1 The effect of the financial crisis on the use of non-financial performance
measurements in CEO compensation
Name: Giovanni Kuhurima Student number: 11397098
Thesis supervisor: Pouyan Ghazizadeh Date: August 20, 2018
Word count: 7779
MSc Accountancy & Control, specialization Control
2 Statement of Originality
This document is written by student Giovanni Kuhurima 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.
3 Abstract
This thesis studies the effect of the financial crisis on the use of non-financial performance measures in CEO compensation. For this research, data of 70 firm stock-listed firms has been collected for the years 2005-2006 and 2012-2013. The years 2005-2006 are considered as pre-crisis years. The years 2012-2013 are considered as post-pre-crisis years. The use of non-financial performance measurements is the dependent variable in this study. This information is found in the proxy statements that are published by the firms every year. The results show that the use of non-financial performance measurements in CEO compensation is not significant related with the post-crisis years. However, when excluding firms that appointed a new CEO, there is a positive and significant relation between the use of non-financial performance measures and the post-crisis years. This study concludes that only in firms that did not appoint a new CEO, the use of non-financial performance measurements in CEO compensation increased in the post-crisis years.
4
Contents
Abstract ...3 1 Introduction ...5 2 Literature review ...7 2.1 CEO compensation ...7 2.2 Performance measurements ...8 2.3 Financial crisis ...10 2.4 Hypothesis ...11 3 Methodology ...12 3.1 Databases ...12 3.2 Data collection ...12 3.3 Model ...15 4 Results ...18 4.1 Descriptive statistics ...18 4.2 Correlation ...20 4.3 Regression ...22 4.4 Hypothesis ...26 5 Conclusion ...27 Bibliography ...29 Appendices ...315 1 Introduction
A CEO compensation plan consists of two types of performance measurement. The two types are financial and non-financial measurements. Financial measurements are measures that can be found in the financial statements such as profit and costs. Non-financial measurements are more firm-specific and looks at measures such as efficiency, customer satisfaction and market share. There are several reasons for choosing either one of the performance measures. According to Ittner et al. (1998) financial measurements are the most important and used measures. However, they find that there is shift in the use of financial performance measures, as non-financial measures are used more often when the noise in financial performance measures increases. This noise reduces the informational value of financial performance measures, and firms are therefore looking for other measures that are more reliable and useful in evaluating the CEO.
The financial crisis of 2008 had a huge impact on the market. The crisis, which started in 2008, has led to uncertainty in the financial markets. This uncertainty has led to a decline in revenues, sales, and impacted the financial performance of firms. According to Stock (2014) the financial crisis has led to financial shocks in the market. These shocks impacted the financial figures, and therefore increased noise in the financial performance measures. In this bad economic situation, most firms still grant a bonus to their CEOs. They have to be
incentivized during a period of uncertainty, which ultimately lead to actions that are taken to improve the firm’s performance. However, determining bonuses for CEO’s during the financial crisis becomes more difficult, as financial measures become less reliable. But when the crisis comes to its end, the uncertainty in the markets decreases. This might lead to a reduction of noise in the financial performance measures. This study will investigate the effect of the financial crisis on the use of non-financial performance measures after the crisis.
Several studies conclude that noise causes a shift in the use of non-financial
performance measures (Ittner, Larcker, & Rajan, 1997; Davila & Venkatachalam, 2004). These studies show that the noise in financial performance measurements influences the use of
performance measures. However, there is not much research in the use of non-financial performance measures after the financial crisis. There are two possibilities of the use of non-financial performance measurements after the crisis. First of all, the shift in the use of performance measures may be short lasting. In this case, with the years after the crisis, the noise in financial performance measurements decreases and the uncertainty in the market decreases. As a consequence, the use of non-financial performance measures returns back to its
6 level before the crisis. Second of all, there is a possibility that this shift is lasting for a longer period. The end of the crisis reduces the noise, while the use of non-financial performance measures does not decreases.
To find out what effect the financial crisis has on the use of non-financial performance measurements after the crisis, the following research question is answered:
What effect does the financial crisis have on the use of non-financial performance measures in CEO compensation after the crisis?
The research question covers the financial crisis and the use of non-financial performances in CEO compensations. The effect of the financial crisis is determined by comparing the pre-crisis years and the post-crisis years. The years 2005-2006 are considered as the pre-crisis years, whereas the years 2012-2013 are considered as the post-crisis years. The reduction of noise in the financial performance measurements is assumed in the years after the crisis. The use of non-financial performance measures are measured using data of firms for the given time period. Data for the non-financial performance measurements is gathered from several
databases, including Compustat, Execucomp and proxy statements that are available in the Lexis/Nexis database. For the periods 2005-2006 and 2012-2013, data from 70 American listed firms is collected. The collected data shows the effect of the use of non-financial performance measures comparing the years before and after the crisis. The findings are as follows. The results provide evidence that when there is a new CEO is hired during the crisis years, less non-financial performance measurements are used in the CEO compensation plan after the crisis. It also provides evidence that in organizations where there is no change in CEO, the use of non-financial performance measurements increases compared to before the crisis. The contribution of this study is as follows. First of all, this paper gives empirical evidence of the use of non-financial performance measures after the financial crisis. It extends existing literature such as Ittner et al. (1997) that looks at the change in the use of non-financial performance measures because of the noise in financial measures. Second of all, it shows that after the crisis, firms do not change the compensation contracts according to the years before the crisis.
