The circumstances under which firms are inclined to more/less frequently use certain performance measures : a focus on financial, subjective and relative performance measures

(1)

1

The circumstances under which firms are

inclined to more/less frequently use certain

performance measures

A focus on financial, subjective and relative performance measures

Tim Schavemaker

10408789

28th of June 2015, version 2

BSc Accountancy & Control, ABS, UVA

Supervisor: H. Kloosterman MSc

(2)

2

Hierbij verklaar ik, Tim Schavemaker, dat ik deze scriptie zelf geschreven heb en dat ik de volledige verantwoordelijkheid op me neem voor de inhoud ervan.

Ik bevestig dat de tekst en het werk dat in deze scriptie gepresenteerd wordt origineel is en dat ik geen gebruik heb gemaakt van andere bronnen dan die welke in de tekst en in de referenties worden genoemd.

De Faculteit Economie en Bedrijfskunde is alleen verantwoordelijk voor de begeleiding tot het inleveren van de scriptie, niet voor de inhoud.

(3)

3

Abstract

This thesis is on the subject of firms' likeliness to more or less frequently use certain

performance measures under certain specific circumstances. The study assesses the positivists line of research, a conflict of interests between shareholders and dishonest managers, in studying the agency problem. The thesis will focus on three different sorts of performance measures to minimize the shareholder-manager problem, namely financial, subjective and relative performance measures. With the help of a literature review, an answer will be provided to the question: when are firms inclined to more/less use certain performance measures? The study contributes to literature by providing an unprecedented overview of 20 situations in which firms are likely to use a particular performance measurement. More (empirical) research is suggested with respect to the specific circumstances given in the overview in relation to the other two, not applied, performance measures.

Samenvatting

Deze scriptie gaat over de waarschijnlijkheid van bedrijven om bepaalde prestatiemeters meer of minder vaak te gebruiken onder bepaalde specifieke omstandigheden. De studie gaat uit van de denkbeelden van de positivisten, een belangenverstrengeling tussen aandeelhouders en onoprechte managers, in het bestuderen van het "agency" probleem. De scriptie zal zich focussen op drie verschillende soorten van prestatiemeters om het aandeelhouder-manager probleem te minimaliseren, namelijk financiële, subjectieve en relatieve prestatiemeters. Met behulp van een literatuuronderzoek zal antwoord gegeven worden op de vraag: wanneer zijn bedrijven geneigd bepaalde prestatiemeters meer/minder vaak te gebruiken? De studie draagt bij aan de literatuur door het verschaffen van een niet eerder vertoond overzicht van 20 situaties waarin bedrijven geneigd zijn een zekere prestatiemeter te gebruiken. Meer

(empirisch) onderzoek wordt voorgesteld naar de specifieke omstandigheden in het overzicht in relatie tot de andere twee, niet toegepaste, prestatiemeters.

(4)

4

List of tables

Table 3.5.1...25

(5)

5

1 Introduction

One of the oldest and most common codified modes of social interaction is the relationship of agency. An agency relationship arises when one party, the agent, acts on behalf of the other, the principal, in a certain domain of decision problems (Ross, 1973). However, because the principal and the agent both have different incentives and will try to maximize their own needs and because it is difficult and expensive to monitor an agent's actions, the principal encounters difficulties in verifying if the agent has behaved appropriately. This problem also rises in the relation between shareholders and managers.

To give a manager the incentive to accomplish the same goal as the shareholders, the manager can be given a specific performance threshold. The manager has to reach the

threshold to obtain some kind of benefit. Choosing which type of performance measure to use in an incentive contract is a central problem in agency theory (Baker, 1992). There are several ways to set the threshold and to measure the performance of the manager with respect to the threshold. Three main categories of performance measures can be distinguished; financial performance measures, subjective performance measures and relative performance measures.

According to Bushman and Smith (2001), the most popular topic of governance accounting research concerns the role of financial accounting information in managerial incentive contracts. They state: "the heavy emphasis on managerial compensation derives from...the success of principal–agent models in supplying testable predictions of the relations between available performance measures and optimal compensation contract" (Bushman & Smith, 2001, p. 238). Because it is an derivative of the heavy emphasis on managerial compensation, using the right performance measures is important. However, it is not always clear which performance measure to use in which circumstances and this makes the decision regarding performance measure harder for firms.

(6)

6

The purpose of this study is to give a clear overview of what a firm's used performance measures are based on different circumstances of the firm. This overview will contribute to the literature by clarifying what is to be expected under different circumstances; to use a performance measure more or less frequently.

There is sufficient literature to be found about the agency problem and about different performance measures and their advantages and disadvantages, including some empirical evidence. The effects of different performance measures are given with respect to particular variables and the measures are described under certain circumstances. However, no clear overview has been given about when firms are more likely to use a certain performance measure. This study contributes to literature by giving a clear overview of which performance measures are more to be used in certain situations and by giving a better understanding as for why a performance measure is best used more or is not in this situation. This study is based on a literature review.

The research question of this study is as follows.

"Under which circumstances are firms inclined to more or less frequently use certain performance measures?"

The study will focus on three relatively different sorts of performance measures; the

traditional and often used financial performance measures, the totally subjective performance measures and the more recent relative performance measures. I find 20 situations in which shareholders are inclined to more/less frequently use one of the three performance measures.

The study is constructed as follows: chapter 2 explains the basic assumptions

(7)

7

extensive overview of the (dis)advantages of the three studied performance measures. Next, chapter 4 provides the analytical part of this study, subdivided into paragraphs based on the three measures. This is where I try to find answers to the research question. Then, chapter 5 shows an overview and discussion of the findings in chapter 4. Finally, I will provide the conclusion in chapter 6.

2 Agency theory

I use agency theory as the basis for this study. The implications of the agency theory indicate some problems for shareholders in verifying the performance of managers. The implications of the agency theory and their problems will be discussed below.

Ross (1973, p. 134) suggests that "an agency relationship has arisen between two (or more) parties when one, designated as the agent, acts for, on behalf of, or as representative for the other, designated the principal, in a particular domain of decision problems". According to Eisenhardt (1989), two problems can occur in agency relationships. These problems are the agency problem and the problem of risk sharing. First, the agency problem arises when the principal and the agent have different desires and goals and when it is difficult or expensive for the principal to monitor the actions of the agent. This makes it difficult for the principal to verify if the agent has behaved appropriately. Shareholders wish the firm to have an optimal profit because their dividend is likely to rise then and their shares are likely to rise in value. However, managers are supposed to be solely self-interested and therefore only thinking about an as high as possible compensation with the least effort possible.

