1
THE RISK TAKING BEHAVIOR OF A STOCKHOLDER-OWNED LIFE INSURANCE
Hui-Hsuan Liu, National Cheng Kung University, Taiwan, NO. 1, University Road, Tainan City, Taiwan
701, Tel: +886-928994582, mimi73718@gmail.com
Yung-Ming Shiu, National Cheng Kung University, Taiwan, NO. 1, University Road, Tainan City,
Taiwan 701, Tel: +886-6-2757575 ext. 53330, yungming@mail.ncku.edu.tw
ABSTRACT
Prior research has not been well examined the firm-level and country-level corporate governance to risk-taking behavior. Data have used on 471 life insurance companies in 13 EU countries from 1995 to 2006. We prove that life insurers’ risk-taking behavior increases with firm-level corporate governance, further examine the interaction effect between the firm-level and country-level to risk-taking behavior respectively. Our argument is the same as the Doidge, Karolyi and Stulz (2007), under high economic freedom countries, life insurers with high firm-level corporate governance has more opportunities to take risk. However, the result is mixed depending on various indexes of country-level governance.
Keywords: European Union, Financial Risk Assessment, Insurance, Corporate governance, Economic Freedom, Risk-taking Behavior
INTRODUCTION
Since the sub-prime mortgage crisis happened, the international market tumble was consequent on the bankruptcy of Lehman Brothers in September 2008. Under the financial tsunami, bank and insurance industry also got involved. For example, American International Group Inc. (AIG) endures the liquidity crisis when its credit ratings were downgraded below "A-" levels in September 2008. Unlike other industries, insurance industry has developed a number of complex products and at the same time has had to reduce the profit loadings in these products to compete in the marketplace (Cummins Phillips Smith, 1997). In response, they undertake the public interest and loss. As long as the financial difficulty bursts out, the effect is the most serious and far-flung than all financial institutions. Especially for the bankruptcy of insurers, they have more significant costs. Overall, the financial storm induces us to review the risk-taking behavior of life insurers.
While most studies suggest that the main factors to risk-taking behavior include ownership structure (Jensen and Meckling, 1976; Staking and Babbel, 1995; Cummins and Sommer, 1996; Chen, Steiner and White, 2001), CEO negation power (Hermalin and weisbach, 1998 and Adams, Almeida and Ferreira, 2005) and board composition (Mayers et al., 1997; Fama, 1980; Fama and Jensen, 1983; He, 2007). Based on the agency theory, managers have competing agency relationships with shareholders and policyholders. In the modes of governance of stockholder-owned life insurance,1 shareholders direct financial flows within the company away from policyholders and towards themselves, especially large shareholders, can utilize more pressure over managers than small disparate policyholders. In addition, while shareholders can diversify theirs’ risk in capital market, managers often attempts to change firm risk by investment a lot projects. For their own private benefits, managers may be conservative and avoid taking risks in investments, including value-enhancing ones (Hirshleifer and Thakor, 1992), especially for they only receive fixed salaries, they would not acquire additional reward as who compensated with shares and share option (Pathan, 2009). However, Staking and Babbel (1995), Cummins and Sommers (1996) and Chen, Steiner and White (2001) advanced by is the wealth transfer hypothesis versus risk aversion hypothesis,2 they argue that if managers enlarge the proportion of ownership, theirs’ behavior and interest will align with shareholders’ interests and have strong motivation to maximize theirs’ value by increase the level of risk. In the case of life insurance companies, shareholders always own significant cash flow rights or votes. Insiders are pressured into taking risk for shareholders’ interests. Shareholders have more incentives and power to avoid being expropriated by managers. Attig, Guedhami, and Mishra (2008) find that mainly in East Asian firms, multiple shareholders structures exert an internal governance role to control private benefits and to reduce information asymmetry, they actively take risks in investments. Overall, we propose high shareholder independence degree leads insurers to engage in high level of risk taking behavior.
Numerous researches use chief executive officer (CEO) duality to proxy corporate governance. Pathan (2009) refers to the Fama and Jensen (1983), Weisbach (1998) and Hermalin and Weibach (2003) and documents bank risk-taking behavior is
1 Stock and mutual are the two major ownership structures in the insurance industry. In the case of stock companies, shareholders as principals who employ managers to act as their agents in the running of the company, policyholders to act as their agents in the management of risk and the provision of financial intermediary serviced. The mutual companies, policyholders and shareholders are the same, the mutual insurance companies which are not exposed to the market for corporate control. However, in the European insurance industries, the mutual structure in many countries mainly focuses on the different kinds of insurance (e.g. mutual life insurance company has significant market share in United Kingdom; mutual property-liability insurance company has mainly been engaged in France; in Germany, life and property-liability insurance have been played an equally important role). The mutual structure of life insurance companies is more difficult to examine the risk-taking behavior in the European insurance industries, so we examine for a sample of stockholder owned life insurance companies in the European.
2
Advanced by risk aversion hypothesis, Smith and Stulz (1985) affirm that if the ownership shares of managers’ increases they become increasingly risk averse and are more likely to purse hedging and other risk reducing strategies. It implies that even managers realize engaging excessive risk taking would increase the market value of owner’s equity, they only care about the their long run compensation. The optimal long run compensation depends on the survival of insurance firms. Therefore, they become increasingly risk averse when as the ownership share of manager’s increases.
2
positively related to strong board while negatively related to CEOs power. However, Wen and Chen (2008) show that CEOs with the dual role as a chairman of the board has a negatively impact on the firm’s risk strategy. In the life insurance industries, CEOs is likely to be the person with the most power and influence within the company. If CEOs are part of the board, they often exploit theirs’ power and negotiate with their board of directors to seize some private benefit, furthermore, they have more power to make risk taking behavior decision (Adams, Almeida, and Ferreira, 2005). So CEOs duality is an important factor to affect risk taking behavior making.