The structure of this paper is as follows. Firstly, the theory and literature needed to answer the research question is covered. This follows with the hypothesis stated for the
7 tests is discussed. Lastly, this paper ends with the conclusion of the results, closing with
suggestions for future research. 2 Literature review
The literature review consists of four parts. The first part discusses the CEO compensation. The second part is about performance measurements. The third part discusses the financial crisis. In the last part, the hypothesis for this test is stated.
2.1 CEO compensation
The agency theory, which is explained by Jensen & Meckling (1976), explains the
relationship between the principal and the agent. The agent is performing services on behalf of the principal. In the case of a CEO, he or she is performing on behalf of the firm. But a CEO doesn’t always perform in the best interest of the firm. This happens when he or she takes decisions that are the most beneficial for the agent, and less for the principal (Grossman & Hart, 1983). Short-term oriented CEOs act in a way that each decision boosts the financial performance for a short period. This will benefit them, as they will be evaluated on the immediate effect. However, this short-term oriented focus is not beneficial for the firm. A way to solve this ‘agency problem’ is by using a compensation plan. The use of a
compensation plan for rewarding CEOs for decisions that enhances the financial performance of the firm, should lead to an alignment of the interest of both parties (Beatty & Zajac, 1994). This is in line with the paper of Mehran (1995), as he finds that when equity bonus is
included in the CEO compensation, the financial performance of the organization improves. The CEO compensation consists of two parts: the fixed and the variable parts. It includes base salary, annual bonus, long-term incentives and several stock-related bonuses (Murphy, 1999). The fixed part of the compensation is the base salary. This is the key component of the contract and consists of the fixed amount the CEO receives, the base salary. Since this part of the compensation is fixed, the financial performance of the firm does not directly influence this type of compensation. This compensation is usually fixed for a given period. The variable part consists of the annual bonus plan, stock options and other forms of compensation. A CEO has influence on these parts of the compensation, as the amount of these forms of compensation are based on the financial performance of the organization. The annual bonus plan is based on the firm’s performance of the year. Achieving the targets that are set prior year should lead to a bonus for the executive, while not achieving the target will lead to a lower or even missing the bonus. Companies use a
8 variety of financial and nonfinancial measurements for determining this bonus. A more
complex incentive are stock options. Stock options are contracts which give the recipient the right to buy a share of stock at a specified "exercise" (or "strike") price for a
pre-specified term (1999). More recent study show that these incentives are complex since they cover multiple years. Its measurement is more complex compared to the short-term oriented incentive plans such as the annual bonus plan. For stock options, the CEO has the right to buy the share, but are not obliged. There are doubts whether this could be considered as part of the compensation, since it depends on if it is exercised or not (Murphy, 2012). The long-term incentive plan is based on the long-long-term financial performance. It is similar as the annual bonus plan, as the bonus will be granted when the objectives are met. While the annual bonus is short-term oriented, the long-term incentive plan is usually focused on a multiple years plan. (Murphy, 1999)
2.2 Performance measurements
Different performance measurements are available for firms to use. The two types of performance measurements are financial and non-financial performance measurements. In the past, most firms only used financial performance measures for rewarding the CEO. These financial performance measures were considered as the most important, and were therefore the most used type of performance measure (Ittner & Larcker, 1998). Examples of financial performance measurements are profit, sales, stock price, earnings per share (Ibrahim & Lloyd, 2011)
Financial performance measures consists of two types; market measures and accounting measures. They differ from each other as they are based on different figures. Market measures are based on stock prices. These measures are forward looking measures, as it reflects future cash flows and risks. Accounting measures are based on figures that can be found in financial statements. These measures are considered as backward looking, since they are based on actions and outcomes that already happened. Examples of accounting measures are: Return on assets, Return on investment, return on equity and earnings before income.
The importance of financial performance measures is mentioned in a study by Bacidore et al (1997). They show that financial measurements provides information about the performance of the firm. Any decisions that are made by the CEOs should therefore be reflected in these measurements. This could be a motivation for using financial performance measures in the CEO compensation.
9 firm. Examples of non-financial performance measurements are workforce development,
product quality, customer satisfaction, on time delivery, innovation measures, attainment of strategic objectives, market share, efficiency, productivity, leadership and employee
satisfaction (Ibrahim & Lloyd, 2011; Ittner, Larcker, & Rajan, 1997). These nonfinancial measurements are based on figures that are not required to be mentioned in financial statements.
Both measurements have their advantages and disadvantages. A major difference
between both is collecting the data. The data for financial performance measurements are found in the financial statements, which means an extraction of the data should be enough for a financial performance measure. Data for non-financial performance measurements is harder to obtain. Most of the time, this data cannot be found in the financial statements, but rather from other reports.
One of the most common used performance measures, which incorporates both types of measurements is the balanced scorecard (Kaplan & Norton, 2005). It consists of both financial and nonfinancial performance measurements. They argue that having only financial
measurements will lead to a short-term oriented focus, and therefore both are needed for a scorecard that is also long-term oriented.
