 2009, p. 8). However, the actual percentages should be considerably higher than these figures, as there were no performance evaluations available for 30% of the sample observations. The report indicates that ‘lots of these observations’ have ex-post received a bonus.
Furthermore, Hay Group distinguishes other forms of variable compensation based on team performance and overall firm performance, for example team bonus, stock grants and stock option grants (Voeten, 2011, p. 1). As explained before, the incentives provided that are based on collective team or firm performance will not be considered in this study. For each empirical model used in the current study, the stock grants and stock option grants will be considered in the regression analyses because prior literature document that these compensation components are associated with future firm performance (Hall & Liebman, 1998).
2.2.3 Discretionary bonus payments
Gibbs et al. (2004) explain three ways in which subjectivity can occur in the assignment of bonuses:
(1) All or part of a bonus is based on subjective judgements about performance; (2) The weights on some or all objective measures are determined subjectively;
(3) Ex-post discretionary adjustments in a bonus that was initially based on objective measures.
13
The authors found that discretionary bonuses based on subjective assignments of rewards provide approximately 20% of the total compensation for their sample managers. The authors argue that these results are similar to those found in studies of top executives and middle managers (Gibbs, 2004, p. 433). Furthermore, subjective bonuses are substitutes to formula-based bonuses, as subjective bonuses are larger when formula-formula-based bonuses are not used. Particularly, subjective bonuses are used to mitigate perceived weaknesses in bonuses based on quantitative performance measures, when formula-based bonuses fail to adequately encourage investments with long-term impacts, when they fail to encourage cooperation and when performance is noisy due to the influence of other departments.
Consistent with the definition of discretionary bonuses in paragraph 2.2.2, the construct is consistent with (1), all or part of the bonus is based on subjective judgements about performance. Due to the empirical design, the other two forms of subjectivity in the assignment of bonuses will not explicitly be considered in the current study.
2.3 Formula-based incentives
2.3.1 Initial conceptual framework (Bonner & Sprinkle, 2002)
The conceptual framework that is proposed in the overview paper of Bonner and Sprinkle (2002), will be used to analyze the effects of performance-contingent incentives on the financial performance of the banking sector. This framework is presented in figure 2.
Figure 2. Conceptual framework for the effects of performance-contingent incentives on effort and
task performance (Bonner & Sprinkle, 2002, p. 304).
As this framework has been used extensively in numerous contexts by researchers in understanding how the features of accounting settings may affect the incentives-performance relationship, I have chosen it to be the theoretical framework of this empirical study. This paper reviews theories and provides evidence regarding the effects of performance-contingent
14
incentives on individual effort and task performance (Bonner & Sprinkle, 2002, p. 304). Thus, for this purpose, both the incentives-effort and effort-performance relationships are analysed and explained in the paper. The standard economics perspective underlies the positive directions for both relationships, as can be seen from figure 2 (2002, p. 305). The reasoning is that incentives typically motivate people and, as a result, will lead to increased effort and improved performance. This line of reasoning is consistent with the notion of the ‘homo economicus’ within the economic science. This notion sees humans as passive and rational, who tend to show utility-maximizing behaviour and actions (Lemmens, 2014). In the context of the conceptual framework, the homo economicus notion implies that economic actors will react to external, contingent incentives in a rational and predictable way. The behaviour of employees can thus be directed through the deployment of external rewards or sanctions. Bonner and Sprinkle (2002) offer different theories and variables, mostly in support of this economic perspective of the incentive-effort and effort-performance relationships. As this paper is the theoretical starting point of this study, both relationships will be explained concisely below.