Shareholders want the firm to measure the performance of managers to give managers the incentive to handle in the interest of the firm. The problem of risk sharing is the second problem mentioned. The principal and the agent prefer different actions because of their

(8)

8

different risk preferences. They do not have the same attitude towards risk. Shareholders prefer less risk because they can lose their money if something goes wrong. Managers have less an incentive to care about the risk of losing money because they cannot lose their own money. An incentive contract is able to provide managers with an incentive to lower risk. In this study I will mainly focus on the first problem; the agency problem.

Agency theory has developed along two lines. A stream of positivists and a stream of principal-agents developed with time (Jensen, 1983). Both lines share the same unit of analysis: the contractual relation between the principal and the agent. The streams also share common assumptions about factors like people, information and organization (Eisenhardt, 1989). However, the two streams differ in their mathematical rigor, dependent variable and style. The distinction that is made that is most related to this study is the distinction in the parties focused upon. The focus of positivist researchers has been on identifying situations of conflicting goals between the principal and the agent. Thereafter, they describe the agent's self-serving behaviour and the governance mechanisms that limit it. Also, the positivists have almost exclusively narrowed their focus on one explicit case, namely the case of the principal-agent relationship between owners (shareholders) and managers of large, public corporations. Bushman and Smith (2001) describe this problem as the classic agency perspective that the separation of corporate managers from outside investors involves an inherent conflict. A more general theory of the principal-agent relationship is under the interest of principal-agent researchers. This more general theory can be applied in agency relationships like employer-employee, lawyer-client, buyer-supplier. This theory is in comparison with the positivist theory less accessible, because it is more abstract and mathematical. In this study, the focus will be on the positivist look at agency theory. This is because this stream is more accessible and narrow in its view. Using the principal-agent researchers theory would broaden the scope

(9)

9

of this study too much. Also, the problem in the shareholder-manager relation occurs repeatedly and often making it interesting to study.

I will now go more into detail about the specific relation between the shareholder and the manager. In this study we will use the term managers for the big decision making officers in large (publicly traded) companies. Firms are known for giving their managers different titles, but managers included are for example chief executive officers (CEO), chief financial officers (CFO), chief operating officer (COO), presidents, senior vice presidents and

executive vice presidents. As stated earlier, different desires and goals exist between the two groups of shareholders and managers (Eisenhardt, 1989). Because of this divergent interests, conflicting situations can arise; as a result of their self-interest managers want the highest pay-off they can get but with the littlest effort possible. Shareholders on the other hand want to maximise firm value and profit to generate dividend pay-outs. The divergent interests are not necessarily in agreement with each other. When they do conflict, the shareholders should intervene to verify what the managers are doing. Only when the different goals of the manager and shareholder are aligned, will the problem within the company be solved.

Based on the information provided in this chapter, agency theory appears to give a good view of the problems that arise in large, public organisations and the theory will therefore be applied as the basis of the writing of this thesis. In this study, I will look at how performance measures can be used to solve the agency problem (the problem of the difference in interests between principal and agent). Furthermore, this study relates to the positivists research stream, which focuses on the relation between shareholders and managers, rather than the principal-agents research stream. The conflicting goals between shareholders and managers mean that it is important to develop a system that aligns the goals of managers in the same direction as shareholders. The question that follows, is which incentives or measures the shareholders should use and of when to use which measure. In the following chapter, an

(10)

10

overview will be given of the different options for shareholders in relation to performance measures and their (dis)advantages.

3 The (dis)advantages of the performance measures

3.1 The performance measures

To minimize the shareholder-manager problem, the manager can be given certain incentives to align the goals of the shareholder and the manager in the same direction. One of these incentives is creating a performance threshold. A threshold has to be reached by the manager to obtain a certain benefit. The threshold is placed on a performance measure. According to Indjejikian (1999), performance measures are important because the observation of those performance measures provides useful knowledge in determining which decisions the manager took rather than because shareholders inherently value customer satisfaction or higher returns. Therefore, he thinks, in contrast to other literature, the relevant question is: what combination of measures adequately reflects the manager's contribution to firm value? In this study, I suppose manager's contribution to firm value equally relevant as, for example, customer satisfaction and higher returns. Employees (managers) are, by relating pay to performance, motivated to exert more effort in order to increase pay through the improved performance (Hölmstrom, 1979). The performance measures used determine the motivational effects of incentive contracts. This is, according to Moers (2005), because employees focus their attention to the measured parts of their work. Therefore the choice of performance measures is crucial in giving the right incentives.

There are several ways to measure performance and set the performance threshold. In this study I distinguish three different groups of performance measures; financial performance

(11)

11

measures, subjective (discrete) performance measures and relative performance measures. In the next three paragraphs I will explain the three performance measures and give their (dis)advantages.

3.2 Financial performance measures

Financial performance measures are extensively used in compensation contracts of executives (Bushman & Smith, 2001). The traditional financial performance measures measure the effort of a manager with accounting data. This quantitative data can be found in the accounting records of the firm. Financial performance measures are often seen as an objective measurement tool. There is a large and diverse pool of choices of financial performance measures that a firm can use to measure the performance of their managers. I divide the financial performance measures in the following groups: profitability, liquidity and investment related measures. I will now give a short explanation of the three groups of financial performance measures.

First, profitability is the way a firm uses its assets to gain profits (Ding, Lebas & Stolowy, 2013). Several performance measures exist to provide sufficient data to measure performance since just a net income figure is usually not sufficient enough (Marshall,

McManus & Viele, 2010). Two very straightforward financial performance measures are the gross profit margin and the net profit margin. The margins state the income per euro of revenue. Other profitability measures take assets into account. Asset turnover measures how the firm generates revenues with available assets. Return on assets measures how effectively the firm is able to generate net income using its assets. Some argue that measuring return only over a firm's equity, cash that owners have brought into the firm plus retained earnings

(12)

12

accumulated over the operational life of the firm, generates a better picture of their

effectiveness and efficiency (Marshall et al., 2010). This measure is called return on equity. Second, liquidity measures show the ability of a firm to fulfil its liabilities as they become due (Marshall et al. 2010). Liquidity measures can provide shareholders with an indication of the risk aversions of managers. Firm's liquidity not being that high, can indicate that the manager takes more risks. The liquidity performance measures can be globally divided into short and long term measures. Cash ratio, quick ratio and current ratio give, in varying degrees, an indication of the capability of a firm to pay-off its liabilities, due within one year, with liquidities that are immediately available if the firm went bankrupt(Ding et al. 2013). An example of a widely used long-term liquidity performance measure is the debt ratio. It gives an insight into the ability of a firm to pay its long-term debt. Debt to equity ratio can also be used as a financial performance measure.