Regarding the determinant of corporate board have found that board size (Boone, Field, Karpoff, and Raheja, 2007; Hermalin and Weisbach, 2003; Jensen, 1993; Dalton et al., 1999; Yermack, 1996), shareholders concentration (La Porta et al., 1998; Guedhami and Mishra, 2008) and board independence (Fama, 1980; Fama and Jensen, 1983; Hermalin and weisbach, 1988 and He, 2007). Among these variables, board size is the most closely related to several firm operating and industry characteristics. Boone, Field, Karpoff, and Raheja (2007) indicate the three hypotheses: scope of operations hypothesis, monitoring hypothesis and negotiation hypothesis are important to the firm attributes. 3 Hermalin and Weisbach (2003) suggest that a small level of board size allows for effectively monitoring and improves firm performance. Jensen (1993) also indicates that larger boards face the problem of coordination and director free-riding, they could be less effectively than smaller boards. Although with twofold arguments of board size, we prefers to the first view is scope of operations hypothesis, because stockholder-owned life insurance companies generally have many different kinds of insurance policy, they can diversification risk and have ability to make more risky investment decision. Overall, we use the internal governance mechanism (shareholder ownership, CEOs and board size) to be the firm-level corporate governance, arguments hypothesis one can be stated as:
H1: Other things equal, higher firm-level corporate governance leads life insurers to engage in higher level of risk taking behavior.
Insurance industry is a risk intermediary, they have a lot of complex products in different financial markets spread over the world. Insurers are more extremely regulated to prevent from any systematic risk and to protect the interest of policyholders. The country-level legal protection can facilitate better monitoring of financial activities including theirs’ risk-taking behavior. Booth (2007) argues that insurance industry have bilateral informational asymmetries characteristic, the regulation can prevent the adverse affects of information asymmetries for illiquid contracts. Furthermore, regulation can be a protection mechanism to guarantee insurers commit to their contracts. In addition, Gonzalez (2005), Konishi and Yasuda (2004) and Park (1997) claim the importance of regulation in banks. About the proxy of country-level legal, John, Litov, and Yeung (2008) refer to the La Porta et al. (1998) and use the investor protection index to examine the relationship between corporate governance and risk-taking. However, several authors (Pagano and Volpin, 2005) criticize the investor protection proxy with conceptual ambiguities. Therefore, we follow Djankov et al. (2008, p.441) for using anti-self-dealing index to be my control variable and to express the legal protection of minority shareholders against expropriation by corporate insiders. In sum, corporate governance varies widely across countries and across firms. Doidge, Karolyi and Stulz (2007) find that country characteristics explain much more of the variance in governance rating than observable firm characteristics. It is important to examine the corporate governance to affect risk-taking behavior by country characteristics of governance.
Economic freedom is a part of country characteristics of governance. The level of economic freedom differs from one country to another, it has been developed and used in their productive abilities, exchange goods and services with others, compete in markets, and keep the fruits of their labor (Gwartney and Lawson, 2007). In generally, high index of economic freedom, it is means that the country’s regulation is the freest. As shown in Figure 1 form the Heritage Foundation Index of Economic Freedom, it shows that after year 2001, the index of Europe started to exceed the index of Americas. So it is important to understand that how to affect life insurance companies risk-taking behavior decision in the freest country. Therefore, the influence of economic freedom in Europe should be taken into considered.
(Insert Figure 1 about here)
McMullen, Bagby, and Palich (2008) document that governance restriction of economic freedom appear to impact different entrepreneurial activities depending on the particular freedom restricted by government and the entrepreneurs’ motivation for engaging in entrepreneurial action. Firms within high economic freedom countries have more opportunities to get different businesses, the policy and legislative power force firms to take more risky behaviors. In addition, Berggren (2003) arguments several reasons which relate to institutions guarantee economic freedom plausibly have capacity to provide the kinds of growth-enhancing incentives: first, they promote a high return on productive efforts by low taxation, an independent legal system and the protection of private property; Second, they enable talent to be allocated to where it can generate the highest
3
The first hypothesis implies that board structure is driven by the scope and complexity of the firm’s operations (Fama and Jensen, 1983 and Coles, Daniel and Naveen, 2008). The second hypothesis implies that degree of information asymmetry and the size monitoring cost determine the board size (Demsetz and Lehn, 1985; Raheja, 2005). The third hypothesis implies that board composition results from a negotiation between CEO and firm’s outside board members (Hermalin and Weisbach, 1998).
3
value (Murphy, Shleifer, and Vishny, 1991); third, because experimentally organized economy in which have a large amount of business trial and error can be take placed (Johansson, 2001). Lastly, they facilitate predictable and rational decision-making by a low and stable inflation rate. Thus, legislation limits the firms’ operation and may influence theirs’ risk taking behavior. The difference of economic freedom helps us to understand why firms are financed and owned so differently in different countries. In sum, we argue that country-level economic freedom interacts with firm-level corporate governance, life insurers actively take risky behavior in a country with a higher level of economic freedom. Overall, the hypothesis is stated as follows:
H2: Other things equal, the country-level of corporate governance by economic freedom is positively correlated with life insurers’ risk-taking if the firm-level corporate governance is higher.