There are several motivations why firms would choose for using non-financial measurements. A reason for using non-financial performance measurements is because
financial performance measurements cannot capture all the performance. The study of Ittner et al. (2003) argues that non-financial performance measures are important and should be used as it provides information that is not captured by financial measures. This is in line with the study of Sliwka (2002) as he finds that non-financial performance measurements are important when the financial performance measurements does not capture the actions.
Another motivation is that the use of non-financial performance measurements improves the financial performance and firm value. Several studies argue that the use of non-financial
performance measurements improves the firm’s performance. Ittner & Larcker (1998) find that non-financial measures about customer satisfaction are important for determining customer behavior, and is positively related to the firm’s performance. These measures provides information that leads to an increase in firm’s value. Said et al. (2003) conclude that the adoption of non-financial measurements leads to improvement of the current and future stock performance. This improvement leads to an improvement in the overall firm’s performance, and increases the value of the firm. Banker et al. (Banker, Potter, & Srinivasan, 2000) find that a high performance on non-financial measurements leads to managers taking decisions that are
10 long-term oriented. Therefore, the use of non-financial performance measurements in the CEO bonus will lead to an improvement in future firm performance.
Key in this paper is the presence of noise in financial performance measurements. Ittner et al. (1997) argued and find that an increase of noise reduces the informativeness of a
performance measure. The informational value of the measure decreases, and becomes less reliable and accurate. This leads to a shift to the use of more non-financial performance measurements, as they become more reliable and accurate than financial performance measurements. In other words, when the noise increases in financial performance measurements, firms will use more non-financial performance measurements
2.3 Financial crisis
Important factor in this study is the economic crisis, which started around 2008 with the bankruptcy of Lehman Brothers (Cordemans & Ide, 2012). This has led to panic in the market, where organizations became more careful in their spending.
Financial measures gives information about how an organization has performed. Financial shocks, which is considered as noise, have affected the financial measures and therefore these financial measures are volatile to the financial crisis (Stock, 2014). The financial crisis has led to financial shocks in the market. These shocks has increased the noise in financial performance measurements. Non-financial measures, however, are more firm-specific and are less vulnerable to noise. As described earlier, the noise in financial performance measurements, leads to more use of non-financial performance
measurements. The decreasing informational value of the financial performance
measurements causes a shift to more weight on non-financial performance measurements (Ittner, Larcker, & Rajan, 1997).
More recent papers provided evidence of the change in use of performance measurements because of noise in financial performance measurements. Davila (2004) find that the use of non-financial measurements is related with the volatility of financial performance measurements. This volatility has caused noise in the financial metrics. The importance of non-financial performance measurements increased and more weight was placed on non-financial performance measurements. Bebchuk and fried (2002) also find that there is a change in the use of performance measurements during the crisis. During the crisis, there is a shift from financial to non-financial performance measurements.
11 2.4 Hypothesis
The studies mentioned before give evidence of an increase in the use of non-financial performance measurements during the crisis. The financial crisis leads to noise in financial performance measurements and organizations are looking for reliable measurements that are useful for evaluating and rewarding the CEO. During the crisis, there is a shift in the use of non-financial performance measurements. The focus of this study is in the years after the crisis. The effect of the financial crisis is known for the years during the crisis. This paper argues whether the decrease in noise, which is the end of the crisis, will lead to a return of the use of non-financial performance measurements on the level before the crisis. To test this argument, the tests will compare the use of non-financial performance measurements in the pre-crisis and in the post-crisis years.
The following hypothesis is stated:
H1: After the financial crisis, firms will use more non-financial measures in the CEO compensation.
Ittner et al. (1997) finds the shift in the use of financial performance measurements due to an increase in noise in financial performance measures metrics. Since the financial crisis is over, less noise in the metrics is expected. However, the assumption of the hypothesis is that firms will not use less non-financial performance measurements after the crisis.
12 3 Methodology
This chapter explains how the hypothesis of this study is tested. This section is structured as follows. First of all, the databases that are used for the data are discussed. After, it is explained how the data was collected and combined into one dataset. At last, the model and variables that are used in this study is discussed.
3.1 Databases
This study looks at stock-listed firms in the USA. The data for this study are extracted from several databases. The data is collected for the period of 2005-2006, which are considered as the years before the crisis, and the years 2012-2013, which are considered as post-crisis years. To test and answer the hypothesis in this study, data from three databases were
collected. These are Execucomp, Compustat, and the Lexis/Nexis database. Execucomp contains information of CEOs. It is a sub database that collects CEO compensation
characteristics such as CEO salary, CEO bonuses. Therefore, the variables used in this thesis that contains CEO characteristics are extracted from this database. The Compustat database contains financial information of publicly traded companies in the US and Canada. It consists of published financial statements, such as the income statement, balance sheet, statement of cash flows. Lastly, the Lexis/Nexis database provides the proxy statements. In these proxy statements, the DEF 14A section is used to study the use of non-financial performance measurements in CEO compensation. Firms are obliged to publish the compensation that are given to CEOs. The use of non-financial measures is collected by observing these proxy statements, and see whether non-financial performance measurements are used in determining the CEO bonuses.