Bonner and Sprinkle review four theories explaining the incentives-effort relationship: the expectancy theory, agency theory, goal-setting theory and self-efficacy theory (2002, p. 308). The first three theories assume the standard economics view while the latter theory takes cognitive mechanisms from the psychological science into account. Both expectancy and agency theory suggest that monetary incentives affect the utility of various outcomes and that effort affects the probability of achieving these outcomes. Incentives increase an individual’s desire to increase performance, which motivates the individual to exert costly effort (2002, p. 308). The goal-setting theory suggests that incentives cause people to set specific and challenging personal goals, which in turn lead to greater effort (otherwise the goals would not be obtained). In addition, incentives may result in higher goal commitment (2002, p. 309). Finally, the self-efficacy (or social-cognitive) theory proposes self-regulatory cognitive mechanisms whereby the motivation created by incentives leads to changes in effort. Self-efficacy, or an individual’s belief about whether he can execute the actions needed to attain the performance required, is an important determinant of effort. In particular, this relation is affected by several cognitive, motivational, affective and task mechanisms. Because self-efficacy affects many factors, the roles of incentives in the incentives-effort relationship are more numerous than those specified in the prior theories (2002, pp. 309-310).
Furthermore, Bonner and Sprinkle have distinguished four categories of factors that influence both the incentives-effort and effort-performance relations: person variables (skill), task variables (complexity), environmental variables (assigned goals) and incentive scheme variables (see figure 2).
2.3.2 Bonner et al. (2000)
Bonner et al. (2000) examine the effects of two interaction variables on the incentive-performance relationship, reviewing 131 experimental studies across a wide variety of disciplines. The first variable, task type, defined as the gap between task complexity and the individual’s skill, is found to interact negatively with the effort-performance relationship (p.19). Thus, as tasks become more complex (and the laboratory subject’s skills are held constant), performance becomes less sensitive to increases in effort. In some studies increased
15
effort led to lower task performance. In their extensive review, the authors ordered the types of tasks on their average complexity of information processing into five broad categories. Instead of rating each individual task on its level of complexity, the authors chose to rate the complexity of the five categories of tasks (p. 24). The types of tasks, in order of increasing complexity, are: (1) vigilance and detection, (2) memory, (3) production and simple clerical, (4) judgement and choice, and (5) problem solving, reasoning, game playing. Only the categories (4) and (5) are relevant for this study, as executives in the banking sector are responsible for making decisions in diverse complex situations involving a high level of problem-solving and judgement. In support of this statement, two job descriptions for banking executives are included in the figures 7 and 8. For the current study the interaction variable task type will be indirectly taken into account, as only banking executives will be included in the sample. Therefore, the effort-performance relationship is expected to be significantly negative.
The second variable from Bonner et al. (2000), type of incentive scheme, ranging from quota, piece or variable rate, tournament and flat rate, is found to interact with the incentive-effort relationship (p. 26). Incentive contracts using a quota rate are found to be most effective on the individual’s efforts while the flat rate is found to be least effective (p. 27). Variable bonuses are in essence quota schemes, as individuals typically receive a flat rate irrespective of performance until a certain targeted level of performance is attained. Above the targeted level, individuals sometimes receive a piece rate for each additional unit of output (p. 26). Flat rate schemes do not link compensation to performance on either an overall basis or an individual output basis. The explanation for the positive effects of quota and piece rate incentive schemes compared to flat-rate schemes lies in the goal-setting theory, as mentioned before. Specific and challenging goals increase motivation and effort, resulting in higher performance. It should be noted that the incentive scheme variable directly influences effort; performance is only indirectly influenced by the incentive scheme variable.
The conclusions of the review concerning the task type (5) are:
In 17%-21% of all studies involving problem-solving, monetary incentives resulted in higher performance on the task at hand (p. 32);
None of the incentive schemes performed well for the problem-solving tasks. Quota incentive schemes had positive effects 38% of the times, while for piece rate schemes this number amounts to 13% (p. 32).
2.3.3 The economic view of formula-based incentives
The economic view on the effect of incentives on effort can be illustrated by Newton’s first law of motion: ‗An object in motion will stay in motion, and an object at rest will stay at rest, unless acted on by an outside force‘ (Pink, 2009, p. 14).