Several financial structure variables that are likely to affect risk-taking behavior are also included as control variables. The insurance specific variables including size, consolidated, leverage and reinsurance are also considered. There have competing hypotheses are relate to the size, the bankruptcy cost of small firms is greater than larger firms from the financial distress viewpoint. However, Colquitt and Hoyt (1997) argue the size is negatively related to risk-taking behavior from the economics of scale viewpoint. John, Litov, and Yeung (2008) tabulate a significantly negative relation between risk-taking behavior in banking industry. So the effect of size is indeterminate. In addition, in view of the consolidated firm have more resource and capital; consolidated insurers have more possibility to take more risky decision. Furthermore, in life insurance industry, the policyholders with the fixed long-term claims act as the bondholders. High leverage increases costly contracting. The residual claimants increase their value of claim for diminishing the outstanding debt. Higher leverage life insurers have higher probability to face underinvestment problem. Therefore, negative relation between risk-taking and leverage is expected. Then, an insurer may also rely on the reinsurance and involves in the risky behavior. Even the use of reinsurance can be a sign that the insurers have more tendencies to avoid risks.
We believe that this study contributes to broaden literatures on the following ways. First, prior literature (e.g., John, Litov, and Yeung, 2008) have used the investor protection to proxy country-level corporate governance and has empirically examined the relationship between corporate governance and risk-taking. However, the investor protection proxy they have used with conceptual ambiguities, we follow Djankov et al. (2008, p.441) and use the completeness of anti-self-dealing index to be my control variable and to express the legal protection of minority shareholders against expropriation by corporate insiders. Second, few studies directly investigate the relationship between corporate governance and risk taking behavior in the insurance industry. Chen, Steiner and White (2001) show that life insurers’ risk taking behavior increases with managerial ownership. However, the paper only considers one aspect of corporate governance. In this study, we use a number of measures to characterize each life insurers’ corporate governance. Third, with regard to the corporate governance and risk taking behavior, especially in the European life insurance industry, the research literature is very limited. Because the data of European life insurance is difficult to get, we get the data which relate to life insurances’ financial statements from the Eurothesys database, moreover, through the ISIS database to get the insurance financial information. Fourth, we further examine the country-level corporate country-level corporate, and argue that stockholder-owned life insurance in high economic freedom countries has more opportunities to take risk, the argument consist with the same as the Doidge, Karolyi and Stulz (2007). Lastly, to our knowledge, literature has documented that risk-taking behavior is a determinant of firm-level corporate governance. However, the country-level corporate governance country-level corporate has not been well examined and empirical evidence varies. Moreover, no prior research has examined the interaction between country-level and firm-level corporate governance to risk taking behavior respectively.
Two prior studies that are closely connected to ours include Lai and Lin (2008) and Chen, Steiner and White (2001). However, several major differences exist. They focus on the determinants of the board structure for property-liability insurers and managerial ownership for US life insurance industry respectively. However, the board structure and managerial ownership are considered as only part of corporate governance. Furthermore, both studies all have not considered the country characteristics of governance. Therefore, we use data from life insurers and include completely characteristics of firm-level governance; furthermore, we examine the interaction between country-level and firm-level corporate governance within cross country data to gauge the relationship.
THE METHODOLOGY AND EMPIRICAL FRAMEOWRK
As discussed above, we examine the relationship between corporate governance and risk-taking behavior, and then investigate the interaction effect of different levels of economic freedom on the relationship of corporate governance and risk-taking behavior. We therefore construct these equations as follows:
t i t i t i t i t i t i
CV
Risk
, , 1 , 4 , 3 , 2 , 1 0,
Board
size
CEOs
duality
Shareholde
r
independen
ce
degree
1
ε
β
β
β
β
β
+
+
+
+
+
=
(1)4 t i t i t i t i t i t i t i t i t i t i t i
CV
Risk
, , 2 , , 2 5 , , 4 , 3 , , 2 , 3 , 2 , 1 0 ,freedom
Economic
degree
ce
independen
hareholder
freedom
Economic
duality
CEOs
freedom
Economic
size
Board
degree
ce
independen
r
shareholde
duality
CEOs
size
Board
2
ε
β
β
β
β
β
β
β
β
+
+
∗
+
∗
+
∗
+
+
+
+
=
(2)Where
Risk
i,t denotes the earnings volatility of life insurer i in year t, we follow John, Litov and Yeund (2008) to compute it to represent the firm-level and country-level risk-taking behavior respectively.Board
size
i,t,t i,
duality
CEOs
and t i,degree
ce
independen
r
Shareholde
are the firm-level of corporate governance mechanism in the life insurance industry.t i,
freedom
Economic
is the economic freedom index of life insurer i in year t, it represents the country characteristics of governance mechanism. Furthermore, we examine the interaction effect by the eight dimensions of economic freedom: Economic Freedom (ECFR), Business Freedom (BUFR), Fiscal Freedom (FCFR), Government Size (GOSI), Monetary Freedom (MOFR), Investment Freedom (INFR), Financial Freedom (FIFR) and Labor Freedom (LAFR).t i
CV
1,, andt i
CV
2,, are two different sets of control variables which can be identified effect the firm-level and country-level risk-taking behavior.ε
1,i,t andε
2,i,t are the error term.Data description and diagnostics
We mainly use the Eurothesys and ISIS database to retrieve the data for all variables which used in this research. The Eurothesys include financial statements of insurers in 76 countries all over the world. The ISIS database can reveal detail insurance financial information: audit, shareholder ownership, manager’s name etc. all over the world. The study employs an unbalanced sample of yearly-based panel database of 471 stockholder-owned life insurance companies in 13 European Union (EU) countries.4 Table 1 presents the number of sample and economic freedom index in every country.