3.2 Data collection
All the data is gathered from the earlier mentioned databases. The combination of the data that are extracted from the databases is the sample that is used for this study. After
extracting the data from the Execucomp and Compustat databases, the data is merged into one dataset. After reducing the amount of observations, the data from the proxy statements is added to this dataset.
First, data was collected from the Execucomp.. The data of CEO compensation for the years 2005-2006 and 2012-2013 was extracted from the Execucomp database. After the
13 extraction, the dataset contained 4844 observations. Not all data was relevant for this
study. Table 1 shows a summary of the data that was used for the final dataset:
Table 1 Execucomp Database
Execucomp database
Data Used for
Name Identification of CEO Company
name Identification of firm
Company ID
Link with Compustat database
Year Period
SIC Industry
Became CEO Determine new CEO Left as CEO Determine new CEO
After the data of the Execucomp was gathered, data from the Compustat was added to the dataset. After retrieving all the data from the years 2005-2006 and 2012-2013, only the firms that was included in the Execucomp dataset remained in the final data set. Not all data was relevant for this study. In table 2, a summary of the data is provided.
14 Table 2 Compustat database
Compustat database
Data Used for
Company name Identification of firm
Company ID
Link with Compustat database
Year Period
SIC Industry
Assets Control variables Liabilities Control variables Net Income
Adjusted Control variables Stockholder's
equity Control variables
This data was merged into one dataset. In both separate datasets, Company ID was present. This was used as an identifier to merge the two datasets into one dataset. Combining the two sets from the two databases, the number of observations remained 4844. This set contained complete information of the CEO compensation and the financial information. The dataset was still too large for performing the hand collecting of the use of non-financial performance measures. Since the collection of the use of non-financial performance measurements
requires a lot of hand collecting work, it was decided to reduce the total amount of
observations. The first step to decrease the amount of observations was to keep only the 10 largest firms for each industry, based on the year 2013. The size is based on the variable FIRMSIZE. This variable is determined by the logarithm of total assets (Aldamen, Duncan, Kelly, & McNamara, 2012). The total observations for this sample after these two steps was 280 observations. This sample is used for the test. In table 3, a summary of the final dataset is given.
15 Table 3 Observations per industry
# of firms by industry Industry SIC # of firms # of total observations Mining & Construction 1000 10 40 Manufacturing 2000 10 40 Manufacturing 3000 10 40 Transportation 4000 10 40 Wholesale/retail sale 5000 10 40 Services 7000 10 40 Services 8000 10 40 280
For each firm, 4 years is taken (2005-2006 & 2012-2013)
The collection of the non-financial performance measurements in CEO compensation are hand-collected from the Lexis/Nexis database. It follows the same way as how the data was gathered for the study of Ittner et al. (1997). Each proxy statement had to be observed individually. After observing the proxy statements, the variable NF was given a 1 if a non-financial performance measure was mentioned in the bonus of the CEO. If it was not explicitly mentioned, a value of 0 was given for this variable.
In total, 280 proxy statements were observed for all the observations. Not all proxy
statements were in the Lexis/Nexis database, as some statements could only be found on the website of the firm.
3.3 Model
The effect of the financial crisis on the use of non-financial performance measures will be tested in the following way. It will be tested whether there is an increase in the use of non-financial performance measurements after the crisis. For this study, the dependent variable, which is called NF, is the use of non-financial performance measures in CEO
non-16 financial objective/criteria. If this criteria is not mentioned it will have a value of 0.
The empirical model used for the test is:
NF = α + β1 POSTCRISIS + β2 NEWCEO + β3 CRISISCEO + β4 LOSSPREVYEAR + β5 FIRMSIZE + β6 LEVERAGE + β7 INDROA + β8 – β14 SICCAT
The first variable is POSTCRISIS. This dummy variable has a value of 1 for the years 2012- 2013 and 0 for the years 2005-2006. Since the sample consists 2 years before the crisis and 2 years after the crisis, it is expected that half of the observations are in the post-crisis years. The next variables look at CEO characteristics. The variable NEWCEO is determined by hiring a new CEO. Hiring a new CEO can influence the use of non-financial performance measurements as it becomes easier to make adjustments in the CEO compensation contract. It has a value of 1 if a new CEO is hired year. When there is no change in CEO, it has a value of 0. Another CEO characteristic variable is CRISISCEO. This variable is determined by hiring a CEO during the crisis. If the new CEO is hired in the year 2008, 2009 or 2010, the so-called crisis years, it has a value of 1. When this is not the case, the variable has a value of 0.
The following variable is based on the financial performance of the firm. Previous years’ performance may have an impact on the use of non-financial measures, as it may be a reason to include more non-financial performance measures. As mentioned before, the use of non-financial performance measures improves the future financial performance of a firm (Banker, Potter, & Srinivasan, 2000). If the firm made a loss in the previous year, the variable has a value of 1. If they made a profit in the previous year, the variable has a value of 0. This data is extracted from the Compustat database.