Finally, investment measures should clarify if managers have been acting in the best interest of the shareholders. As stated earlier, in this study we will mainly look at publicly traded companies. This means they are publicly traded on a stock exchange and current and future, potential shareholders wish to make a profit on their investment. The most used investment related measure firms use to measure managers is earnings per share (EPS); income earned per outstanding share (Marshall et al. 2010). Next to the EPS, price earnings ratio, market to book ratio, dividend yield ratio and dividend pay-out ratio are used as investment performance measures (Ding et al. 2013).

Bushman and Smith (2001) offer some impressions of the existing literature on managers' incentive contracts and accounting based measures. Accounting based measures can be seen as financial measures, because managers are rewarded based on the information in the accounting data, which is financial data. This line of existing literature is in a mature stage of development (Bushman & Smith, 2001). It shows that financial accounting measures

(13)

13

are widely used in managers' incentive contracts. Especially profitability measures are extensively used. They state that the literature provides evidence for extensive use of the profitability measures in annual incentive plans as in the longer term incentive plans. The widespread use of financial performance measures for the compensation of high officers is supported by a positive statistical correlation between the use of profitability measures and different measures of offices' compensation. This correlation is robust according to Bushman and Smith (2001).

3.2.1 Advantages of financial performance measures

First of all, I will give some advantages of the use of financial performance measures. First of all, financial performance measures are easy and relatively cheap to calculate and interpreted. Next to this, short-term performance will be boosted upwards when using financial

performance measures. Furthermore, it is the most aggregate performance measures because it shows the consequences of managers ultimately.

Using financial performance measures is a very objective way of measuring performance (Ittner, Larcker & Meyer, 2003). The quantitative financial data presented to shareholders is easy to calculate and to interpret. This saves some of the firms' time and money. Indjejikian (1999) calls this advantage of the use of accounting data in performance measurement understandability. According to most studies, short term performance increases when a firm uses financial performance measures. This is a logical result of managers' desires for an as high as possible personal payoff in form of a bonus payment. The managers are so motivated and will act to improve their performance in order to receive a bonus payment (Gerhart & Milkovich, 1990). So, increased short term performance is an advantage of using financial performance measures.

(14)

14

According to Moers (2006) comprehensive financial (accounting) performance measures are the most aggregate performance measures. They are more aggregate because they reflect the consequences of all decisions ultimately. Namely, all the decisions a manager takes flow through the financial statements. Aggregate performance measures provide

principals with an advantage; when the compensation of an agent depends on an aggregate measure, principals are allowed to constrain the agent's actions to those in the interest of the principal (Prendergast, 2002). A distinction can be made between aggregate performance measures and specific performance measures. While aggregate performance measures provide (some) information about all decisions of managers, specific performance measures provide (some) information about a subset of decisions. The degree to which this aggregation can be exploited depends however on contracting costs of financial performance measures. It is less expensive to contract based on financial performance measures the more precise, verifiable and sensitive financial performance measures are compared to other, nonfinancial measures. Relative and other nonfinancial performance measures can, to some extent, also be aggregate performance measures. Moers (2006) gives an example for both performance measures. First, when using market-share (relative performance measure) as a performance measure,

information about all the manager's customer acquisition decisions is obtained. Second, when using product defect rates (nonfinancial performance measure), information about all of the manager's quality improvement activities is obtained. The problem is that neither

measurement is able to reflect the full consequences of these actions; costs of acquiring market-share or reducing defect rates or their implications for revenues are not included in the measurements. Most importantly, the measures do not include decisions other than market-share acquisition and quality improvements. The described performance measures provide (some) information about a specific subset and are so specific performance measures.

(15)

15

3.2.2 disadvantages of financial performance measures

On the other hand, there are two disadvantages to be found about the use of financial performance measures; these measures do not encourage long-term thinking which impacts stock returns and these measures increase noise which leads to an outcome of the performance measurement that is not fully a consequence of the manager's actions.

Financial performance measures have a focus on the short-term rather than the long-term because they only measure performance over the previous year. Higher performance will be encouraged when using financial performance measures. A significant correlation between the short-term firm performance and the yearly bonus payments proves this statement.

Inevitably, this kind of performance measures do not encourage long-term thinking and so the term thinking within the firm will be undermined (Gerhart & Milkovich, 1990). If long-term thinking is undermined, long-long-term investments will be undermined (Ittner et al., 2003). This is a disadvantage for shareholders because the impact of long-term investment strategy and long-term opportunities on stock returns is significant (Bushman & Smith, 2001). Ittner et al (2003) state that there is sufficient literature to be found that all came to the conclusion that the focus of financial performance measures it too much on the short-term profits of firms.

Furthermore, incentive contracts should substitute away from financial performance measures, according to a substantial theory-based empirical research that studies the trade-off between risk and incentives in relation to general principal-agent models (Bushman & Smith, 2001). The study finds evidence of an increased noise in financial performance measures relative to alternative performance measures. Noise is the influence of other factors than the manager's effort on the financial reports. Increased noise leads to an outcome of the

performance measurement that is not fully a consequence of the manager's actions. So, shareholders are not able to fully control the manager's performance. A possible explanation

(16)

16

of the increased noise is that it is to aggregate and therefore involves irrelevant factors. In totality, the findings are mixed. There is evidence for a shift but others do not find this evidence. Bushman and Smith (2001) also find literature that suggests that firms substitute away from financial performance measures toward alternative performance measures when growth opportunities of the firm increase and that the intensity with which earnings are impounded into stock price determines how much the incentive weight on earnings increases.

Altogether, it can be concluded that the use of financial performance measures comes with three advantages and two disadvantages. The fact they are easy and relatively cheap to calculate and interpreted can be seen as an advantage. Next to this, short-term performance will be boosted upwards when using financial performance measures. Furthermore, this kind of performance measures are the most aggregate performance measures because they show the consequences of managers ultimately. On the other hand financial performance measures do not encourage long-term thinking because of their focus on the short run. Also, evidence can be found in literature for an increased noise in the financial measures compared to other performance measures.

In the literature important trends can be found about the use of financial measures over the last three decades (1970-2000) in the United States (Bushman & Smith, 2001). Their statistical data suggests that profitability accounting measures have become less important. This is because of a trend of firms shifting more from accounting profitability measures toward alternative approaches when compensating their managers.