Thus, the expectancy theory, agency theory and to a lesser extent the goal-setting theory are based on this standard economics view of incentives. Assuming the rational homo economicus will react to external incentives in a predictable way (= maximize his utility). The reasoning of this economic perspective can be further illustrated by Lazear (1999), who investigated the productivity effects associated with the switch from paying hourly wages to paying piece rate wages (thus from fixed to variable pay). This resulted in:
16
1. An increase in productivity amounts to anywhere from about 20% to 36% of output; 2. The firm attracted a better work force;
3. The variance in output across individuals at the firm rises as well.
In short, individuals will increase their working effort most when the incentive system links rewards as closely as possible to performance. However, the paper does not take the effort-performance relationship explicitly into consideration. The study assumes that an increase in effort will automatically result in higher productivity and performance. However, based on the psychology studies in paragraph 2.4.3, we will see that this assumption is not always valid.
2.4 Discretionary incentives
2.4.1 Economic theory (optimal incentive contracting)
The economic theory interprets discretionary bonus payments as the result of implicit incentive contracts, agreements that are not legally enforceable and thus non-verifiable to outside parties (Ederhof, 2010, p. 1921). According to Bull (1987), implicit contracts constitute Nash equilibria in a game-theoretic sense and therefore are sustained through economic mechanisms. Indeed, in a Nash equilibrium, none of the game players (agent and principal) has the incentive to unilaterally deviate, because both players chose the best response, given the action of the other player. Although the implicit contract agreements are non-verifiable to outside parties, the contract could be credibly enforced when there are repeated interactions between the contract parties (Bull, 1987, p. 149). The author proves the existence of an implicit contract that supports efficient trade in a simple model, using intra-firm reputations as an economic enforcement mechanism (p. 147).
Based on this notion of discretionary bonus payments, Hayes and Schaefer (2000) have introduced an approach to indirectly test whether the unobservable performance metrics underlying the bonus schemes are likely to be non-contractible in nature. The authors regress future financial performance on current cash compensation, while controlling for current financial performance in a first difference form. The current compensation measure is defined as the sum of salary and bonus. Alternatively, one can view this as regressing the unexplained variation in current compensation on the future variation in firm performance. The unexplained variation in current compensation is an indirect measure of discretionary bonus, as it is the residual term in the regression of current cash compensation on current, observable performance measures. The future variation in firm performance is measured by the change in RoEt+1 (p. 273). This empirical approach assumes the optimal incentive contracting theory, namely that corporate boards optimally use both observable and unobservable measures (to outsiders) of executive performance in their tasks of monitoring and rewarding executives. This theory implies that the current unobservable performance measures are correlated with future firm performance. The authors’ findings are consistent with this theory: the unexplained variation in current compensation is significantly positively associated with the variation in future firm performance. Thus, current compensation is incrementally informative about and predictive of future financial performance, after controlling for current financial performance. Additionally, Hayes and Schaefer found that the quality of observable performance measures, as specified by the variation in these measures, negatively influences
17
the positive effect of discretionary bonuses on future financial performance (p. 292). Thus, the results are consistent with the hypothesis that boards of directors use information that is not available to outside parties as part of an implicit incentive compensation contract, especially when the quality of observable measures is low.
Ederhof (2010) extends the aforementioned study by separating the current cash compensation in two components, namely a discretionary and formula-based bonus. The author uses an unique dataset, collected from the public SEC website, allowing the separation of bonus payments in these two components. The specification of the main empirical model is similar to Hayes and Schaefer (2000), except for the splitting the current cash compensation variable into its two components. Besides, future financial performance is measured in this study as the change in ROAt+1 (p. 1936). Ederhof finds empirical support for the notion that discretionary bonuses are paid based on non-contractible performance measures that are related to future financial performance, consistent to the optimal incentive contracting theory (p. 1941). In contrast, the formula-based bonus component is not incrementally informative about future financial performance, since this component is based on performance metrics such as accounting earnings, which tend to be backward looking.
