08-06-2017 Groningen
The value of Enterprise Risk Management: A new measure
for the empirical analysis of the insurance sector in Europe
and the US
W.T. Smida, supervised by Dr. M.A. Lamers
a Faculty of Economics and Business, University of Groningen, the Netherlands
Abstract: This thesis investigates the value of Enterprise Risk Management (ERM) for insurers. In order to do so, the way of identifying ERM users is researched and an ERM rating is constructed to provide a better understanding of the ERM use of insurers. Using this new proxy, the added value of an ERM framework is determined for the biggest insurers in both the US and Europe in the previous ten years. This thesis concludes that using ERM can create value for insurers, once the insurers implement multiple aspects of ERM. Furthermore, this thesis concludes that the new ERM rating creates a better understanding of the value of the use of ERM than previously used proxies for the use of ERM. Next to this, the impact of regulation on risk management in the US and European insurance sectors is investigated. This thesis concludes that ERM is significantly beneficial for the value of insurers in Europe, but does not affect the value of insurers in the US.
Keywords: Enterprise Risk Management, Firm value, Insurance sector, ERM rating, Chief Risk Officers, Value creation, Insurance regulation
1. Introduction
2
Modigliani and Miller (1958) by stating market imperfections and real costs that are brought forth by those risks. Stulz (1996, 2003) argues that risk management adds value by reducing or eliminating the costly lower-tails outcomes. These lower-tail outcomes consist of direct costs such as bankruptcy costs and missing debt covenants. It also contains indirect costs, for example having to pass on an Net Present Value (NPV) project due to a lack of cash or the loss of reputation. Reducing or eliminating the probability of these costs is valuable. In this thesis, a relatively new concept in risk management literature is examined: Enterprise Risk Management.
The key concept of Enterprise Risk Management (henceforth ERM) is that all different types of risk are managed together and controlled as a whole. This is in contrast with the traditional view on risk management where every department of a company hedged its own risk individually (Bromiley, Mcshane, Nair, & Rustambekov, 2014). An effective risk management framework is vital for financial companies in these economically turbulent years and therefore it affects the value of a company. If ERM is an improvement of the efficiency of risk management, this should be visible in the value of firms as well. This thesis aims to investigate this impact on the value of insurers. Insurers are in the business of pricing risks and they should therefore experience the benefits of a more efficient method of risk management the most. This results in the following Research Hypothesis.
RH: Enterprise Risk Management has a positive impact on the value of companies in the insurance sectors of Europe and the US.
3
based on the core elements of ERM as defined in section 2. The ERM rating contributes to a clear view of the extent to which the insurers are using ERM. This leads to a clearer understanding of the added value of ERM in comparison with using only one aspect of the ERM framework as a proxy.
The second contribution of this thesis can be found in the inclusion of insurers in both the European and the US market. Special care is given to the regulatory regimes in these two insurance markets. The impact of these two regimes on the risk management system employed by the insurers is examined. Previous research focused either on the European or on the US market alone.
Thirdly, this thesis contributes to the literature by constructing a new panel dataset consisting of the 59 biggest insurers in Europe and the US that are listed on the Morgan Stanley Capital International (the MSCI) world index. The data in this thesis’ dataset are collected from the period 2007 till 2016 to investigate the most recent results of this method of managing a firm’s risk. The previous ten years have focused on recovering from the financial crisis in 2007 and the focus on risk management has never been higher. This makes it possible to capture the effects of ERM in a period in which it could contribute most to firm value.
The results of this study are that ERM on average adds value to insurers. Not only does ERM improve the efficiency of risk mitigation, it also encourages firms to incorporate risk management in their decision making. This supports them to allocate their capital to the most risk-adjusted profitable activities.
This is not uniformly true though. Firm and market specific characteristics are of an influence on the value of ERM. The US insurers in this sample do not experience added value from their ERM frameworks. The European insurers in this sample do. One of the reasons for this phenomenon is that the regulatory regime in the US, in contrary to the regulatory regime in Europe, does not show interest in the way in which US insurers manage their risks. Therefore, the value of ERM for US insurers does not become visible for the outside world. In Europe, the performance of the risk management framework of insurers is recognized by regulators and the insurers can therefore reap the benefits of their efficient risk management framework.