3.3 Subjective performance measures

An alternative approach of measuring a manager's performance is subjective performance measurement. Moers (2005) identifies subjective performance measures (e.g. using discretion

(17)

17

in performance measures) as "superior's subjective judgements about qualitative performance indicators". In incentive contracts, subjective performance measures can be used in roughly three different ways (Gibbs, Merchant, Van der Stede & Vargus, 2004). First of all, the whole or a part of the bonus depends on discretion in the performance measure. Secondly, the weight on all or some quantitative performance measures can be subjectively determined. And lastly, a subjective performance threshold can be used. In this case, a bonus will be paid if the manager's measured performance exceeds a subjectively determined threshold.

Gibbs et al. (2004) agree with Prendergast and Topel (1993) by saying that when a manager's trust (proxied by a manager's tenure at a firm) is higher, subjective performance measures can be effectively used. Agency theory assumes the principal is honest. If subjective performance measures are used, an honest principal is crucial; managers should be convinced of the fact that their principal is to be trusted because only then are they able to believe they are being evaluated fairly (Prendergast & Topel, 1993). If managers do not fully trust in the fairness of principals, managers will lower their effort and the subjective incentive will fail. In Contrast, they will put more effort in their work if they believe principals are honest because they know the firm will compensate them fairly for their extra effort.

3.3.1 Advantages of subjective performance measures

I will now discuss some advantages of using subjective performance measures. I begin with explaining increased goal congruence between the shareholders and the managers when using alternative measures, like subjectivity. Thereafter, I will discuss the relation between

subjective performance measures and increased manager's satisfaction with the pay scheme and increased profitability and net profit when a manager has a long tenure

(18)

18

Literature indicates that performance measurement can be improved by including more alternative performance measures next to using financial performance measures; subjective performance measures included. Feltham and Xie (1994) suggest that alternative measures such as subjective performance measures can lead to more goal congruence by improving the agent's effort allocation. According to them, the goal of the principal and effort allocation of the agent come together more, which makes subjective performance measures valuable to use in incentive contracts. This is because firms will compensate managers on the basis of performance that is in the best interest of the shareholders. Consequently, managers will perform appropriately and have the same goal as the firm. In addition, using subjective performance measures in incentive contracts of managers is positive for shareholders. Consequently, goal congruence can be seen as an advantage of the use of subjective

performance measures. Another advantage suggested by Gibbs et al. (2004), is that manager's satisfaction with the pay scheme increases when the owners of a firm use discretion in

performance measurement. Furthermore, they find evidence for a positive relation between subjective performance measures and productivity and net profit when a manager has a long tenure at the particular firm. From this can be concluded that the use of subjective

performance measures provides firms with an advantage (increased productivity and net profit) when managers have a long tenure at the firm.

Using divers performance measures provides a firm with discretion in the evaluation of managers, similar to using subjective performance measures. Subjective performance measures provide firms with discretion because no clear standards exists for measuring the effort of managers, only the subjective judgements of the firm. Similarly, a portfolio of performance measures can provide a firm with conflicting outcomes. These conflicting outcomes give firms the opportunity to differentiate the weight on the different performance measures afterwards to one that they think is optimal.

(19)

19

3.3.2 Disadvantages of subjective performance measures

I will now provide two disadvantages of the use of subjective performance measures. I discuss an increased vulnerability to reneging and the problem of bias when using discretion in

performance measures.

Empirical evidence shows that a principal can not commit to being honest when subjectively evaluating an agent (Prendergast & Topel, 1993). This brings some

disadvantages. The most obvious disadvantage of dishonest shareholders should be not compensating managers while they performed sufficient enough according to the incentive contract. Baker, Gibbons and Murphy (1994) think a performance measure based on discretion is more vulnerable to reneging (by dishonest shareholders) because an objective contract can be enforced by a court while a subjective contract cannot. Prendergast and Topel (1993) refute this by saying the incentives for shareholders to renege are often missing because they are not the residual claimants of managers' output. In other words, they do not receive the money that is not paid as a bonus to the managers. Baker et al. (1994) provide the notion that a contract based on subjective performance measures is a self-enforcing contract. They suggest a required trust in measuring performance so that the subjective performance assessments do not have to be enforced by court but by the firm itself for the sake of its reputation on the labour market. In other words, firms want be seen as honest because they have a reputation to maintain and they want the best managers' willingness to work at the firm. Therefore, firms have an intrinsic motivation not to renege.

Moers (2005) describes in his study the bias problem that can appear when using subjective performance measures. Shareholders are able to compensate managers to their preferences. Psychological research suggests that performance figures are low when a firm is

(20)

20

using subjective performance measures for incentive pay purpose. More bias in performance measures leads to more payouts but poorer performance. A problem that comes with bias of performance measures in a shareholder-manager relation is indirect costs (Moers, 2005). They relate to the difficult decision of personnel of shareholders. If performance figures are biased, then all managers seem to perform above average. This makes it difficult to pick the right manager for the right function. Also, when managers know about the bias, they could lose motivation and perform less and less from then on. Incentives and having the right manager for the right job are important for firm performance. Therefore bias in subjective performance measures has to be considered when designing an incentive contract.

Everything at a glance, I have found three clear advantages of the use of subjective performance measures; goal congruence increases between the shareholders and the managers, manager's satisfaction with the pay scheme increases and profitability and net profit increase if a manager has a long tenure. An increased vulnerability to reneging is a disadvantage of subjective performance measurement (Baker et al., 1994). However, this disadvantage has been refuted by Prendergast and Topel (1993) and Baker et al. (1994). They say that there will be no reneging because respectively, the shareholders are not the residual claimants of managers' output and for sake of the reputation of the firm. More concern exists on the bias problem which is the biggest disadvantage of discretion measurement.

3.4 Relative performance measures

A rationale for relative performance measures based on risk-sharing is provided by economic theory (Gibbons & Murphy, 1990). Although rewarding based solely on individual

performance provides incentives to perform better, measured performance and therefore compensation is almost always affected by random factors beyond the employee's control. It

(21)

21

is common to measure performance of employees relatively to other employees. Examples of this in practice are explicit contests, fixed bonus pools within a firm or division, forced-distribution performance appraisals and awarding promotions to one employee of a particular group. The same random factors that affect one employee's (manager's) performance often affect employee's performance in the same division, company, industry or market. A relative performance measure can insulate a manager from common uncertainty and therefore provide the manager with incentives.

3.4.1 Advantages of relative performance measures

I give the advantages of using relative performance measurement in this section. I see the insulation of managers from the common risk and uncertainty, the increased informative function for shareholders, the positive relation with financial performance and low costs of obtaining as the advantages of relative performance measures.

