Executive Compensation and Insurance Sector Performance: Evidence from Nigeria

This study investigates how six different measures of firm performance affect executive compensation in Nigerian insurance sector (2011-2016). The Im, Pesaran and Shin (IPS) and Kao Residual tests were employed to ascertain stationarity and cointegration of the variables. Mixed stationary and no cointegration were observed. The results indicate that profitability variables (return on asset, return on equity (ROE) and net profit margin) are insignificant to executive compensation while efficiency variables are significant to executive compensation. That is profitability variables does not have significant effect on executive compensation while corporate performance measured by efficiency variables have effect on executive compensation. It is recommended that the board should focus on efficiency measures in setting executive compensation levels as these ultimately drive profitability and corporate performance in the insurance sector. This study make a meaningful contribution to the literature as very little work has been done in this area.


INTRODUCTION
The remunerations paid to chief executive officers (CEO's) and other executives who run the affairs of firms are termed Executive compensations and these include basic salaries, bonuses, stock options and stock grants (Yusuf and Abubakar, 2014;Omoregie and Kelikume, 2017). The debate as to whether or not the perceived excessive compensations paid to corporate executives is justified by the reported corporate performance across various sectors and geographies remains inconclusive, and continues to attract the interest of researchers, academics and practitioners (Southam and Sapp, 2010;Erick et al. 2014;Omoregie and Kelikume, 2017). Omoregie and Kelikume (2017), Yusuf and Abubakar (2014) amongst others, find that compensation paid to bank executives in Nigeria is apparently not justified by the reported performance in that sector. An investigation of similar relationship in the insurance sector is logical given the critical nature of this sector to the financial system. The fact that the recent global financial crisis emanated from the inappropriate financial practices of financial executive further draws more attention to the discourse (Rajan, 2009;Fahlenbrach and Stulz, 2011).
Like the banking sector, the insurance sector is very critical to the financial sector due to its complex operations, risk impact on policyholders and the economy at large (Adams and Jiang, 2016). It is thus important to assess how executive compensation has been driven by performance in this sector.
performance. Other explanations are the Tournament Theory, and Equity Fairness Theory. This study therefore investigates the relationship between the financial performance and executive compensation in Nigeria's insurance sector. Understanding this relationship will afford a comparison with results of a similar study for the banking sector (Yusuf and Abubakar, 2014;Olalekan and Bodunde, 2015;Omoregie and Kelikume, 2017) and a better understanding of how executive compensation can be used as an instrument for driving business performance and shareholder value in the insurance sector.
The rest of this paper is structured such that section two, three, four and five present the literature, methodology, result, and conclusions respectively.

AT
The AT has over-time gained prominence in explaining the relationship structure that exists in the ownership and administrations of organizations, wherein shareholders (principal) engages the services of executives (agents) in overseeing the affairs of the business (Erick et al. 2014;Olalekan and Bodunde, 2015;Adams and Jiang, 2016). The AT hinges on the agency relationship and the dilemma that exists between both parties. While shareholders and executives are working towards a common goal of creating value, their interest may differ, thus leading to conflict of interest (Tosi et al. 2000;David, 2011;Ozkan, 2011). Mitnick (1973) in line with Eisenhardt (1989) categorised the agency problems into three: (i) the principal's problem of motivating the agents to act to achieve its goal; (ii) the agent's problem of deciding either to act in line with the principal's interest, his self-interest or direction of compromise in-case of conflict; and (iii) monitoring mechanisms and incentives to curtail discretionary behaviour of agents. The AT attempts to strike a balance between shareholder and executive interests through incentives and monitoring (Jensen and Mecking, 1976;Bosse and Phillips, 2016). A lingering problem from this is the determination of an efficient system for setting executives' incentive that would ensure they work in the interest of shareholders and improve corporate performance. Bebchuk et al. (2002) argued that optimal contracting as in the AT would hold more ground if board holds symmetric information power with the executives. They suggest that the executives wield information power that gives them bargaining power over board. Given this power, Bebchuk and Fried (2003) noted that executives are better able to negotiate compensations in their favour rather than shareholders' interest, irrespective of performance outcome. Hence, executive compensation might be insensitive to performance outcomes and shareholder value ( Van-Essen et al. 2015). Duffhues and Kabir (2008) among other studies clearly support the MPT by emphasizing the information power imbalance between the executives and the board. However, Van-Essen et al. (2015) outlined that the increased independence of board and executive in the 1990s disproves MPT as executive power has been weakened (Hall and Murphy, 2003;Conyon, 2006).