4
developed. In the third section the empirical study is described, designed to test the research question and the hypotheses. The results of the empirical study are shown in the fourth section of the thesis. Finally, in the fifth section of this thesis the conclusion of this research is described.
2. Literature Review
In this section of the thesis, the most relevant literature in the field of ERM is described. In section 2.1, a definition of ERM and the differences with the traditional, more separated approach to risk management is given. In section 2.2, the ways in which ERM could contribute to the value of insurers is examined. Finally, in section 2.3, the differences between the institutional rules and requirements of insurers in the US and Europe are defined.
2.1. Enterprise Risk Management
Traditionally, risk management was compartmentalized in the firm. Companies used to consider each type of risk separately, without coordinating the hedging methods with their fellow risk managers. Each risk manager had to focus on his own pure risk, such as interest rate, credit, market or foreign exchange rate risk (Bromiley et al., 2014; Liebenberg & Hoyt, 2003; McShane, Nair, & Rustambekov, 2011). This fragmentation occurred as each department of an organization was concerned with its own processes. The Finance department would focus on exchange and interest rate risk, operations managed quality and safety risks and so on. Because of this, the techniques of handling these risks were developed separately and so was the entirety of risk management (Bromiley et al., 2014).
5
Imagine an insurer that sells both annuities and life insurance. These products are managed by two different departments. Managed individually, these products have to be hedged by, for example, buying reinsurance. When the risk of these two departments is considered together, one finds that annuities and life insurance form a natural hedge. If an insurer sells both to similar customers, the risks of these two products will even each other out (McShane et al., 2011).
With ERM, risk is managed holistically, by a centralized and dedicated unit which helps the company to create a clear overview of the total risk that the organization faces. This makes it easier to keep risk within the firm’s own risk appetite and to steer an organization’s risk taking to fields that are essential for their core business (Altuntas, Berry-Stölzle, & Hoyt, 2011; Liebenberg & Hoyt, 2003). Insurers might choose to take on more risk in fields where they have an information advantage and where they know that their risk-adjusted returns are highest. To keep the firm’s risk within the firm’s risk appetite the insurer needs to reduce their risk exposure in fields where this return is lower. With ERM, risk management becomes an essential part of decision making in the company and is not limited to solely mitigating all risks as they come.
From the literature on ERM three core elements of ERM can be drawn. These are summarized in table 1.
Table 1. The core elements of Enterprise Risk Management
Core Element
Description References
1. Managing the risk portfolio of the entire corporation in an integrated fashion
(e.g. Bromiley et al., 2014; Hoyt & Liebenberg, 2011; McShane et al., 2011)
2. Inclusion and quantification of non-traditional and non-financial risks
(e.g. Bromiley et al., 2014; Gordon et al., 2009)
3. Incorporating risk management in the decision making of the company
(e.g. Altuntas et al., 2011; Bromiley et al., 2014)
2.2. Enterprise Risk Management & Value creation
6
Liebenberg and Hoyt (2003) and Beasley, Clune and Hermanson (2005) state in their studies on the determinants of ERM, the appointment of a Chief Risk Office (a CRO) usually means that a firm is starting to manage their total risk exposure in a holistic fashion, considering all potential risks together. Having a CRO or some other dedicated risk entity corresponds with the first core element of ERM.
The impact of hiring a CRO or the appointment of a similar dedicated risk unit on firm value has been researched by various authors (Beasley et al., 2008; Hoyt & Liebenberg, 2011). In the remainder of this thesis, when the presence of a CRO is mentioned, it envelops any entity with a similar function as well. Grace, Leverty, Phillips and Shimpi (2015) researched what aspects of ERM add value to a company. They used a survey of Tillinghast Towers Perrin, an insurance consultancy, regarding the ERM usage of their customers for 2004 and 2006. This resulted in a sample of nearly 4000 insurers in the life and the property/loss insurance industry. They found that the appointment of a CRO increases ROA and therefore has a positive influence on firm value. According to Nocco and Stulz (2006), CRO’s help improve the collection of risk data and reduce the information asymmetry between firm managers regarding the firm’s risks.