2.4.2 Managerial power theory
The preceding paragraph is based on the hypothesis that discretionary bonus payments are an outcome of optimal incentive contracts. Alternatively, the managerial power theory interprets discretionary bonus payments as managerial rent extraction stemming from corporate governance weaknesses (Ederhof, 2010, p. 1928). Thus, the assignment of discretionary bonuses is not the result of an implicit incentive compensation contract. Rather, it is the outcome of executives exerting considerable power over the board of directors and the compensation committee, who are mainly responsible in setting executive compensation. Extending the logic of this theory, one would expect to a lesser extent a positive effect of discretionary bonus payments on future firm performance, as the current unobservable performance measures are less correlated with future firm performance due to the rent extraction efforts by executive officers. Thus, the managerial power variable will serve as a moderating variable in this study, interacting negatively in the relationship between discretionary bonus payments and future firm performance. In the relationship between formula-based bonus payments and future firm performance, managerial power theory is less applicable, as the contractible performance measures are agreed upon and ex-ante and the incentive targets are laid down in formal employment contracts. The prior literature on this subject uses various specifications of the managerial power variable to explain the level of CEO cash compensation, consisting of both formula-based and discretionary components. These papers will be summarized below, as they provide further insight into the managerial power theory.
Boyd (1994) uses a measure of board control in explaining the level of CEO cash compensation, which consists of salary and bonus. The author develops a conceptual model for measuring the construct of board control, consisting of five measures: CEO duality (CEO serves as chairman of the board), ratio of insiders on the board (ratio of inside directors to outside directors in the board), board stock ownership (percentage of common stock owned by the board of directors), number of directors representing ownership groups and level of
18
director compensation. The operationalisation of the board control variable is consistent with the agency theory, viewing the CEO as agent who strives to maximize his own self-interest in regards to compensation. This is a function of his ability to circumvent/minimize board control. Boyd found that CEO compensation is significantly negatively related to levels of board control, as measured by the five sub-measures outlined above. Thus, the CEO could command a larger salary when he dominates the board of directors. Besides, the author found that CEO compensation is not significantly driven by firm size or firm performance. This conclusion is inconsistent to Gabaix and Landier (see paragraph 2.2.1), who found that the sixfold increase of CEO pay in the U.S. between 1980 and 2003 can be fully attributed to firm size. Indeed, the market capitalizations of the sample companies during that period have increased sixfold (2006, p. 93). Also, the conclusion seems inconsistent with Hayes and Schaefer (see paragraph 2.4.1), who found that current CEO cash compensation is incrementally informative about and predictive of future financial performance, after controlling for current financial performance. Finally, it should be noted that Boyd (1994) used a total measure of CEO cash compensation, thus not examining the effect of managerial power (board control) on discretionary bonus in particular.
Core et al. (1999) confirm the findings of Boyd (1994), namely that measures of board and ownership structure for managerial power/board control explain a significant amount of variation in CEO compensation, after controlling for standard economic factors of pay. Thus, CEOs obtain excessive pay levels (above the level explained by standard economic factors) when corporate governance structures are less effective (p. 372). In addition, the authors examine whether the compensation predicted by the aforementioned governance measures is related to subsequent firm operating and stock market performance. This relationship is significantly negative, implying that firms with greater agency problems perform worse (p. 404). Summing up, prior empirical research has documented that measures of CEO power are associated with compensation patterns that can be interpreted as rent extraction.
Besides examining the informativeness of discretionary and formula-based bonuses on future financial performance, Ederhof (2010) studies whether discretionary bonus payments are more likely to occur in companies where management has considerable power, consistent to managerial power theory. The reasoning is that executives prefer compensation structures that enable the extraction of rents to be camouflaged as optimal contracting (pp. 1928-1929). Besides, the author clarifies that the incentive plans approved by shareholders typically allow for the payment of discretionary bonuses. Therefore, it is difficult for outsiders to determine whether a discretionary bonus is paid in accordance with the optimal incentive contract or whether it constitutes rent extraction (p. 1929). This reasoning results in the hypothesis that discretionary bonus payments may be an effective way for powerful executives to extract rents. Ederhof has not found results which support the hypothesis that discretionary bonuses are positively related to executive power (p. 1941), inconsistent with the managerial power theory. However, the author acknowledges (p. 1941; p. 1929) that this finding could be partially attributed to the low statistical power generated by the relatively small sample (the discretionary sample consists of 234 firm-years). Also, possibly omitted variables could bias this result (p. 1941).