MPT
Other relevant theories are the tournament theory and the equity fairness theory. The tournament theory as proposed by Lazear and Rosen (1981) describe a reward structure that emphasizes relative ranking as against absolute performance. It hinges on the career ambition of lower level executive to achieve a performance level that is capable of propelling them to higher executive position to obtain associated reward. Given the higher compensation associated with the CEOs and senior executives, junior executives will improve effort, reduce shirking, and generally work towards improving performance and achieve a rise up the corporate ladder and earn higher compensation, thus reducing the agency risk (O'Reilly et al. 1988;Conyon and Sadler, 2001;Ridge et al. 2015). Lee et al. (2008), Ridge et al. (2015) and Elsayed and Elbardan (2018) supplied evidence in support of the tournament theory.
The equity fairness theory as proposed in (Akerlof and Yellen, 1988;Milgrom 1988) argues that quality of social relation in an organization influences performance. As such undue pay disparity among CEOs and junior executives influence counter-productive behaviour with associated adverse impact on performance and shareholder value.
Executive compensation does not often relate to firm performance and shareholder value because more often than not the talent pool for "good/experienced" CEOs and executives is scare. Thus competition amongst firms to recruit the "best talent" is intense. For this reason, huge executive packages is often used as an incentive to attract the "best talents" without any evidence of the ability of these executives to improve corporate performance. The Board of Directors that is responsible for hiring and setting the remuneration of these executives thus attract these executives with huge compensation packages in the hope that they will work to improve performance. More often than not, the result is always disappointing and no significant correlation is seen between the high compensation paid to these executives and the on-going performance of the firms. Most findings by studies within the Nigerian context support more of the AT than the managerial theory (Olaniyi and Obembe, 2017;Ogbeide and Akanji, 2016). However, the result of Omoregie and Kelikume (2017) validates the MPT more than the AT.