(Insert Table 1 about here) Measure of variables
We expect that corporate governance influence risk taking behavior. We first examine the corporate governance and risk-taking, then investigate the interaction effect of different levels of legal protection on the relationship of corporate governance and risk-taking. Prior research suggests a number of factors which may affect the relation between corporate governance and risk-taking. These factors in both equations are examined as follows and a list of variables and their definitions are described in Table 2.
(Insert Table 2 about here)
The measure of Risk-taking, adopting risky investment behavior leads to have more volatile returns on capital, we follow John, Litov, and Yeung (2008, p.1687-1693) and utilize two proxies for the degree of risk-taking in firms’ operations based on the volatility of corporate earnings: (i) the market-adjusted volatility of firm level earnings over the sample period from 1996 to 2005; (ii) a country average of the volatility of firm earnings. To calculate risk 1, we gather each firm’s earnings before income tax and total asset data from Eurothesys. We keep the company if they provide more than six years EBIT and total asset data over the 1995 to 2006 period. First, we divide EBIT by total asset each year. Second, we average all the EBIT / Total Asset ratio. Then, the average is subtracted by each year’s EBIT / Total Asset ratio. Finally, the calculated deviations of each firm would have one standard deviation. This standard deviation is viewed as the firm-level risk. Thus, every company has a specific risk value regardless the year. This method is conducted to cope with the data disclosure. The data of corporate governance and legal protection only reveal the latest number or a specific index. Risk 2 is the country-level risk, we use the average Risk 1 in every country to express Risk 2.
The measures of explanatory variables, we inspired by the concept of Pathan (2009, p.1340-1343) and La Porta et al. (1998, p.1145-1150), the corporate governance is gauged by shareholder independence degree, CEOs duality, and board size. La Porta
4
Some EU countries are excluded. For instance, Cyprus, Czech, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Romania, Slovak Republic, and Slovenia are excluded because their life insurers’ information cannot be retrieve from our data sets. In addition, Ireland and Luxembourg are also excluded for lack of legal protection indices on the research of La Porta et al. (1998).
5
et al. (1998) collected data on 10 largest nonfinancial firms in 49 countries and computed the combined (cash flow) ownership stake of its three largest shareholders. First, we get the data of shareholder independence degree from ISIS database. According to the definition of ISIS database, the independence indicator categorizes companies into four groups: A means no shareholders recorded with more than 25% direct or total ownership. That is, A represents the independent companies. B is attached to any company with a known recorded shareholder, none of which with an ownership percentage over 50%, but having one or more shareholders with an ownership percentage above 25%. C means no shareholders recorded with more than 50% direct ownership. One shareholder recorded with more than 50% total ownership. That is, these companies can be viewed as indirectly majority owned firms. While the companies which classified into D is the directly majority owned. One shareholder recorded with more than 50% direct ownership. We transfer A, B, C, and D into 1, 2, 3, and 4. In addition, CEOs duality is used to capture the influence of CEOs over the board decision making. CEOs duality equals one when CEOs duality takes place. In other words, CEOs duality equals one if the CEOs is also the chair of the board or if CEO is internally-hired. Moreover, since an individual director has little incentive to search information and monitor the managers within large board. Free-riding director problem results in a relative weak board. Therefore, we check the report on ISIS database and calculate the board size (BS). Board size only includes the number of directors on the board, the secretary or accountants on the board are eliminated.
About the measures of economic freedom, the index is published by the Heritage Foundation and available for each country since 1995. The Index of Economic Freedom is a simple average of 10 individual freedoms, each of which is vital to the development of personal and national prosperity. Each dimension has a score between 0 and 100 percent. The ten dimensions of economic freedom include: Economic Freedom (ECFR), Business Freedom (BUFR), Fiscal Freedom (FCFR), Government Size (GOSI), Monetary Freedom (MOFR), Investment Freedom (INFR), Financial Freedom (FIFR) and Labor Freedom (LAFR). The business freedom (ECFR) score for each country is a number between 0 and 100 percent, with 100 equaling the freest business environment. The score is based on 10 components, all weighted equally, based on objective data from the World Bank’s Doing Business study: Starting a business—procedures (number); Starting a business—time (days); Starting a business—cost (% of income per capita); Starting a business—minimum capital (% of income per capita); Obtaining a license—procedures (number); Obtaining a license—time (days); Obtaining a license—cost (% of income per capita); Closing a business—time (years);Closing a business—cost (% of estate); and Closing a business—recovery rate (cents on the dollar). Fiscal freedom (FCFR) is composed of three quantitative components in equal measure: the top tax rate on individual income, the top tax rate on corporate income, and total tax revenue as a percentage of GDP. Scoring of the government size factor (GOSI) is based on government expenditures as a percentage of GDP. Monetary freedom (MOFR) score combines two components: the weighted average inflation rate for the most recent three years and price controls. Investment freedom (INFR) is an assessment of the free flow of capital, especially foreign capital. Index with 100 percent equals foreign investment (FI) is encouraged and treated the same as domestic investment, with a simple and transparent FI code and a professional, efficient bureaucracy. The financial freedom factor (FIFR) measures the relative openness of each country’s banking and financial system. The authors score this factor by determining the extent of government regulation of financial services; the extent of state intervention in banks and other financial services; the difficulty of opening and operating financial services firms (for both domestic and foreign individuals); and government influence on the allocation of credit. The authors use this analysis to develop a description of the country’s financial climate and assign it an overall score between 0 percent and 100 percent. Labor freedom (LAFR) composites four quantitative components are equally weighted as 25 percent of the labor freedom factor: minimum wage, rigidity of hours, difficulty of firing redundant employees, and cost of firing redundant employees.