Gibbons and Murphy (1990) and, the more recent study of, Gong, Li and Shin (2011) think the biggest advantage of the use of relative performance measures is the insulation of managers from common risk and uncertainty when there exists an important source of it in the industry or firm. They explicitly say that the occupation for which the risk-sharing advantages of relative performance evaluation likely exceeds its counterproductive side-effects is top-level corporate management. The problem with incentive contracts for top managers based solely on performance of the firm is the fact that they make managers subject to the vagaries of the financial markets. These vagaries are far beyond the control of the managers. As explained earlier, relative measures of performance can partially insulate top managers from market shocks. Gong et al. (2011) also believe that relative performance measures provide a shareholder with a more informative measure to assess the manager's actions. Matsumura and

(22)

22

Shin (2006) used annual performance data of 214 postal stores in Korea between 1997 and 1999 to study the effect of implementing relative performance measures. They found that adding the relative measurement tool is positively related with financial performance. Also, the degree of common uncertainty is positively associated with store profitability. They believe agents are induced to exert more effort when common uncertainty is high and their performance is measured with relative performance measures, which leads to better financial performance. The incentives created by relative performance measures and manager's actions following this incentives increase current shareholder return (Gibbons & Murphy, 1990). This is comparable with the increased financial performance Matsumura and Shin (2006) find. The results of the study of Gibbons and Murphy (1990) do not support the hypothesis of the relation between relative performance measures and increased shareholder wealth. They also provide us with the cost advantage of relative performance measures. The costs of obtaining performance data from other firms is low; this information is accessible to everybody via the stock market or financial reports.

3.4.2 Disadvantages of relative performance measures

I begin this section with providing the disadvantages of Gibbons and Murphy (1990) that appear when using relative performance measures; sabotaging the measured performance of any other reference group, generating an incentive to collude with colleagues and so to shirk and applying for jobs with inept colleagues. Thereafter, I explain the influence of fixed peer groups in the long-run. I end with the disadvantage of perceived unfairness that managers can feel when relative performance measures are used.

Gibbons and Murphy (1990) sum up the following disadvantages of the use of relative performance measures: an incentive is generated to sabotage the measured performance of

(23)

23

any other reference group, an incentive is generated to collude with colleagues and so to shirk and to apply for jobs with inept colleagues. These disadvantages seem to be unlikely in the shareholder-manager relation since a top manager tends to have limited interaction with top managers in rival firms. The above is not related to managers with the highest office because they have no colleagues on the same level. However, it does hold for presidents and vice presidents. According to Baker (1992), a disadvantage of relative performance measurements appears with the assumption of fixed peer groups (the industry to which the firm will be compared too when measuring performance). Fixed peer groups only hold in the short-run. In the long-run a manager can make product line decisions and acquisition choices that shift the firm into a different industry. In such a setting, it is suggested that top managers make

decisions that shift the firm into an industry where they can outperform. Managers will not necessarily invest in industries with the highest returns. Matsumura and Shin (2006) find evidence of a mitigated relative performance incentive effect in stores at which the employees likely experience a perceived high unfairness. So, the incentive effect of relative performance measures decreases when employees believe they are not fairly measured. They also suppose that the net benefits of introducing relative performance measures may be conditional on the degree of common uncertainty. Matsumura and Shin (2006) find three implications when empirically studying the use of relative performance measures. First, a significant level of dysfunctional behaviour can occur. A situation in which employees believe the threshold to be unfair, for instance, can cause a decreased morale and scepticism. Second, this perceived unfairness may result in a mitigated incentive effect. Third, the empirical evidence from the research suggests that the perceived unfairness increases when common uncertainty

decreases. This suggests that setting the threshold should be very carefully done, otherwise perceived unfairness increases. From the three implications I make up that the change of a

(24)

24

perceived unfairness, and its consequences, by managers is a disadvantage of the use of relative performance measures.

From the above I formed four groups of advantages of the use of relative performance measures and five groups of disadvantages. I see the insulation of managers from the common risk and uncertainty, the increased informative function for shareholders, the positive relation with financial performance and low costs of obtaining as the advantages of relative

performance measures. First, the disadvantages I found were sabotaging the measured performance of any other reference group, generating an incentive to collude with colleagues and so to shirk and applying for jobs with inept colleagues. Next, I found the fixed peer groups, which can be influenced in the long-run, and the perceived unfairness that managers can feel.

3.5 Overview of the (dis)advantages

The advantages and disadvantages I found in literature are combined in an overview in table 3.5.1. I was able to find 10 advantages and 9 disadvantages. So, there is no large difference between the number of advantages and disadvantages I found. Also, looking at the single performance measures, no measure has a large difference between their number of advantages and disadvantages.

(25)

25 Table 3.5.1* advantages disadvantages financial performance measures

1.easy+cheap to calculate and interpreted 1. decreases long-term thinking; lower stock return

2. increases short-term performance 2. increased noise 3. most aggregate measure

subjective performance measures

4. increases goal congruence 3. increases vulnerability to reneging**

5. increases manager's satisfaction with pay scheme

4. more bias in performance measure

6. increases profitability+net profit when long manager's tenure

relative performance measures

7. insulation of manager from common risk

5. sabotaging reference group's performance

8. increases information for shareholders 6. colluding with colleagues and shirking

9. increases financial performance 7. applying for jobs with inept colleagues

10. low obtaining costs of data 8. managers influencing the peer groups in the long run

9. danger of perceived unfairness *table 3.5.1 provides an overview of the (dis)advantages as given in chapter 3

**Prendergast and Topel (1993) and Baker et al. (1994) contradict this disadvantage

4. The use of performance measures

4.1 The use of multiple performance measures

Feltham and Xie (1994) suggest that implementing alternative measures, next to the more traditional financial measures, can lead to more goal congruence by improving the agent's effort allocation. According to them the goal of the principal and effort allocation of the agent come together more which makes subjective performance measures valuable to use in

incentive contracts. In addition, shareholders will value subjective performance measures in incentive contracts of managers. According to Datar, Kulp and Lambert (2001) a shareholder can be allowed to estimate different weights for performance measures that bring the

(26)

26

multiple performance measures. Combining the studies of Feltham and Xie (1994) and Datar et al. (2001) brings Moers (2005) to the conclusion that diversity in performance

measurement can lead to two things. Namely, increasing effort intensity from managers and more congruity of the goals of managers and shareholders by changing managers' effort allocation.