Empirical Review
The banking industry appears to have attracted more interest from researchers but in recent times, focus is shifting towards insurance. On a general note and contrary to AT, Ghosh (2003) and Parthasarathy et al. (2006) reported that profitability was not found to have significant affect on executive compensation. However, firm size was revealed to be a major determinant. Similarly, Lin et al. (2013) in a study of "fat cat CEOs" found that firm size as well as CEOs' tenure is a positive determinant of executive compensation. They further revealed that there is substitution effect between CEOs' compensations and CEOs' ownership level. This highlights the fact that executive pay does not depend on firm performance. Also, Duffhues and Kabir (2008) examined the pay-performance effects and they found no evidence of support. Bebchuk et al. (2002), find evidence to support the significant roles played by managerial power and they opined that market forces Electronic copy available at: https://ssrn.com/abstract=3506634 Electronic copy available at: https://ssrn.com/abstract=3506634 which serve as the background of executive compensation, is not solid enough to push for the outcome of optimal contracting theory. On the issue of insurance coverage for directors' and officers' liability, in the compensation and performance relationship, Wang and Chen (2016) embraced the notion of positive correlation between executive compensation and firm's performance. They found that the sensitivity of pay-performance relationship is lower for insured firms than is the case for uninsured firms and agency problem may worsen with the insurance. Kato and Long (2006) examined the sensitivity of executive compensation to firm performance of listed companies on the stock markets in China, USA and Japan. Their findings indicate that executive compensation of Chinese firms had a stronger correlation with performance and shareholder value than was in USA and Japan. They however submitted that their findings do not suggest however, that Chinese executives are better compensated to pursue shareholder's interest than in USA or Japan, as much of the executive compensation in USA are in the form of stock options, as against cash. Cole et al. (2016) studied the effect of executive compensation on the performance of bank-owned life insurance (BOLI). They found that executive compensation has a significant effect on the performance of BOLI. It was revealed in Kang and Han-Kim (2017) that the CEOs' compensation is positively related to their appearances on major news articles and CNBC interviews. They further showed that the relationship is stronger with companies with strong stock market performances, but weaker with large companies. However, Ogbeide and Akanji (2016) found that executive remuneration was insignificant to the performance of financial firms in Nigeria. Based on the notion that board effectiveness depends on its independence, Hermalin and Weisbash (1998) investigated the relationship between board effectiveness and its constituents. Their finding suggests that CEO-turnover negatively affect performances, which is strengthened by board independence.
Matousek and Tzeremes (2016) studied the dynamic effect of CEO's salary and bonus payments on both the technological change and technological efficiency of US banks. They found evidence of nonlinear relationship between salary and bonus payment to CEOs and the technical efficiency level of banks. As such, there is a threshold for executive pay. The study further found that bank's technological change and technical catch-up is affected positively by CEO's salaries, but negatively affected by CEO bonuses. The study further found that higher bonuses and salary payment do not always align with higher level of technical efficiency. Their findings are further supported by Ridge et al. (2015).
Similar study on Nigeria by Olalekan and Bodunde (2015) using the generalised method of moments found that bank performance, measured by their earnings per share, is negatively and significantly affected by CEO pay, measured by the annual pay of the CEO of banks. This confirms that CEO pay is indeed part of the agency problem in the Nigerian banking industry, instead of it being aligned to shareholders interest. Also, Omoregie and Kelikume (2017) employed panel vector autoregressive impulse response on the subject matter for 12 Nigerian banks. The result revealed that executive pay responds positively to equity-asset ratio and customers' deposit, while it responds negatively to their returns on equity. The result further revealed that the greater variations in executive compensation of these banks are attributed to their customers' deposit. However, higher customer deposits do not often translate to higher profits. This is attributable to the fact that with higher customer deposits and poor credit risks management systems, a lot of non-performing risk assets (loans and advances) are created by the banks thus leading to losses and poor performance. For example, the banks with the highest customer deposit to asset ratios tend to have the highest levels of non-performing loans, lower ROE, but higher relative executive compensation.
As revealed by Sakawa et al. (2017), banks' executive compensations in Japan are designed in such a manner that it provides incentives for their executives to improve their ordinary profit higher than the average regional profit level, suggesting that executive compensation serves as incentives for performance and competitiveness. Likewise, Aprilia et al. (2016) found for Indonesia that there is direct effect of cash compensation on bank performance. They further found that while earning management serves as a significant mediating variable in the compensation-performances relationship, credit risk do not. In a test of the asymmetric pay-for-performances hypothesis in Chinese banks, Cordeiro et al.(2016) found that there is an asymmetric relationship between executive compensation and firm accounting performances. The asymmetry is stronger when firm performance is above the regional median and when the accounting performance is positive. The pay-performance relationship was found to be moderated by internal governance and ownership structure.
Adopting both tournament theory and equity-fairness theory, Yu and Luu (2016) investigated the relationship between executive pay dispersion and bank performance. They found that executivepay-dispersion impacts performance negatively. As such, the higher the dispersion of executive pay, the lower the performance of banks. Their findings also supported equity-fairness theory as very low or very high pay-dispersion leads to higher performances, and bank profit is negatively affected by executive pay. In the same spirit, Ahmed and Ndayisaba (2017) studied the role of regulatory standards on CEO remunerations-performance relationship in Austrian's deposits institutions (ADIs). This study suggests that the short-term compensation of ADI-executive as regulated by the authorities appears closely related to performance.
Particular to the insurance sector, Adams and Jiang (2016) emphasized the role of external directors on the performance of insurance companies in the United Kingdom. Their findings however suggest that the proportion of outsiders in the board does not affect the financial performance of firms, rather, the financial expertise of both the internal and external board members impacts financial performances significantly. Adams and Jiang (2017) found that in addition to financial expertise of board members, CEOs' experiences also improve the financial performances of their firms. However, CEOs' age and power appears insignificant Electronic copy available at: https://ssrn.com/abstract=3506634 Electronic copy available at: https://ssrn.com/abstract=3506634 to financial performances. Meanwhile, an earlier study by Erick et al. (2014) investigated the relationship between executive pay and financial performance of insurance companies. They found that the relationship between executive pay and financial performance is insignificant, suggesting that financial ratio are uninfluential in determining executive pay. However, Sun et al. (2013) find otherwise as they revealed that financial performances with cost efficiency of US property-liability insurance firms positively impact CEO's compensation.
Besides dearth of literature on executive pay-performance relationship in the insurance industry, the sensitive nature of the industry within the financial system and the economy calls for attention, especially in Nigeria. This study tends to look into this relationship and at the same time test the non-contemporaneous hypothesis of the pay-performance relationship. By noncontemporaneous, we mean that executive compensation might not depend on present performance level of firms, but on previous level of performances.

METHODOLOGY
In order to test the pay-performance relationship in the Nigerian insurance industry, this study adopts a simple pay-performance model following Jiang (2016, and, expressing executive pay as a function of firm performance indicators. Return on assets (ROA), calculated as profit after tax divided by total assets; ROE computed as profit after tax divided by equity; Net profit margin (MARGIN) computed by dividing earnings by gross premium written; Solvency position (SOL) measured as one minus surplus, divided by total assets; Loss ratio (LR) as ratio of total incurred claims to total earned premium; and Combine operating ratio (COR) as the sum of total incurred claims and expenses, divided by total earned premium, are adopted as proxy for financial performance. The relationship is expressed in a semilogged form as

Unit Root and Cointegration Results
To test and ascertain the stationarity properties of the six variables used, this study employs the IPS unit root test. The result, estimated at both intercept, alongside intercept and trend specifications are presented in Table 1. Using the intercept specification, MARGIN, LR and COR tends to be stationary at level, while ROA is stationary at level when tested at intercept and trend. As expected, a better performance of stationarity is arrived when the variables are tested at first difference. All the variables are stationary at first difference when tested at intercept. However, not all the variables are stationary when tested at intercept and trend. This lends the idea that the model that will be estimated in our regression will only have an intercept, and not have a trend term. We can therefore conclude that our variables are first differenced stationary. Results presented in Table 2 show KAO test and given a t-statistics of −3.9802 and probability value of 0.0000, null hypothesis of no cointegration will be rejected to embrace alternative of cointegration among the variables.