We also control the legal protection to shareholders. Anti-self-dealing is a revised estimate of the anti-director rights index which reveal from La Porta (1998, p.1130, 1142). The data based on their questionnaire received from 102 firms. Djankov et al. (2008, p.432) contend that anti-self-dealing index is ‘an index of the strength of minority shareholder protection against self-dealing by the controlling shareholder’. Anti-self-dealing is designed as the average of ex ante and ex post private control of self-dealing. Principal component of: approval by disinterested shareholders, disclosures by Buyer, disclosures by Mr. James, independent review, each of the elements in the index of disclosure in periodic filings, standing to sue, rescission, ease of holding Mr. James liable, ease of holding the approving body liable, and access to evidence. In addition, the insurance-specific variables including size, leverage, and reinsurance are also considered. To measure firm size of the insurer is the natural logarithm of the insurer’s gross premiums written. Different from other industry, we use the gross premiums written instead of total asset. From the financial distress viewpoint, the bankruptcy cost of small firms is greater than larger firms. The size is negatively related to risk-taking behavior. On the other hand, from the economics of scale viewpoint, the size is negatively related to risk-taking behavior (Colquitt and Hoyt, 1997). John, Litov, and Yeung (2008) tabulate a significantly negative relation between risk-taking behavior and the natural logarithm of initial corporate asset in banking industry. Owing to the competing hypotheses, the effect of size is indeterminate. In addition, in life insurance industry, the policyholders with the fixed long-term claims act as the bondholders in other industry. High leverage increases costly contracting. The residual claimants increase their value of claim for diminishing the outstanding debt. Higher leverage life insurers have higher probability to face underinvestment problem. Therefore, negative relation between risk-taking and leverage is expected. In this study, we define leverage as total liability divided by surplus. Moreover, the measure of an insurer’s reinsurance activity is the ratio of reinsurance ceded to total direct premiums plus reinsurance assumed. Insurers diversify the risks through reinsurance,
6
remaining the retention to be the consideration. An insurer may also rely on the reinsurance and involves in the risky behavior. Even the use of reinsurance can be a sign that the insurers have more tendencies to avoid risks, in this study, we assume that reinsurance is positively relevant.
EMPIRICAL RESULTS Univariate analysis
The sample is comprised of 471 stockholder-owned life insurance companies in 13 European Union (EU) countries during the period from 1995 to 2006. We separately present the summary statistics with regard for all variables in Table 3. In addition, we show a Pearson correlation coefficient matrix for all the variables, which indicates the strength and direction of the linear relationship between them. The results show a significant relation between corporate governance, risk taking behavior and the relevant control variables. In addition, the variable of risk taking is significantly related to the relevant control variables. Moreover, all the coefficient values for all the control variables in this study are less than 0.5, and this indicates that there is no collinear relationship between them.
(Insert Table 3 about here)
Multivariate analysis
The relationship between firm-level risk-taking behavior and corporate governance, Table 4 presents OLS estimates of the determinants of the firm-level risk-taking. The importance of board size, CEO duality and shareholder’s ownership concentration leads to significant positive relationship with firm-level risk-taking. In addition, the positive relationship between Anti-self-dealing index and risk-taking behavior interprets that high hurdles to self-dealing induce the investors to eliminate the fear of being expropriated. In our data sample, 72% firms are independent company. Leverage is negatively related to risk-taking behavior. In life insurance industry, the policyholders with the fixed long-term claims act as the bondholders in other industry. Higher Leverage life insurers have higher probability to face underinvestment problem. Therefore, the negative relation result is as expected. Reinsurance is negatively related to the risk-taking behavior. The result may due to the use of Reinsurance is a sign that the insurers have more tendencies to avoid risks.
(Insert Table 4 about here)
Furthermore, the interaction effect of economic freedom on the relationship between country-level risk-taking behavior and corporate governance, Table 5 reports the regression results of ordinary least squares (OLS) estimates of country-level risk-taking behavior from the economic freedom. Board size, Shareholder independence degree and CEOs duality respectively have significant opposite effect in countries with a high and low level economic freedom. The different of economic freedom affect the insurer’s risk-taking behavior decision. In addition, Shareholder independence degree and CEOs duality are statistically significant to the all dimensions of economic freedom. This illustrates that, under the high economic freedom environment, the shareholder ownership has more positive impact on the risk-taking behavior. Furthermore, Board size is only not statistically significant to the two dimensions of economic freedom (LAFR and FIFR). In the closely-held companies, shareholders’ profit or lost is highly rely on the firm’s risk-taking behavior and performance. Although they expect abnormal return from risky investment, shareholders in closely-held company have more opportunities to suffer unsystematic risk. This result is also consistent with corporate bank evidence by John, Litov, and Yeung (2008).
With regards to Firm size, at the outset, we assume that firms affiliate with groups have more resource and investment opportunities to take risk. Large Firm size would lead managers to undertake risky behaviors in countries. The statistically significant positive coefficient on Firm size indicates that no matter the Leverage of economic freedom, the large firms have more preference to take risk for their abundant financial support and investment opportunities. With regards to Leverage, as anticipated the coefficient on Leverage is negative and statistically significant. Reinsurance is positively related to insurance risk-taking behavior and statistically significant. The result conveys that an insurer may also rely on the Reinsurance and involves in the risky behavior. With respect to anti-self-dealing index, as anticipated the coefficient on Anti-self-dealing index is positive across all equations and statistically significant. This illustrates that better legal protection for shareholders, more obstacles the corporate insiders would face when converting corporate benefits back to themselves. With less fear of being expropriated by managers, shareholders would urge corporate insiders to take risky but value-enhancing investment. At odds with the expectation, insurers affiliated to consolidation prefer more risk-taking behavior. This result is consistent with the relationship between Firm size and risk-taking behavior. We assume firms operate under group have more financial support and investment opportunities would have same relationship with risk-taking behavior as the high gross premium written insurers.