So, using divers performance measures is generally speaking optimal for shareholders (Moers, 2005). In this study, we will assume the use of multiple performance measures as the best way for shareholders to measure managers' performance. I will now describe some very specific situations and try to explain the optimal choice that shareholders are able to make with respect to their differentiation in using-frequency of performance measures. In which situation are shareholders more likely to use which performance measure? I will look at significant findings in literature. As in the previous section, I first discuss the shifts related with financial performance measures, then subjective performance measures and at last the relative performance measures.

4.2 Financial performance measures

Financial performance measures are extensively used in compensation contracts of executives (Bushman & Smith, 2001). This is partly due to the fact that this traditional way of

performance measurement is relatively easy and cheap, boosts short-term profitability and is the most aggregate. I will first study the specific cases of firms' strategy and regulations. Thereafter, I will study the influence of volatility on a firm's performance measurement.

Firstly, the paper of Ittner , Larcker and Rajan (1997) examines the relative weights placed upon performance measures in chief executive officer bonus contracts. Using a cross-sectional regression analysis, they conducted an analysis of data from 314 firms for the year

(27)

27

1993 or 1994. They involved two types of firms in their research. Firstly, firms where CEO's bonus only dependents on a financial measures function. So, no nonfinancial performance measures are used in this firms. Secondly, firms where both financial and nonfinancial performance measures are used but where the weights placed on the measures can be

determined. From the 314 firms studied, 114 firms used nonfinancial measures, with 5 firms using solely nonfinancial measures. They find three significant relationships; firms place less weight on financial performance measures when they follow a more quality focused strategy, are more focused on innovation and face stricter regulations. So, a firm's strategy and the regulations it has to follow influence its choice of performance measures. First, firms that follow a more quality focused strategy are generally more likely to use non-financial

performance measures because financial measures do not provide the sufficient information about the quality; as explained in chapter 3, financial performance measures are focused on the short-term. Ittner et al. (1997) describe firms with more quality focus as firms that won or have been a finalist of a major quality award. Second, innovative firms that try to differentiate themselves from the market generally focus less on cost leadership. These firms are

concerned with results on the long-run and with market-share. This is not included in the short-run focused financial performance. Vice versa, firms focussing on costs will be likely to use financial performance measures. Third, regulators can partially influence the prices of regulated firms, usually firms in the utilities industry (e.g. electric, gas and water). Regulators desire the best possible quality and therefore, for example, high customer satisfaction is used as a (non-financial) performance measure for the firms in the industry. The firms will

compensate their managers on the basis of the same measure, because they are rewarded for achieving high scores on these measures.

Secondly, Stathopoulos, Voulgaris and Walker (2014) examined the effect of IFRS (International Financial Reporting Standards) on a firm's choice of performance measures for

(28)

28

their managers using a sample of more than 3,000 UK firms over a period of eight years. The implementation of IFRS includes the use of fair value accounting. Therefore, accounting data becomes more sensitive to worldwide market movements; a firm's volatility increases.

Market-wide noise is no longer eliminated and therefore financial (accounting) performance measures provide less additional information about managers' performance (Bushman & Smith, 2001). An increase in earnings volatility due to fair value accounting is likely driven by factors outside management's control. Consequently, the traditional and easy to use financial performance measure of earnings will become less noise-signalling than it already was and therefore become less attractive. Similarly, the analysis of Stathopoulos et al. (2014) shows a lower weight of firms on financial performance measures after the introduction of IFRS. Their results are consistent with the decreased noise-signalling ratio of Bushman and Smith (2001). From this, it can be concluded that an increase in earnings volatility declines the use of financial performance measures and because of fair value accounting, IFRS using firms will be using relatively less financial performance measures compared to non IFRS using firms. So, firms with high volatility, like IFRS using firms, are likely to less frequently use financial performance measures.

4.3 Subjective performance measures

In this paragraph I will study some specific circumstances in which subjective performance measures are more or less likely to be used. First, I will consider the situation in which no accurate financial performance measures are to be found. Thereafter, I will provide some subjective performance measurement related information about firm size, the number participants and undiversified firms, followed by a situation with high organizational interdependencies.

(29)

29

First, Dess and Robinson (1984) believe the use of objective (financial) performance measures is more important than subjective measures. They suggest subjective performance measures are only useful in specific circumstances. This is when subjective measures are able to complement objective measures, like when shareholders encounter problems in obtaining accurate financial measures. Gibbs et al. (2004) find that the use of subjective measurement is larger where financial performance measures are not used and so subjective measures seem to be a substitute for financial performance measures. Dess and Robinson (1984) conducted empirical research under 26 American firms, active in the paints and allied products industry. Data was conceived from the CEOs by way of an interview, a decision making case and a questionnaire. They focused on two relationships; between subjective and objective measures of return on assets and growth of sales and between measures of return on sales and growth of sales and performance measures that measure firms' overall performance. Return on assets is used because it is viewed as a measure of efficiency with respect to a profitable use of assets. Growth of sales shows firm's performance by showing how a firm relates to its environment by setting its products in the market in a successful way. Their results show a powerful correlation between objective (financial) measures of the absolute changes of return on assets and sales growth and subjective perceptions of improvements relatively of the same variables. They provide a conclusion of the possibility of using subjective performance measures to assess the changes in return on assets and sales growth when an objective measurement about these variables is not be found. This possibility suggests a relationship between the use of subjective performance measures and a lack of accurate financial performance. However, the results of the study of Dess and Robinson (1984) do not provide enough evidence to assume such a general statement. I can only suspect this relationship but not assume it.

Second, Murphy and Oyer (2001) offer empirical data with respect to the use of subjective performance measures, based on a confidential survey of 280 incentive contracts.

(30)

30

This survey was gathered by a large compensating-consulting firm. The results of the study of Murphy and Oyer (2001) show that subjective measurement among publicly traded firms increases with firm size and the participants in the incentive plan. In addition, they show less received compensation based on subjective performance measures for managers in multi-segment or multi-industry firms in relation to managers in undiversified, single-multi-segment firms. So, the use of subjective performance measures is more likely in undiversified firms in relation to diversified firms.