19
2.4.3 Cognitive psychology
Psychology is the science of the human mind (attitudes, cognition, motivation) and human behaviour (actions, communications). Psychology differs from other social science theories which are frequently used in management accounting research in focusing on individual rather than organizational behaviour, and on subjective phenomena (Birnberg et al., 2006, p. 113). The assumptions about human behaviour are based on a more realistic and empirically established foundation (Weibel, Rost & Osterloh, 2007, p. 5). Psychology theory has been used to study management accounting practice for over 50 years, relying primarily on theories from three subfields – cognitive, motivation and social psychology. Cognitive psychology is the study of processes that influence human thinking, including attention, knowledge, judgements, decisions and learning (Birnberg et al., 2006, p. 114). Motivation psychology examines processes that influence behaviour – arousal, direction, intensity and persistence of effort. Eventually, social psychology is concerned with how other people influence individuals’ minds and behaviour. The use of different psychological theories from these three subfields has proven useful in understanding the effects of management accounting practices on individuals’ minds and behaviour (pp. 114-115).
In particular, the motivational effects of management accounting practices depend not only on how these practices influence objectively measured rewards but also how they influence individuals’ mental representations of these rewards through psychological processes (Birnberg et al., 2006, p. 114). For this study, the possible motivational effect of formula-based and discretionary incentives need to be indirectly and implicitly taken into consideration, so that alternative explanations could be provided for the effects on future firm performance. The psychological experiment conducted by Lepper et al. (1973) is relevant in this context, offering an alternative view on the effect of formula-based and discretionary bonus payments on current and future performance – complementary to the optimal incentive contracting theory in paragraph 2.4.1.
The authors tested the ‘overjustification hypothesis’ in a field experiment, which is rooted in the self-perception theory from the cognitive subfield. According to this theory, people have no direct knowledge of their motives, but infer them from their actions and external contingencies (self-attribution processes). Thus, people engage in post-behavioural inference processes about their own behaviour and its meaning, in order to self-justify or explain their behaviour (p. 129). When the external contingencies controlling his behaviour are salient, unambiguous and sufficient to explain it, on the one hand, the person will attribute his behaviour to these external circumstances (in our context: the formula-based incentive). On the other hand, if the external contingencies are not perceived, unclear or psychologically insufficient to account for his actions, the person attributes his behaviour to his own dispositions, interests and desires (p. 129).
The overjustification hypothesis is an implication of the self-perception theory and means that a person’s intrinsic interest in an activity may be undermined by inducing him to engage in that activity as an explicit means to some explicit goal (p. 130). This occurrence of this hypothesis depends on the level of the external contingencies. If the external justification provided to the person to engage in the activity is ‘unnecessarily high and psychologically oversufficient’, the person might eventually infer that his actions were mainly motivated by the external contingencies of the situation. In other words, the person induced to undertake an
20
activity as a means to some ulterior end will see the external contingencies as an end themselves (p. 130).
Furthermore, the authors argue that contracting explicitly to engage in an activity for a reward should undermine interest in the activity while receipt of an unforeseen/unexpected reward after engaging in an activity should have no detrimental effect on the intrinsic interest, even when the reward is a highly prized material one (p. 130). The authors tested these assertions on children in an educational setting and found empirical evidence for the overjustification hypothesis for the expected, contractible reward. The children were randomly divided into three groups: the expected-reward group, unexpected-reward group and no-reward group. Each group is divided into two subgroups, based on initial observations of the children’s interest in the drawing activity. Prior to the experiment, the researchers showed the expected-award group a personal certificate and asked the children to draw in order to receive the award. The second group was handed a certificate for drawing without prior knowledge of a reward. The third group was not promised a certificate for drawing nor received one in the end.
For the expected-reward group, their interest (alternatively: effort) during the experiment was the highest; however, the quality of their drawings was rated the lowest by three blind judges. In addition, two weeks after the experiment, the group’s interest and effort in the drawing activity have decreased significantly. This is consistent with the overjustification hypothesis These findings are significant based on a alpha of 1% and are similar for children with low and high prior interest in the drawing activity. For the unexpected-reward group, the children’s interest during the experiment was middle while their performance were the highest, rated by the blind judges. The subgroup with low initial interest in the drawing activity increased the interest and effort in the activity after the experiment. These findings are significant based on an alpha of 1%.