7
(Insert Table 5 about here) CONCLUSIONS
The paper not only the firm-level corporate governance and risk taking behavior, but also examine the interaction effect between the firm-level and country-level to risk-taking behavior respectively. First, we prove that life insurers’ risk taking behavior increases due to well firm-level corporate governance. Life insurers risk taking behavior increases with larger board size and high CEOs duality and shareholder independence degree. It implies that corporate governance which influenced by the Board size, Shareholder ownership concentration and CEOs duality is an important determinant of insurance risk-taking. Furthermore, we found the corporate governance proxies related to insurance risk-taking behavior consistent with the insurance contract environment. Given that better corporate governance is instrumental to insurance risk-taking, firms in high economic freedom countries have more opportunities to take risk. The argument is the same as the Doidge, Karolyi and Stulz (2007), stockholder-owned life insurance in high economic freedom countries has more opportunities to take risk. However, the result is mixed depending on various indexes of country-level corporate governance.
In addition, large and have Reinsurance life insurance companies have ability to undertake risky behaviors. Furthermore, the result of coefficient on Leverage is significant negative to the risk-taking behavior; it is consist with my expectation. High Leverage increases costly contracting. The residual claimants increase their value of claim for diminishing the outstanding debt. Higher Leverage life insurers have higher probability to face underinvestment problem. However, consolidated insurers may have a lot opinion and could not easily to take more risky decision making.
We recognize that findings in this paper are subject to several limitations. ISIS database provides the corporate governance information such as shareholder ownership, board composition and the name and position of board members at the latest disclosure date. The panel data like gross premium written and liability need to transfer into one specific value in one firm. We follow the method of John, Litov, and Yeung (2008, p.1688), take the average value to code variables.
8
REFERENCES
Adams, R.B., H. Almeida, and D. Ferreira, 2005, Powerful CEOs and their impact on corporate performance, Review of Berggren, N., 2003, The benefits of economic freedom, The Independent Review, 8: 193–211.
Boone A. L., L. C. Field, J. M. Karpoff, and C. G. Raheja, 2007, The determinants of corporate board Firm size and composition: An empirical analysis, Journal of Financial Economics, 85: 66-101.
Booth P., and A. D. Morrison, 2007, Regulatory Competition and Life Insurance Solvency Regulation in the European Union and United States, North American Actuarial Journal, 11: 23-41.
Brookman, J., and P. D. Thistle, 2009, CEO Tenure, The Risk of Termination and Firm Value, Journal of Corporate Finance, 15: 331-344.
Chen, C. R., T. L. Steiner, and A. M. White, 2001, Risk Taking Behavior and Managerial Ownership in the United States Life Insurance Industry, Applied Financial Economics, 11: 165-171.
Cummins, J. D., and D. W. Sommer, 1996, Capital and risk in property-liability insurance markets, Journal of Banking & Finance, 20: 1069-1092.
Dalton, D. R., C. M. Daily, J. L. Johnson, and A. E. Ellstrand, 1999, Number of directors and financial performance: a meta-analysis, Academy of Management Journal, 42: 674–686.
Demsetz, H., K. Lehn, 1985, The Structure of Corporate Ownership: Causes and Consequences, Journal of Political Economy, 93: 1155-1177
Djankov, S., R. La Porta, F. Lopez-de-Silanes, and A. Shleifer, 2008, The law and economics of self-dealing, Journal of Financial Economics, 88: 430–465.
Doidge, C., G. A. Karolyi, and R. Stulz, 2008, Why Do Foreign Firms Leave U.S. Equity Markets? An Analysis of Deregistrations under SEC Exchange Act Rule 12h-6, Working Paper.
Fama, E., and M. Jensen, 1983, Separation of ownership and control, Journal of Law and Economic, 26: 301–325.
Gonzalez, F., 2005, Bank regulation and risk-taking incentives: An international comparison of bank risk. Journal of Banking & Finance, 18: 1153-1184.
Gwartney, J. and R. Lawson, 2007, Economic Freedom of the World: 2007 Annual Report, the Fraser Institute.
Hermalin, B. E., M. S. Weisbach, 1998, Endogenously chosen boards of directors and their monitoring of the CEO, American Economic Review, 88: 96–118.
Hermalin, B. E., and M. S. Weisbach, 2003, Boards of directors as an endogenously determined institution: A survey of the economic literature, FRBNY Economic Policy Review, 9: 7–26.
Hirshleifer, D., and A. Thakor, 1992, Managerial Conservatism, Project Choice, and Debt, The Review of Financial Studies 5, 437–470.
Jensen, M. C., and W. H. Meckling, 1976, Theory of the Firm: Managerial Behavior, Agency Costs, and Capital Structure. Journal of Financial Econometrics, 3: 305-360.
Jensen, M. C., 1993, The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems, Journal of Finance, 48: 831-880.
Johansson, D., 2001, The dynamics of firm and industry growth: The Swedish computing and communications industry. Stockholm, Sweden: Royal Institute of Technology.
John, K., L. Litov, and B. Yeung, 2008, Corporate Governance and Risk-Taking. The Journal of Finance, 63: 1679-1728. Justesen, M. K., 2008, The effect of economic freedom on growth revisited: New evidence on causality from a panel of countries 1970–1999, European Journal of Political Economy, 24: 642–660.