Third, Gibbs et al. (2004) examined when firms make greater use of subjective

performance measures in incentive contracts. They did so by using data from a sample of 526 department managers in 250 car dealerships. Obviously, the studied managers are no high office executives in large firms, but still the results can add value to this thesis because there is an agency relationship between higher office managers (principals) and department

managers (agents). After all, Gibbs et al. (2004) say: "Subjective performance evaluations are present and important in virtually all jobs, from the lowest levels to CEOs". This implies that they perform their research also with respect to learn about subjective performance

measurement for CEOs. The finding of Gibbs et al. (2004) that is most closely related to this study refers to organizational interdependencies. Organizational interdependencies can partially be compared with the term common risk that Gong et al. (2011) use, because it is both about risk from outside. It is about firms that are subsidiary to risk caused by other firms in the production cycle, like in joint production. For example, if a firm produces cars, it is dependent on other firms producing parts of the car. If something went amiss within another firm in the production cycle and therefore performance of the firm decreases, the decreased performance is out of the range of the manager. Gibbs et al. (2004) find a positive relation between organizational interdependencies and the use of subjective performance measures. Therefore, firms are more likely to use subjective performance measures when the extent of

(31)

31

organizational interdependencies is higher. It can be suggested that firms want to eliminate this interdependencies by subjectively measuring managers. When subjectively measuring performance, firms are able to disregard organizational interdependencies when detecting them.

4.4 Relative performance measures

To give a better understanding of the number of firms using relative performance measures I use the studies of Bannister and Newman (2003) and Gong et al (2011). Bannister and Newman (2003) used a sample of 160 large size firms to see if they use relative

performance evaluation in some aspect in CEO's compensation contracts and found a

percentage of 28 percent.Gong et al. (2011) found a quite similar percentage. They used S&P 1500 firms' first proxy disclosures to find that 25.44 percent of their sample firms use relative performance measures to determine top manager compensation. I will begin the literature review of the relative performance measurement part with seven circumstances Gong et al. (2011) investigated. I have divided this into four cases which are more easy to conduct with economic literature and three cases which are respectively less easy to conduct. After this, I will describe the situations of high manager's flexibility to invest in other industries, more stakeholder monitoring and concern about pay and performance and at last a small CEO talent pool.

First of all, Gong et al. (2011) believe their findings to be in line with economic theories predicting that firms tend to use more relative performance measures when they are more exposed to common risk and when their executives are less able to self-hedge their risk. Summarizing their findings they find that the likeliness of firms using relative performance measures is higher when a firms' common risk is higher, activities are in a less concentrated

(32)

32

industry, growth opportunities are lower and chief executive officers are less wealthy. I will now further explain these findings. First, Janakiraman, Lambert and Larcker (1992) put the intuition forward that relative performance measurement is more effective at insulating exogenous shocks when a firm shares more common risk with its peer group. The intuition of Janakiraman et al. (1992) is in line with the finding that firms with less idiosyncratic risk are more likely to use relative measures (Gong et al., 2011). Second, the prediction that more competition in a market (low industry concentration) discourages shareholders to use relative performance measures is in contrast to the findings of Gong et al. (2011) and Aggarwal and Samwick (1999) that firms are more likely to use this measures when operating in a less concentrated industry. Aggarwal and Samwick (1999) found empirical evidence of a positive sensitivity of compensation to rival firm performance (relative performance measurement) that is increasing in the degree of competition in the industry. The results of Gong et al. (2011) show an higher level of common risk in a more competitive industry and hence brings relative performance measurement more benefits. Third, Murphy (2000) contradicts Gong et al. (2011) with the suggestion that firms with more growth opportunities are more likely to use relative performance measures. However, the findings of Gong et al. (2011) are supported by Albuquerque (2013). She provides the argument that a high level of growth opportunities is less beneficial when using relative performance measures since performance of peer groups provides less information of external shocks for these firms. I assume firms to be more likely to use relative performance measures when they have got less growth opportunities because Albuquerque (2013) and Gong et al. (2011) provide more empirical evidence than Murphy (2000). The contradictions in the findings related to competiveness and growth opportunities indicate that in some cases other factors together with competiveness might influence the use of relative performance measures. Fourth, Gong et al. (2011) find evidence suggesting that

(33)

33

firms are less likely to use relative performance evaluation when chief executive officers hold a larger value of stock and options (are more wealthy).

Moreover, Gong et al. (2011) find some evidence that is more difficult to assess with economic literature. Namely, firms are more likely to use relative performance measures when they are having a larger market capitalization, having a more independent board and hiring compensating consultants. Having a more independent board is positively related with the likelihood of using relative performance measures because of the stronger internal

governance which is in favour of using the measurement. Compensation consultants appear to have the greatest impact, increased likelihood of 16%, on the use of relative performance measures. This probably suggests that their specialized knowledge and expertise facilitates the use of relative performance measures. So Gong et al. (2011) provide us with respectively seven situations in which shareholders are likely to intensify the use of relative performance measures.

Furthermore, in chapter 3.4, I explained that fixed peer groups may only hold in the short-run. In the long-run, a manager can make product line decisions and acquisition choices that shift the firm into a different industry. In such a setting, it is suggested that top managers make decisions that shift the firm into an industry where they can outperform. Therefore the value of a contract based on relative performance measures depends on the degree of

flexibility of a manager to invest in alternative industries (Baker, 1992). If a manager's flexibility to invest in other industries is high, shareholders will likely make less intensively use of relative performance measures and vice versa.

Next, Bannister and Newman (2003) performed a statistical analyses on the 160 large size firms mentioned earlier. Their results suggest that environments with greater monitoring and concern over pay and performance are characterised by higher use of relative

(34)

34

is positively related with relative performance evaluation usage means a likeliness of increased weight on relative measures in this environment.

Finally, Rajgopal, Shevlin and Zamora (2006) find the relatively complex relation between relative performance measurement and outside employment opportunities of CEOs. They directly test and find evidence for the hypothesis that firms are better off when

decreasing the use of relative performance evaluation if the CEO's reservation wages from outside employment opportunities fluctuate with the market. In particular, the supply of talented CEOs is relatively inelastic if CEO talent is scarce. When CEO talent is scarce, it may be for the best to compensate CEOs for market-wide shocks. Namely, the shocks can raise the market value of a firm and the CEO's outside employment opportunities. If a firm refuses to compensate their managers when the market has a shortage, managers are inclined to leave the firm for a better compensation at another firm. This compensation can be done by way of relative performance measures. The use of relative performance measures provides managers with higher compensations when other firms offer these managers higher

reservation wages. Therefore, managers are inclined to stay at the firm they are currently working. So, from the study of Rajgopal et al. (2006) we can conclude that the use of relative performance measurement is more likely when CEO talent is scarce.