However, for the subgroup with high initial interest, no significant effect of the unexpected reward is found for the subsequent interest/effort. The authors explain this finding by clarifying that the children’s initial interest was already high, so there is no reason to expect the unexpected reward to have any impact on their subsequent interest (Lepper et al., 1973, p. 135). I aimed to summarize the findings of this paper in table 1.
Table 1. Main conclusions from Lepper et al. (1973).
Subsample groups Effort during
experiment
Performance during experiment
Natural setting post-experiment 1A Expected reward group, low initial
interest
Highest Lowest* (sign. level = 1%) Lower interest, lowest efforts*
1B Expected reward group, high initial interest
Highest Lowest* (sign. level = 1%) Lower interest, lowest efforts*
2A Unexpected reward group, low initial interest
Middle Highest* (sign. level = 1%) Higher interest, higher efforts*
2B Unexpected reward group, high initial interest
Middle Highest* (sign. level = 1%) Lower interest, lower efforts
3A No reward group, low initial interest
Lowest Middle Higher interest,
higher effort
3B No reward group, high initial interest
Lowest Middle Higher interest,
21
The authors conclude the paper by stating that their study provides evidence of an undesirable consequence of the unnecessary use of extrinsic rewards, supporting the case for the exercise of discretion in their application (p. 136). This conclusion implies that each study of monetary rewards should specify rewards in the expected and unexpected component. For the purposes of this study, formula-based bonus payment could be viewed as an expected reward, while discretionary bonus payment can be viewed as an unexpected reward, as the bonus payment is not known until after performing the activities at hand.
Kunz and Pfaff (2002) reviewed empirical evidence and identified that under certain specific conditions, high intrinsic motivation undermines agency theory predictions related to performance pay in corporations. However, these conditions are special and easily avoidable in real life (p. 275). Also, the experimental findings cannot be interpreted as being contradictory to agency theory (p. 291). Therefore, while concern with intrinsic-extrinsic motivation did not seem promising in understanding agency theory predictions, the authors recommend that agency theory could include consideration of implicit contracts, self-perception and social interactions, fairness and reciprocity and social norms (p. 292).
Ariely et al. (2005) studied the effect of excessive performance-contingent rewards on motivation, efforts and the task performance of rural workers in India. The abstract of this paper indicates that psychological research suggests that excessive rewards can produce supra-optimal motivation, resulting in performance declines (p. 1). To test this finding, the authors conducted a set of experiments in rural India, where a relatively low reward for the Western world is considered high. The 87 subjects in this experiment, recruited from a rural town in India, worked on different tasks and received performance-contingent rewards that were either relatively small, moderate or very large (compared to the typical levels of pay in India). The subjects would only receive the very large reward if they reached the ‘very good’ performance level for each game at hand. There were three experiments comprising different games which mainly required creativity, concentration, motor skills or problem solving (pp. 6-7). The authors conclude that in eight of the nine tasks across the three experiments, higher incentives led to worse performance (although the motivation increased). Comparing the ‘moderate reward group’ and the ‘very large reward group’, performance in the latter group was significantly lower under different robustness tests (p. 19). The differences between the ‘low reward group’ and the ‘moderate reward group’ were less significant and less robust. As the authors conclude:
―Many existing institutions provide very large incentives for exactly the types of tasks we used here – those that require creativity, problem solving and concentration. Our results challenge the assumption that increases in motivation necessarily lead to improvements in performance. […] The prevalence of very high incentives contingent on performance in many economic settings raises questions about whether administrators base their decisions on empirically derived knowledge of the impact of incentives or whether they are simply assuming that incentives enhance performance.‖ (Ariely et al., 2005, pp. 19-20)
These findings have been replicated and confirmed in various other papers from both the economic and psychological fields (Cameron, Banko & Pierce, 2001; Deci, Ryan & Koestner, 2001; James Jr., 2005; Jenkins Jr. et al, 1998; Gneezy & Rustichini, 2000; Murdock, 2002; Cheng, Subramanyam & Zhang, 2005). There are some similarities in the task characteristics for which monetary incentives can have detrimental effects, which require
22
mostly the right-side of the human brain. These task characteristics are summarized by Pink as follows:
non-routine, conceptual, requiring problem-solving, drawing unusual connections between elements and highly competitive tasks (Pink, 2009, p. 21).