Konishi, M., and Y. Yasuda, 2004, Factors affecting bank risk-taking: Evidence from Japan. Journal of Banking and Finance, 28: 215–234.
La Porta, R., F. Lopez-de-Silanes, and A. Shleifer, 1998, Law and Finance, Jounrnal of Political Economy, 106: 1113-1155. Lee, S. J., D. Mayers, and C.W. Smith, 1997, Guaranty funds and risk-taking evidence from the insurance industry, Journal of Financial Economics, 44: 3-24
McMullen, J.S., D. R. Bagby, and L. E. Palich, 2008, Economic Freedom and the Motivation to Engage in Entrepreneurial Action, Entrepreneurship Theory and Practice, 32: 875-895.
Murphy, K.M., A. Shleifer, and R. W. Vishny, 1991, The allocation of talent: Implications for growth, Quarterly Journal of Economics, 106: 503–530.
Pagano, M., P. Volpin, 2005, The political economy of corporate governance, American Economic Review, 95: 1005–1030. Park, S., 1997, Risk-taking Behavior of Banks under Regulation, Journal of banking and finance, 21: 491-507.
Pathan, S., 2009, Strong Boards, CEO Power and Bank Risk-taking. Journal of Banking & Finance 33: 1340-1350.
Raheja, C. G., 2005, Determinants of Board Firm size and Composition: A Theory of Corporate Boards, Journal of Financial & Quantitative Analysis, 40:283-306.
Rechner, P. K. and D. R. Dalton, 1991, CEO duality and organizational performance: a longitudinal analysis, Strategic Management Journal, 12: 155–160.
Staking, K., D. Babbel, 1995, The Relation between Capital Structure, Interest Rate Sensitivity, and Market Value in the Property-Liability Insurance Industry, Journal of Risk and Insurance, 62: 690-718.
Ward, D., 2003. Can Independent Distribution Function as a Mode of Corporate Governance? An Examination of the UK Life Insurance Market, Journal of Management & Governance, 7: 361-384.
9
Weisbach, M. S., 1988, Outside directors and CEO turnover, Journal of Financial Economics, 20: 431–460.
Wen, M. M., and C. R. Chen, 2008, The Relationship between Executives' Option-Based Compensation and Risk-Taking in the Property/Liability Insurance Industry: A Simultaneous Equation Approach, Journal of Insurance Regulation, 26: 85-107.
10
FIGURE 1
The Average of Worldwide Economic Freedom Index
Notes: We refer the website (http:// www.heritage.org/index) and the the book (Holmes, K. R., E. J. Feulner, and M. A. O'Grady, 2008, 2008 Index of Economic Freedom, The Wall Street Journal) to average the worldwide of individual economic freedoms respectively. The figure covers the period from 1995 to 2008, it further appears the tendency that the index of Europe started to exceed the index of Americas after year 2001. A majority of the freest economies are in Europe.
TABLE 1
Summary of All Sample’s Economic Freedom Index
Country
Numbers of stockholder-owned life insurance firms in
this country Economic Freedom Business Freedom Trade Freedom Fiscal Freedom Government Firm size Monetary Freedom Investment Freedom Financial Freedom Property Rights Freedom from Corruption Labor Freedom ECFR BUFR TRFR FCFR GOSI MOFR INFR FIFR PRRI FRCO LAFR
Austria 27 70.0 80.6 86.0 51.2 25.3 81.4 70 70 90 86 59.2 Belgium 25 71.5 93.7 86.0 43.9 17.9 80.4 90 80 80 73 69.9 Denmark 18 79.2 99.9 86.0 35.0 19.8 86.5 90 90 90 95 99.9 Finland 12 74.8 95.2 86.0 64.3 29.1 88.5 70 80 90 96 48.8 France 68 65.4 87.1 81.0 53.2 13.2 81.2 60 70 70 74 63.8 Germany 125 71.2 88.9 86.0 58.4 34.0 81.4 80 60 90 80 52.8 Greece 3 60.1 69.5 81.0 65.6 57.8 78.5 50 50 50 44 54.3 Italy 33 62.5 76.8 81.0 54.3 29.4 80.6 70 60 50 49 73.5 Netherlands 24 76.8 88.0 86.0 51.6 38.2 86.9 90 90 90 87 60.5 Portugal 8 64.3 79.6 86.0 61.3 32.6 79.4 70 50 70 66 48.0 Spain 21 69.7 77.5 86.0 54.5 56.2 78.1 70 80 70 68 56.7 Sweden 9 70.4 94.8 86.0 32.7 3.9 82.8 80 80 90 92 62.0 UK 98 79.5 90.8 86.0 61.2 40.1 80.7 90 90 90 86 80.7 Total 471
Notes: We mainly use the Eurothesys and ISIS database to get my sample data. However, the Eurothesys include financial statements of insurers in 76 countries all over the world. Due to the completeness of data, we only cover the period from 1996 to 2005. In addition, Due to data availability, we only include the EU countries that appear in La Porta et al. (1998, p.1124~1125, 1156). However, some EU countries are excluded (etc. Lithuania and Romania, Cyprus, Bulgaria, Malta, Slovak Republic and Slovenia) because their life insurers’ information cannot be retrieve from my data sets, for example: Bulgaria does not enforce the Directive 2002/83/EC. In addition, we use the anti-self-dealing index to be my one of control variables from the research of La Porta et al. (1998, p.1124~1125, 1156), Ireland and Luxembourg are excluded for lack the index. In sum, my study employs an unbalanced sample of yearly-based panel database of 552 life insurance companies in 13 EU countries. Furthermore, we refer the website (http:// www.heritage.org/index) to appear the rank of countries of my data. In addition, we refer the research of La Porta et al. (1998, p.1124~1125, 1156) to choice the 10 individual freedoms which are relate to the investment activities around the cross country. Then, we average the 10 individual freedoms to be the index of Economic Freedom (ECFR).