Contemporaneous relationship
This study focuses on the pay-performance relationship of insurance firms in Nigeria and adopted six performance indicators to deepen our understanding. Table 3 presents the analysis of contemporaneous pay-performance relationship. The table shows that ROE, LR and combined operating ratio (COR) all have negative relationships with executive performance. However, ROE was revealed to be insignificant and only LR and COR proves to be significant to executive pay. On the other hand, return on asset (ROA), net profit margin (MARGIN) and solvency position (SOL) have positive relationship with the executive pay. But the result shows that ROA and MARGIN appear insignificant to executive pay as only SOL is significant.

Non-contemporaneous relationship
To account for the possibility of non-contemporaneous payperformance relationship, this study also estimated a model of lagged performance variables as presented in Table 4. This is hinged on the fact that the pay accrued to executives of insurance firms might not be a function of the present performances of the firm, but on past performances. It is worth noting that only COR retains its sign and relevance in terms of significance. It appeared to have a negative relationship with executive pay.
Electronic copy available at: https://ssrn.com/abstract=3506634 Electronic copy available at: https://ssrn.com/abstract=3506634 The adjusted r-squared of both the contemporaneous and non-contemporaneous results shows that approximately 83% of executive compensations are determined by the six firm performances indicators adopted. At the same time, the significant probability values of the F-statistics reveals that all the variables are jointly significant, at 1% significance.

DISCUSSION
Following Duffhues and Kabir (2008) argument of lagged response of executive compensation in response to performance, this study attempt to investigate the impact of performance on executive compensations. Though a number of the performance variables used in this study appear significant to executive compensation at the contemporaneous level, all except COR affects executive pay in the non-contemporaneous model. Major variables popularly used to assess profitability, return on asset, ROE and net profit margin appeared insignificant at both contemporaneous and non-contemporaneous level. However, management efficiency variables such as LR, solvency ratio and combine operation ratio are revealed to be significant, especially at the contemporaneous level. Meanwhile only combine operation ratio appeared significant in the non-contemporaneous level.
The implication of this is that profitability does not drive executive compensation in Nigeria's insurance sector but efficiency does. This suggest that insurance companies in Nigeria emphasize the efficiency of their business in terms of sustenance over profitability and given the nature of the risk associated with the sector, ability of executives to sustain the insurance business given the inflow of premium and outflow of indemnity. Intuitively, solvency ratio was revealed to positively affect executive pay; as the executive is able to ensure a high solvency rate the insurance business remain a going-concern and executive are continuously rewarded. Also, as established by negative effect of LR on executive pay, LR is associated with depleting efficiency in terms of managing premium and indemnity and executive pay depletes too. Findings for COR is also intuitive; negative effect on executive compensation suggests that increasing COR is associated with falling profitability from underwriting, as such executive pay falls. This finding is insightful for board of directors to actually capture the variable that matter to executive pay-performance relation in executive performance appraisals and makes them realized that the different categories of performance indicator that are important for executive compensation decision. This study is limited by the fact that it focuses on the insurance sector. Also for data availability reasons, this study needs to be expanded to the financial sector as a whole to better capture how event turns out for all financial institutions.

CONCLUSION
This study looked into the pay-performance relationship in the Nigerian insurance industry. The study made use of annual data from eleven (11) insurance firms spanning from 2011 to 2016.
The IPS unit root test was employed to test for the stationarity properties of the variables, while the KAO residual cointegration    test was used to ascertain the long run properties of the variables. The result shows that the variables are first differenced stationary, the KAO residual cointegration shows that long-run relationship exists among the variables used.
This study adopts a two-stage approach using contemporaneous and non-contemporaneous estimation following Duffhues and Kabir (2008) who argued for lagged response of executive compensation to performance. The results show that profitability variables (return on asset, ROE and net profit margin) are insignificant to executive compensation while efficiency variables (solvency ratio, LR) are significant at contemporaneous level and combine operation ratio is significant at both contemporaneous and non-contemporaneous level. The implication is that profitability variables are not important in the decision of executive pay in insurance companies in Nigeria while efficiency variables appear the most important. Board members are advised to be watchful in executive performance appraisals to capture the indicators that matter. Further studies are required on the financial market as a whole.