Using parts of two current vacancies below, I aim to indicate that these task characteristics correspond to the top executives’ main responsibilities in the banking sector, besides strict requirements on the bankers’ analytical and interpersonal skills. The first quotation is a selected part of the job description of a senior strategy manager at the Royal Bank of Scotland in London. The second quotation is a selected part of the job description for the executive director of Norges Bank Investment Management (NBIM). The NBIM is a division of the Norwegian central bank.
―The role we are offering – In this role, you‘ll have regular engagement with the Bank and Strategy leadership teams driving problem-solving in the delivery of a varied portfolio of projects. You‘ll plan and allocate team resources to meet project requirement and delivery timelines. You‘ll lead team interactions with individual customer-facing businesses relevant to the development of their strategic plans and provide challenges to provide the quality of thinking and output. You‘ll engage with Finance and other functions regularly as part of your role in delivering strategic plans of customer-facing businesses and the overall Bank, and building consensus with key business leader to execute transformational changes to the business.‖ (Royal Bank of Scotland, 2014)
―Principal duties – The position of Executive Director and Chief Executive Officer of NBIM entails responsibility for:
Development of an investment strategy to achieve the highest possible return for the funds within the investment constraints laid down
Management of the above funds within the constraints laid down for their management in the investment mandates, rules and guidelines laid down by the Executive Board of Norges Bank and the funds‘ owners
Cost-effective operation of NBIM and preparation of budgets and accounts.‖ (Norges Bank, 2008)
Bank executives worldwide work in extremely competitive environments, wherein the daily work requires high levels of problem-solving capacities in the management of different project portfolio’s and in the development of firm-wide strategies under various constraints. The requirements on problem-solving capabilities, the competitive environment and the continuous internal and external pressures characterize the daily work of banking executives in comparison with executive officers from other industries. Different psychological studies emphasize the destructive effects of mental and physical load, caused both by the individual’s environment and the increasing compensation incentives, on the individual’s task performance (Easterbrook, 1959; Baumeister, 1984; Baumeister & Showers, 1986; Wegner, Ansfield & Pilloff, 1998).
23
2.5 Incentives in the recent financial crisis 2.5.1 Motivation for the research setting
The existing literature from the cognitive psychology field, outlined in paragraph 2.4.3, motivate the choice of the current research to study cash incentives in two separate components, namely formula-based and discretionary bonus. According to this literature, and the prior studies from economic scholars, the effects of cash incentives on firm performance will differ depending on the bonus component. In addition, the literature reviewed in paragraph 2.4.3 imply that the largest effects will be found in the banking sector, given (1) the nature of the executive officer’s work (non-routine, conceptual, problem-solving), (2) the oppressive environment both internally and externally, and (3) the sixfold increase of executive compensation in the U.S. between 1980 and 2003, resulting in higher pressures for the executives. In 2015, the New York state Controller estimated the development of the New York securities industry bonus pool from 1986 to 2014 for employees who work in the New York city. The estimates reflect solely the cash payments and deferred compensation, thus excluding stock options (DiNapoli, 2015). The bonus pool development shows that the average bonus for the employees in the banking industry was the highest in 2006, namely $191,360. In addition, the total bonus pool in 2006 amounts to $34,3 billion. In the financial crisis year 2007 and 2008, both numbers decreased substantially. For 2008, both the average bonus and the total bonus pool for New York banking employees decreased by 43-47% compared to the 2007’s numbers. These numbers motivate the current study to investigate executive compensation structures for the U.S. banking industry in 2006, as the compensation practices for the New York city are likely to be similar for other cities in the U.S.
2.5.2 Prior literature
In this paragraph, two recent studies are outlined which concern the effects of executive incentives on the U.S. banking performance during the financial crisis. These studies do not explicitly investigate cash incentives, nor do they separate these incentives into bonus components.