4.5 Overview of the findings

Adding all the findings in chapter 4, I find 20 situations in which there is something to say about when shareholders are likely to be inclined to put more or less weight on one of the three studied performance measures. Especially with respect to relative performance measures is much to be found, namely 10 situations. The findings of this study are provided by an overview in table 4.5.1. The 20 situations provided in table 4.5.1 are the answer of this study

(35)

35

on the research question as stated in chapter 1: "Under which circumstances are firms inclined to more or less frequently use certain performance measures?". The significance of my

findings are to the extent in which my overview is complete. It is recommended that more literature research has to be done to complement this overview. One of the pitfalls of this study is that it is a literature review. In order to come up with definitive answers it is

recommended to conduct an empirical research. Moreover, the overview could be improved by conducting empirical research to the specific circumstances given in the overview in relation to the other two, not applied, performance measures. For instance, what is the

influence of strategy on the use of subjective and relative performance measures. The findings of this study can be applicable in real-world practice by predicting which performance

measures shareholders under specific circumstances will use or by shareholders to see what is commonly used in the specific situation of their firm.

(36)

36

Table 4.5.1 *

performance measure situation likely to more/less frequently use this measure reason financial performance measures

1. more quality focus less provide no sufficient info about quality

2. more innovation focus

less provide only short-term info

3. more costs focus** more cost info included

4. stricter regulation less firm rewarded by non-fin. measures

5. increased earnings volatility

less less noise signalling

subjective performance measures

6. no accurate fpe*** more complement/substitute (especially ROA + sales growth)

7. large firm size more positive correlation 8. more plan

participants

more positive correlation 9. diversified firm less less spe use

10. organizational interdependencies

more Joint production causes outside risk

relative performance measures

11. common risk more insulate outside risk 12. less concentrated

industry

more more competition so more common risk

13. less growth opportunities

more peer groups provide more info of external shocks 14. wealthy CEOs less positive correlation 15. larger market

capitalization

more positive correlation 16. more

independent board

more stronger internal

governance, in favour of rpe

17. hiring compensation consultants

more their knowledge + expertise facilitates rpe use

18. more flexibility to invest in other industries

less managers shift to

industries where they can outperform

19. more concern and monitoring over pay and performance

more positive correlation

20. CEO talent is scarce

more to offer compensation to hold/hire talented CEOs *table 4.5.1 provides an overview of the findings of this study as explained in chapter 4 **this finding follows partly from finding number 2

(37)

37

5 Conclusion

In this thesis, I try to find an answer to the question: under which circumstances are firms inclined to more or less frequently use certain performance measures? To answer this question the theory underlying the problem had to be explained. Agency theory describes the conflict between the different desires and goals of the principal and the self-interested agent. This study assesses the positivist line of research, a conflict of interests between shareholders and managers, in studying the agency problem. The conflicting goals between shareholders and managers (in this study high office executives) means that it is important for shareholders to develop a system that puts managers in the same direction as shareholders. The study focuses on three relatively different sorts of performance measures to minimize the shareholder-manager problem; the traditional and often used financial performance measures, the totally subjective performance measures and the more recent relative performance measures. They are explained and their (dis)advantages are provided to give a better understanding of the different performance measures. Thereafter, I gave an overview of findings in literature. I provide 20 situations in which shareholders are likely to more or less frequently use one of the three studied performance measures.

The overview contributes to literature, because no such an overview can be found in literature. It can help researchers and shareholders with clarifying which performance measures are used in practice. One of the pitfalls of this study is that it is a literature review. In order to come up with definitive answers it is recommended to conduct an empirical research. Moreover, the overview could be improved by conducting empirical research to the specific circumstances given in the overview in relation to the other two, not applied,

performance measures. For instance, what is the influence of strategy on the use of subjective and relative performance measures.

(38)

38

6 References

Aggarwal, R., & Samwick, A. A. (1999). Executive compensation, strategic competition, and relative performance evaluation: Theory and evidence. The journal of finance, 54(6), 1999-2043.

Albuquerque, A. M. (2013). Do growth-option firms use less relative performance evaluation? The accounting review, 89(1), 27-60.

Baker, G. P. (1992). Incentive contracts and performance measurement. Journal of political economy, 100(3), 598-614.

Baker, G., Gibbons, R., & Murphy, K. J. (1994). Subjective performance measures in optimal incentive contracts. The quarterly journal of economics, 109(4), 1125-1156.

Bannister, J. W., & Newman, H. A. (2003). Analysis of corporate disclosures on relative performance evaluation. Accounting horizons, 17(3), 235-246.

Bushman, R. M., & Smith, A. J. (2001). Financial accounting information and corporate governance. Journal of accounting and economics, 32(1), 237-333.

Datar, S., Kulp, S. C., & Lambert, R. A. (2001). Balancing performance measures. Journal of accounting research, 39(1), 75-92.

Dess, G. G., & Robinson, R. B. (1984). Measuring organizational performance in the absence of objective measures: the case of the privately‐held firm and conglomerate business unit. Strategic management journal, 5(3), 265-273.

Ding, Y., Lebas, M. J., & Stolowy, H. (2013). Financial Accounting and Reporting (4). Cengage Learning.

Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of management review, 14(1), 57-74.

(39)

39

Feltham, G. A., & Xie, J. (1994). Performance measure congruity and diversity in multi-task principal/agent relations. The accounting review, 69(3), 429-453.

Gerhart, B., & Milkovich, G. T. (1990). Organizational differences in managerial

compensation and financial performance. Academy of management journal, 33(4), 663-691.

Gibbons, R., & Murphy, K. J. (1990). Relative performance evaluation for chief executive officers. Industrial & labor relations review, 43(3), 30S-51S.

Gibbs, M., Merchant, K. A., Stede, W. A. V. D., & Vargus, M. E. (2004). Determinants and effects of subjectivity in incentives. The accounting review, 79(2), 409-436.

Gong, G., Li, L. Y., & Shin, J. Y. (2011). Relative performance evaluation and related peer groups in executive compensation contracts. The accounting review, 86(3), 1007-1043.

Hölmstrom, B. (1979). Moral hazard and observability. The bell journal of economics, 10(1), 74-91.

Indjejikian, R. J. (1999). Performance evaluation and compensation research: An agency perspective. Accounting horizons, 13(2), 147-157.

Ittner, C. D., Larcker, D. F., & Meyer, M. W. (2003). Subjectivity and the weighting of performance measures: Evidence from a balanced scorecard. The accounting review, 78(3), 725-758.

Ittner, C. D., Larcker, D. F., & Rajan, M. V. (1997). The choice of performance measures in annual bonus contracts. The accounting review, 72(2), 231-255.

Janakiraman, S. N., Lambert, R. A., & Larcker, D. F. (1992). An empirical investigation of the relative performance evaluation hypothesis. Journal of accounting research, 30(1), 53-69.

The circumstances under which firms are inclined to more/less frequently use certain performance measures : a focus on financial, subjective and relative performance measures