The following control variables, which are also used in the study by Said (2003), are used in the model. These variables are included in the model, as they may influence the use of non-financial performance measurements. The variable INDROA is calculated by subtracting the firm’s ROA (Return on Assets) from the industry average ROA, where ROA is measured by dividing the net income with the total assets. FIRMSIZE is determined by the log of assets of the firm (Aldamen, Duncan, Kelly, & McNamara, 2012). A study by Hoque & James (2000) shows that the size of a firm has an influence on the use of the balanced scorecard. The variable LEVERAGE measures the leverage of a firm. It is calculated by dividing the firm’s debt by its
17 equity. The variables INDROA, FIRMSIZE, and LEVERAGE use data that are extracted from the Compustat database.
The sample is also divided into different industries. Each industry may have a different outcome, as some are more regulated than the other (Said, Hassabelnaby, & Wier, 2003). The industries are based on the SIC code, as it tells in what industry it operates.
18 4 Results
In this chapter, the results of the tests are covered. Firstly, the descriptive statistics are discussed. Secondly, the correlations between the variables are analyzed. Thirdly, the regression on the two samples is analyzed and discussed. Lastly, with the results of the regression, the hypothesis is tested and discussed.
4.1 Descriptive statistics
In table 4 the descriptive statistics of this sample are shown. The minimum, maximum, mean and standard deviation of all variables except the industries are in the table.
Table 4 Descriptive statistics
Descriptive statistics Descriptive statistics of the variables
Mean
Std.
Dev. Min Max
NF 0.518 0.501 0 1 POSTCRISIS 0.500 0.501 0 1 NEWCEO 0.443 0.498 0 1 CRISISCEO 0.089 0.286 0 1 LOSSPREVYEAR 0.054 0.226 0 1 FIRMSIZE 10.639 1.068 7.812 13.073 LEVERAGE 1.331 8.687 -81.056 31.544 INDROA -0.023 0.055 -0.207 0.180
Total number of observations are 280. The variables are explained in the methodology section.
Of all variables, 5 of them are dummy variables. They have a minimum of 0 and a maximum of 1. The mean of the variable NF is around 0.52. This means that in almost 52% of the observed CEO compensation plans, non-financial measures are used to determine the bonus a CEO earns. The other 48% bonus plans do not contain any non-financial performance
19 measures. The variable POSTCRISIS has a mean of 0.5. This is because for all 280
observations, exactly half of them are in the post-crisis period. The variable NEWCEO has a mean around 44%. One tenth of these changes, a new CEO is hired during the crisis. This is shown by the CRISISCEO variable. The variable LOSSPREVYEAR has a mean of around 0.05. This means that for all the observations, 95% of all observations had made profit in the previous year. Recovering of the financial crisis may be an explanation of this low mean. Another explanation is that the financial crisis may only lead to a lower profit, as cutting costs would be the reaction of the firms. The leverage is calculated by the log of assets, and has an average of around 1.33. This variable explains that the observations has 1.33 times more debt than equity, as leverage is the debt to equity ratio. The variable INDROA has a negative mean of -0.02. Since it is negative, the investments of the firms do not cover the costs. In fact, any investments will not result in any profit but rather a loss.
Table 5 Use of non-financial performance measurements in CEO compensation
Table 5 shows that the use of non-financial performance measurements has increased in the years after the crisis. It shows that before the crisis, for each year 34 CEO compensation plans consisted of non-financial performance measurements. In 2012 and 2013, respectively 38 and 39 times non-financial measures were used in the CEO compensation. A regression will be performed to investigate the relation between the use of non-financial performance
measurements and the post-crisis years.
31 32 33 34 35 36 37 38 39 40 2005 2006 2012 2013
Use of non-financial performance
measurements
20 4.2 Correlation
In table 6 the correlation between the variables is shown.
In overall, there is a weak and moderate correlation. The relations with the dependent variable is relatively small. The relation between FIRMSIZE and NF is the largest of all relations, while a loss in the previous year is the smallest. However, not all variables are significant related with the dependent variable. The correlation matrix show some relationships between the variables. The variables that are significant related with the use of non-financial measures are
CRISISCEO, FIRMSIZE and INDROA. This means that these variables have an influence on the NF variable. Interesting is that CRISISCEO has a significant negative correlation with the use of non-financial performance measures. There is some evidence that hiring a CEO during the crisis is negative related with the use of non-financial performance measurements.
21 Table 6 Correlation matrix
Correlation model Correlation of the variables
NF POSTCRISIS NEWCEO CRISISCEO LOSSPREVYEAR FIRMSIZE LEVERAGE INDROA
NF 1 POSTCRISIS 0.064 1 NEWCEO 0.026 0.446*** 1 CRISISCEO -0.107* 0.333*** 0.374*** 1 LOSSPREVYEAR 0.007 0.143*** 0.075 0.026 1 FIRMSIZE 0.262*** 0.284*** 0.243*** 0.141** -0.012 1 LEVERAGE -0.022 -0.034 0.016 0.02 -0.262*** -0.064 1 INDROA 0.169*** 0.209*** 0.048 0.075 0.389*** 0.164*** -0.111* 1
The asterisks denotes the following:
* significant at α = 0.10
** significant at α = 0.05
22 4.3 Regression
To test the hypothesis, a multivariate regression will be performed on two samples. The regression will run on a sample with all observations, and a smaller sample that only contains firms that did not hire a new CEO.