Fahlenbrach and Stulz (2011) studied the effects of CEO incentives through stock holdings, stock options and cash bonuses on the performance of U.S. banks in the credit crisis. They include 98 banks in their sample, that are obtained from Execucomp using the SIC code 6000-6299 (p. 25). They conclude that a bank’s stock return performance and return on equity in 2007-2008 are significant negatively related to the CEO incentives from stock holdings in 2006. In addition, the authors note that the CEOs’ incentives were aligned with the shareholders’ interests, as they possess large amounts of stock holdings and stock options prior to the crisis. In other words, the CEOs in the sample were in fact in part shareholders themselves prior to the crisis, because they shared the firm’s results and risks (as the volatility of the firm’s stock price influences the value of the CEOs’ option holdings). Because these CEOs were also affected by the worsened bank’s performance, and because the CEOs have not changed their stock and option holdings before the crisis, the authors conclude that the CEOs did not focus knowingly and sub-optimally on the short term (p. 13). Therefore, banks with CEOs whose incentives were better aligned with shareholder interests, performed seemingly worse during the credit crisis (p. 11). The authors explain this by stating that CEOs with better incentives to maximize shareholder wealth took risks that other CEOs did not.
Ex-24
post, these risks had unexpected poor outcomes (p. 25). This is referred to as the Unforeseen Risk Hypothesis, namely the poor financial performance of banks during the crisis was the result of unforeseen risk of the bank’s investment and trading strategy (Bhagat & Bolton, 2014, p. 314). Concerning the cash bonuses awarded to the CEO’s, neither cash bonus nor stock options had any impact on bank performance during the crisis. Finally, separating the sample into banks which received TARP-funds and No-TARP banks (see below), the relation between bank performance and CEO incentives does not differ significantly among the subsamples.
Bhagat and Bolton (2014) build upon the prior study by using nearly the same sample of firms (100 American banks), but distinguishing three different subsamples. 14 system banks are required by the U.S. Treasury to participate firstly in the Troubled Asset Relief Programme (TARP) in October 2008, because of their role in the U.S. financial crisis. These banks are the so-called ‘Too-Big-To-Fail’ (TBTF) banks. The CEO compensation structures of these TBTF-banks during 2000-2008 are studied and compared to that of CEO’s of 37 non-TARP banks. These banks neither sought nor received non-TARP funds from the U.S. Treasury. Besides, the CEO compensation structures of the TBTF-banks during 2000-2008 are compared to that of CEO’s of 49 later-TARP banks. These banks received TARP funding several months after the TBTF-banks (p. 319). Using three measures of risk-taking (asset write-downs, bank’s Z-score, amount of borrowed capital from Fed bailout programs), the authors find that CEOs in the TBTF banks engaged in significantly more discretionary stock sales than CEOs in the No-TARP and later-TARP banks (p. 314). This result implies that incentives generated by CEO compensation led to excessive risk-taking by banking executives, resulting in the financial crisis (Managerial Incentives Hypothesis). Thus, banking executives focused on their own short-term gains at the expense of long-term shareholder wealth.
This paper provides empirical evidence for the assertions of Daniel Pink and the Dutch Central Bank in the introduction. However, the authors do not find support for the prior study (Fahlenbrach & Stulz, 2011), that the poor performance of banks during the crisis was the result of unforeseen risk. Based on their findings, Bhagat and Bolton recommend the banking executives compensation to solely consist of restricted stock and stock options. Besides, the equity/book value ratio should contribute to a minimum of 25% (p. 314).
2.5.3 Current study
This study uses a larger sample than in the studies outlined above in order to account for the bias towards larger banks in these studies. As explained in the introduction, none of the existing research on formula-based and discretionary bonuses has come to my attention that has used the financial crisis setting to study the effects on financial performance. By exploiting this gap in the prior literature, I aim to contribute to the existing knowledge on the subjects of variable executive compensation and the recent financial crisis. It should be noted that the use of this sample requires further data collection, as information on formula-based and discretionary components of the bonus payments are not readily available in the databases. A direct estimation method will be used to obtain data on discretionary bonus and formula-based bonus payments, as existing public information concerning bonus payments of