TABLE 2
Variables and Definition
Variables Expected
sign Definition
Risk-Taking Behavior
Risk 1 We compute company earnings volatility.
Risk 2 It is a country-level risk-taking proxy. Defined as the average of the company risk-taking proxy RISK1. Firm-Level Corporate Governance
Board Size (+) We calculate only the number of directors on the board. The secretary or accountants on the board are eliminated. CEOs Duality (+) A dummy variable which is 1 if CEO duality takes place and 0 otherwise. In other words, CEOS DUALITY equals
one if the CEO is also the chair of the board or if CEO is internally-hired. Shareholder Independence Degree (+)
It represents the degree of shareholder independence. We apply a SHAREHOLDER INDEPENDENCE DEGREE of ISIS database. A represent the independent companies. D is the directly majority owned. One shareholder recorded with more than 50% direct ownership. We transfer A, B, C, and D into 1, 2, 3, and 4.
Country-level Corporate Governance
Economic Freedom Annual Index of Economic Freedom that ranges from 100% to 0% with a higher percentage indicating less government control on financial institutions. This index is available for each country since 1995.
Control Variables
Anti-self-dealing Index (+/-)
It includes: (1) approval by disinterested shareholders; (2) disclosures by Buyer; (3) disclosures by Mr. James; (4) independent review; (5) each of the elements in the index of disclosure in periodic filings; (6) standing to sue; (7) rescission; (8) ease of holding Mr. James liable; (9) ease of holding the approving body liable; and (10) access to evidence.
Firm Size (+/-) Defined as the natural logarithm of the insurer’s gross premium written. Leverage (-) Defined as total liability divided by surplus.
TABLE 3
Descriptive Statistics of All Variables and Correlation Matrix
Notes: Panel A separately reports the descriptive statistics for all independent variables by showing minimum (Min.), maximum (Max.), median (Median), mean (Mean), first quartile (Q1), third quartile (Q3), and standard deviation (Std. Dev.). Panel B reports the pair wise of the Pearson correlation matrix for all variables. ***, **, and * represent statistical significance at the 0.01, 0.05, and 0.1 levels, respectively.
Min. Max. Median Mean Q1 Q3 Std. Dev.
Panel A: Descriptive statistic
RISK 1 -50.00 62.04 0.01 110.38 0.00 10.02 23.85
RISK 2 -50.01 --1.72 0.01 110.18 0.01 10.03 20.48
Board size -50.00 74.00 6.00 116.91 4.00 19.00 25.12
CEOs duality - 50.00 -1.00 1.00 110.54 0.00 11.00 20.50 Shareholder independence degree -51.00 --4.00 4.00 113.60 4.00 14.00 20.88 Anti-self-dealing index -50.33 --0.66 0.33 --0.38 0.33 10.39 20.09
Firm size -53.11 17.64 6.84 19.01 5.48 12.88 23.92
Leverage -50.22 158.25 63.50 85.58 9.89 1138.41 17.70
Reinsurance -00.25 79.00 43.04 26.59 7.83 27.16 22.91
Panel B: Correlation matrix
(a) (b) (c) (d) (e) (f) (g) (h) (i)
(a) RISK1 ---
(b) RISK 2 0.001***
(c) Board size -0.051*** ---
(d) CEOs duality -0.082*** -0.112* ---
(e) Shareholder independence degree -0.068*** -0.084* -0.144*** ---
(f) Anti-self-dealing index -0.251*** -0.127** -0.133*** -0.005 ---
(g) Firm size -0.091** -0.018** -0.089** -0.085* -0.383*** ---
(h) Leverage -0.015 -0.031 -0.066* -0.057** -0.011** -0.154*** -0.037* ---
TABLE 4
Firm-level Risk-taking Behavior and Corporate Governance
Dependent Variable: Firm-level Risk-Taking, RISK 1
(1) (2) (3) (4)
Board size -(0.0016** (-0.0011*
( (0.0498) ((0.0653)
CEOs duality ( -0.0021** (-0.0016**
( (0.0431) ((0.0263)
Shareholder independence degree (-0.0024** (-0.0013**
((0.0378) ((0.0312) Constant (-2.1276** (-2.5178** (-2.4559** (-2.7763** ((0.0299) ((0.0468) ((0.0472) ((0.0291) Anti-self-dealing index (5.9234** ((6.1235** ((7.1243** ((5.8792* (0.0297) ((0.0364) ((0.0386) ((0.0676) Firm size (0.0195** ( 0.0366** ((0.0426** ((0.0148** (0.0347) ((0.0218) ((0.0345) ((0.0217) Leverage (-0.0033* (-0.0017* (-0.0028* (-0.0017* ((0.6245) ((0.0864) ((0.0617) ((0.0624) Reinsurance (-0.0001* (-0.0001* (-0.0000* (-0.0014* ((0.9567) ((0.7863) ((0.8105) ((0.0694) Adjusted R 2 ((0.2348 ((0.2454 ((0.2458 ((0.2774 F-statistic (3.7352*** ((3.7843*** ((3.7755*** ((3.8411*** (p-value) ( (0.0017) ((0.0019) ((0.0019) ((0.0042)
Note: This table presents the results of the ordinary least squares (OLS) estimates of firm-level risk-taking behavior. The reported results are corrected for heteroskedasticity. ***, **, and * represent statistical significance at the 0.01, 0.05, and 0.1 level, respectively.