The first test is performed on the whole sample. The outcome is presented in table 7, with an R squared of 0.122. The results of the regression show that the variable POSTCRISIS has a coefficient of 0.1745. It suggests that there is a positive relationship between the use of non-financial performance measurements and the post-crisis years. However, this variable is not significant, as the p-value is 0.6. This indicates that the use of non-financial performance measurements does not significantly increases in the post-crisis years.
The only control variable that is significant in this test is INDROA. It has a coefficient of 2.861, which means a positive relationship with the use of non-financial performance
measurements. With a p-value of 0.020, it is significant at the level of 0.05.
The variable CRISISCEO has a coefficient of -1.164. This indicates that there is a negative relationship between the use of non-financial performance measurements and hiring a CEO during the crisis. This variable is significant, since the p-value is 0.02, which is significant at the 0.05 level. In other words, firms that hired a new CEO during the crisis uses less non-financial performance measurements in the CEO compensation.
Some industry variables are significant related. First of all, the ‘mining and construction’ industry has a coefficient of 1.559, which indicates a positive relationship with the use of non-financial performance measurements. The p-value of this industry is 0.01, and is therefore significant at the 0.01 level. Finally, the ‘transportation and utilities’ industry has a coefficient of 1.226. This means that there is a positive relationship with the use of non-financial performance measurements. Its p-value is 0.1, which makes it also significant at the 0.1 level. This is in line with the findings of the paper of Said et al. (2003).
23 Table 7 Coefficients (Full sample)
Regression results 1 Test with all variables
NF
Variable Coefficient p-value
(Constant) 0.174 0.040 POSTCRISIS 0.108 0.600 NEWCEO -1.164 0.740 CRISISCEO -0.656** 0.020 LOSSPREVYEAR 0.320 0.320 FIRMSIZE 0.203 0.120 LEVERAGE 0.017 0.800 INDROA 2.861** 0.020 SIC1000 0.608*** 0.010 SIC2000 0.706 0.290 SIC3000 0.699 0.380 SIC4000 0.733* 0.100 SIC5000 0.593 0.550 SIC7000 0.612 0.920 Adj./Pseudo R2 0.1219 Observations 280
A logistic regression is performed with NF as the dependent variable. The variables are explained in the methodology chapter.
Asterisks denotes the following: * significant at α = 0.10 ** significant at α = 0.05 *** significant at α = 0.01
24 The second test is performed on the sample that only consists of firms that did not hire a new CEO during 2005 and 2012. When a new CEO is appointed, the compensation contract can be easily adjusted. Without hiring a new CEO, it is more complicated to adjust the measurements in the compensation contract. With excluding this group of firms, the possible effect of the change in CEO is left out.
Because the original model includes the variables with a change in CEO, the variables NEWCEO and CRISISCEO are left out of the model. This is because there are no firms that includes new CEOs. Therefore the following model will be used:
NF = α + β1 POSTCRISIS + β2 LOSS_PREVYEAR + β3 FIRMSIZE + β4 LEVERAGE + β5 INDROA + β6 – β11 SICCAT
Table 8 shows the change in the use of non-financial performance measurements for firms that did not hire a new CEO. Just as the original sample, the use of non-financial performance measurements increased after the financial crisis. In the years 2005-2006, 14 observations contained non-financial performance measurements in the CEO compensation. After the crisis, it increased to 29 observations.
In table 9, the outcome of the test with the adjusted model and sample is shown. The results has an R squared of 0.2311.
Table 8 Use of non-financial performance measurements in CEO compensation (Sample 2)
0 5 10 15 20 25 30 35 2005-2006 2012-2013
Use of non-financial performance
measurements
25 Table 9 Coefficients (Sample 2)
Regression results 2 Test with all variables except NEWCEO and CRISISCEO
NF
Variable Coefficient p-value
(Constant) -7.494 0.001 POSTCRISIS 0.943** 0.050 LOSSPREVYEAR 0.358 0.760 FIRMSIZE 0.603** 0.030 LEVERAGE -0.013 0.620 INDROA 4.559 0.270 SIC1000 1.612** 0.050 SIC2000 1.493 0.170 SIC3000 1.054 0.310 SIC4000 2.496** 0.020 SIC5000 0.787 0.350 SIC7000 0.51 0.540 Adj./Pseudo R2 0.2311 Observations 156
A logistic regression is performed with NF as the dependent variable. Firms that appointed a new CEO are excluded from this sample. The variables are explained in the methodology chapter.
Asterisks denotes the following: * significant at α = 0.10 ** significant at α = 0.05 *** significant at α = 0.01
The results show that the variable POSTCRISIS has a coefficient of 0.943. This means that there is a positive relationship between the use of non-financial performance measurements and the post-crisis years. With a p-value of 0.05, this relationship is significant at the 0.05 level. This gives evidence that the use of non-financial performance measurements has increased in the post-crisis years.
The results also show that the control variable FIRMSIZE is significant related. It has a p-value of 0.03, which is significant at the 0.05 level. With a coefficient of 0.603, there is an
26 indication that this variable has a positive relationship with the use of non-financial performance measurements. This means that the size of a firm influences the use of non-financial performance measurements. As the relationship is positive, the size will increase the use of non-financial performance measurements.
As in the previous test, the same industries are positively and significantly related.
SIC1000, with a coefficient of 1.612, and a p-value of 0.05, which is significant at the 0.05 level. This variable refers to the ‘mining and construction’ industry. And SIC4000, with a coefficient of 2.496, with a p-value of 0.020, which is significant at the 0.05 level. This refers to the
‘Transportation, Communications, Electric, Gas and Sanitary service’ industry.
4.4 Hypothesis
The hypothesis mentioned in this study looks at the use of non-financial performance measures in the years after the crisis.
Based on the results of the first test, the hypothesis would be accepted when the coefficient of the POSTCRISIS variable has a positive coefficient. The findings show a positive relationship. However, this relationship is not significant. Therefore, the hypothesis is rejected.
The second test shows different results. The hypothesis would be accepted when the coefficient of the POSTCRISIS variable is positive. The results show a positive coefficient. Unlike the previous test, in this test the variable is significant related with the dependent variable. Therefore, for this sample the hypothesis is accepted.
27 5 Conclusion
The purpose of this study is to look at the effect of the financial crisis on the use of non-financial performance in CEO compensation. Data of the years before and after the crisis is used to
investigate whether the financial crisis has increased the use of non-financial performance
measures for the years after the crisis. Data from American firms were collected and investigated for the years 2005-2006 and 2012-2013. In each CEO compensation plan, the use of
non-financial performance measures in the CEO bonuses was examined.
The tests performed in this paper come with the following results. This test is run on two different samples. The first sample consists data of 70 firms for the previous mentioned periods. The second sample consists of only firms that did not hire a new CEO, as the expectation is that hiring a new CEO influences the choice in performance measures. For the first test, there is no significant relationship found in the use of non-financial performance measurements and the post-crisis years. However, there is evidence that firms that hired a new CEO during the crisis uses less non-financial performance measurements in the CEO compensation plan. In other words, the appointment of a new CEO during the crisis influences the use of non-financial performance measures in CEO compensation. For the second test, there is a significant relationship found in the use of non-financial performance measurements and the post-crisis years. The results of the test provide evidence that in the post-crisis years the use of non-financial performance measures has increased.
With the above mentioned results, the research question of this study can be answered. The results that are given in this paper does provide evidence that the financial crisis may affect the use of performance measures in the CEO compensation. The evidence of this effect is only seen in firms that did not appoint a new CEO. Even though there is less noise as the financial crisis is at its end, there is still an increase in the use of non-financial performance measures in the CEO compensation.
This study faces some limitations. First of all, the time period used in this study are the years before and after the crisis. The chosen time period of 2005-2006 and 2012-2013 may not be the right period to use for the pre- and post-crisis years. Also, the beginning of the financial crisis may increase the noise immediately, but the decline of the financial crisis does not have to lead to an immediate decrease in the noise. Even though the financial circumstances improves, this does not have to lead to immediate certainty in the firms. Second of all, the SEC changed the rules for proxy statements released after 2006. From this year on, firms were obliged to publish more detailed information, concerning CEO bonuses. This may influence the data, as
non-28 financial performance are much more shown in the years after the implementation of the SEC rules. Thirdly, this study only investigates whether non-financial performance measurement are mentioned in CEO compensation plans. Other ways of a change in the use of performance measurements, is a change in the weight given to financial and non-financial performance measurements. Consequently, this study does not cover all changes in performance measurements.
Future research on the use of non-financial performance measures can be done in several ways. As mentioned as a limitation, the weight of the non-financial performance measurements is not covered in this study. By looking at the change in the weight, there could be more insight given in the effect of the financial crisis on the use of non-financial performance measurements. In this study, it is shown that there is evidence of the influence of hiring a new CEO during the crisis on the use of non-financial performance measurements. This finding could be a potential research topic for future studies about the use of non-financial performance measurements. Last of all, this study finds that there is an increase in the use of non-financial performance
measurements after the crisis. Further research on the explanation of this increase gives insight of the reasoning behind this finding.
29 Bibliography
Aldamen, H., Duncan, K., Kelly, S., & McNamara, R. (2012, January 2). Performance of family firms
during the global financial crisis: Does governance matter? Retrieved March 1, 2018, from
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1976789
Bacidore, J., Boquist, J., Milbourn, T., & Thakor, A. (1997). The search for the best financial performance measure. Financial Analysts Journal, 53(3), 11.
Banker, R., Potter, G., & Srinivasan, D. (2000). An Empirical Investigation of an Incentive Plan that Includes Nonfinancial Performance Measures. The Accounting Review, 75(1), 65.
Beatty, R., & Zajac, E. (1994). Managerial Incentives, Monitoring, and Risk Bearing: A Study of Executive Compensation, Ownership, and Board Structure in Initial Public Offerings.
Administrative Science Quarterly, 39(2), 313-335.
Bebchuk, L., Fried, J., & Walker, D. (2002). Managerial Power and Rent Extraction in the Design of Executive Compensation. The University of Chicago Law Review, 69(3), 751-846.
Cordemans, N., & Ide, S. (2012). Monetair beleid in de Verenigde staten en in het Eurogebied tijdens de crisis. Economics Tijdschrift, pp. 1-27.
Davila, A., & Venkatachalam, M. (2004). The Relevance of Non-financial Performance Measures for CEO Compensation: Evidence from the Airline Industry. Review of Accounting Studies, 9(4), 443-464. Grossman, S., & Hart, O. (1983). An analysis of the principal-agent problem. Econometrica : journal of
the Econometric Society, an internat. society for the advancement of economic theory in its relation to statistics and mathematics, 51(1), 7-45.
Hoque, Z., & James, W. (2000). Linking Balanced Scorecard Measures to Size and Market Factors: Impact on Organizational Performance. Journal of Management Accounting Research, 12(1), 1-17. Ibrahim, S., & Lloyd, C. (2011). The association between non-financial performance measures in executive
compensation contracts and earnings management. Journal of Accounting and Public Policy,
30(3), 256-274.
Ittner, C., & Larcker, D. (1998). Are nonfinancial measures leading indicators of financial performance? An analysis of customer satisfaction. Journal Of Accounting Research, 36, 1-35.
Ittner, C., Larcker, D., & Meyer, M. (2003). Subjectivity and the weighting of performance measures: Evidence from a balanced scorecard. Accounting Review, 78(3), 725-758.
Ittner, C., Larcker, D., & Rajan, M. (1997). The choice of performance measures in annual bonus contracts.
Accounting Review, 72(2), 231-255.
Jensen, M., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305-360.
Kaplan, R., & Norton, D. (2005). he balanced scorecard: measures that drive performance.(The High-Performance Organization: The Best of HBR). Harvard Business Review, 83(7), 172. Mehran, H. (1995). Executive compensation structure, ownership, and firm performance. Journal of
30 Murphy, K. (1999). Chapter 38: Executive Compensation. In K. Murphy, Handbook of Labor Economics
(Vol. 3, pp. 2485-2563). Elsevier.
Murphy, K. (2012). Executive Compensation: Where We are, and How We Got There. In K. Murphy,
Handbook of the Economics of Finance. Elsevier.
Said, A., Hassabelnaby, H., & Wier, B. (2003). An empirical investigation of the performance
consequences of nonfinancial measures. Journal of Management Accounting Research, 15, 193. Sliwka, D. (2002). On the Use of Nonfinancial Performance Measures in Management Compensation.
Journal of Economics & Management Strategy, 11(3), 487-511.
Stock, J. (2014). The Economic Recovery Five Years after the Financial Crisis. Business Economics, 49(1), 21-26.
31 Appendices
List of firms used in sample
SIC industry
group Company name
1000 FREEPORT-MCMORAN INC 1000 APACHE CORP
1000 ANADARKO PETROLEUM CORP 1000 CHESAPEAKE ENERGY CORP 1000 CONOCOPHILLIPS
1000 DEVON ENERGY CORP 1000 HESS CORP
1000 MARATHON OIL CORP
1000 OCCIDENTAL PETROLEUM CORP 1000 SCHLUMBERGER LTD
2000 MONDELEZ INTERNATIONAL INC 2000 PEPSICO INC
2000 COCA-COLA CO 2000 DOWDUPONT INC 2000 JOHNSON & JOHNSON 2000 MERCK & CO
2000 PFIZER INC
2000 AMGEN INC
2000 CHEVRON CORP 2000 EXXON MOBIL CORP 3000 DEERE & CO
3000 CATERPILLAR INC 3000 HP INC
3000 CISCO SYSTEMS INC 3000 APPLE INC
3000 INTEL CORP 3000 FORD MOTOR CO 3000 GENERAL MOTORS CO 3000 BOEING CO
3000 UNITED TECHNOLOGIES CORP 4000 SPRINT CORP
4000 AT&T INC
4000 VERIZON COMMUNICATIONS INC 4000 COMCAST CORP
4000 DISNEY (WALT) CO 4000 TIME WARNER INC 4000 EXELON CORP
4000 NEXTERA ENERGY INC 4000 KINDER MORGAN INC 4000 DUKE ENERGY CORP
32 SIC industry
group Company name 5000 MCKESSON CORP 5000 HOME DEPOT INC 5000 LOWE'S COMPANIES INC 5000 TARGET CORP
5000 WALMART INC 5000 MCDONALD'S CORP 5000 CVS HEALTH CORP
5000 EXPRESS SCRIPTS HOLDING CO 5000 WALGREENS BOOTS ALLIANCE INC 5000 AMAZON.COM INC
7000 OMNICOM GROUP
7000 EBAY INC 7000 ALPHABET INC
7000 INTL BUSINESS MACHINES CORP 7000 MICROSOFT CORP
7000 ORACLE CORP
7000 AUTOMATIC DATA PROCESSING 7000 FIRST DATA CORP
7000 CAESARS ENTERTAINMENT CORP 7000 MGM RESORTS INTERNATIONAL 8000 HCA HEALTHCARE INC
8000 TENET HEALTHCARE CORP 8000 UNIVERSAL HEALTH SVCS INC 8000 QUEST DIAGNOSTICS INC
8000 LABORATORY CP OF AMER HLDGS 8000 DAVITA INC 8000 GRAHAM HOLDINGS CO 8000 URS CORP 8000 PAYCHEX INC 8000 ACCENTURE PLC