Another potential benefit of managing risk in a holistic fashion is that the companies using ERM will have less reinsurance costs, while still effectively reinsuring their risk portfolio. After all, the hedge needed to reinsure the total corporation’s risk portfolio is smaller than the separate hedges needed for each individual department added together. The saved amount of cash can be invested in NPV projects of the firm, adding value to the company. This was one of the conclusions of Hoyt and Liebenberg (2011). As a proxy for ERM adoption, they used the announcement of hiring a CRO or another enterprise-wide risk unit. In their sample of 275 US insurers in the period of 1995 till 2005, they found that on average, ERM using insurers had to rely less on reinsurance than non-users. Next to this, the average ERM user was found to be valued 4 percent higher than non-users (Hoyt & Liebenberg, 2011). This means that when insurers have to rely less on reinsurance, their valuation goes up.
7
to objectively distribute resources among its departments, increasing the capital efficiency and the return on equity (Hoyt & Liebenberg, 2011).
Once the firm has a better understanding of the firm’s total risk exposure, including all non-financial risks, their risk position can be determined more accurately. This way, ERM supports firms to reduce their risk exposure in fields where the risk taking does not pay off enough. Furthermore, it allows firms to increase their risk exposure in fields where they discover that they can increase their leverage and still be within their risk appetite. This means that ERM contributes to an improved asset/liability management and increases the ability of the firm to steer their risk taking to more profitable ventures. This cash injection gives the company more funds to invest in NPV projects, increasing the firm value (Pagach & Warr, 2010). In their study, Pagach and Warr (2010) used the appointment of a CRO or another enterprise-wide risk manager as their proxy for ERM adoption. In total 106 firms that announced the appointment of a risk officer were examined from 1992 until 2004. On average, the leverage and the return of equity increased after the announcement of hiring a CRO, which means that companies were able to increase their debt position after they had acquired a better understanding of their risk portfolio. This result is supported by the study of Hoyt and Liebenberg (2011).
Because an ERM framework clarifies the total risk exposure, financial and non-financial, and incorporates these risks in the firm’s decision making, bankruptcy costs and probability of financial distress are expected to be lower (Beasley et al., 2008; Grace et al., 2015). Therefore, firms that use ERM are expected to invest more in growth options that pay out more in the future but might be difficult to liquidate against real value when in financial distress. This increases firm value on the long run (Beasley et al., 2008; Grace et al., 2015).
8
value of 82 insurers that are rated by S&P in 2008. They found that firm value increases with increasing effective traditional risk management, but firm value stops increasing once a basic level of ERM implementation has been achieved.
2.3. Institutional differences between the US and Europe
There are differences in the regulatory frameworks of the US and Europe. These regulatory regimes have an influence on the possible incentives of insurers to adopt ERM and these regimes also influence the effectivity of an ERM framework (Klein, 2012; Klein & Wang, 2009). In order to research these effects, the two regulatory frameworks in the US and Europe are compared.
The National Association of Insurance Commissioners (the NAIC) handles the regulation on the insurance sector in the US. In Europe, the European Committee (EC) has introduced the Solvency II framework in January 2016. Rules and guidelines have been set up for solo insurers and for conglomerates. In this thesis, only the biggest, listed insurers are selected. All of these are conglomerates and therefore, group-wide rules and guidelines are required. For an extensive description of the two frameworks, see Eling and Holzmüller (2008).
In the US, the NAIC introduced the risk-based capital (RBC) system in 1994. Since 2010 the Federal Insurance Office is the leading regulatory authority after the Dodd-Frank act, which updated the regulatory regime. In the US every insurance business line, e.g. life insurance and property and casualty insurance, has a specific RBC formula, based on the various risks that insurers in that particular business line experience. These formulas calculate the minimum required capital level of insurers. The RBC of the insurer is then compared to its Total Adjusted Capital (TAC), which is the total amount of capital that the insurer has available. The ratio of these capital levels determines the regulatory action, ranging from no regulatory intervention (𝑅𝐵𝐶
9
insurer’s risk management, no incentive is created for insurers in the US to change their risk management system to an ERM framework. The NAIC adjusts for insurance groups by aggregating all capital requirements of its subsidiaries and adjusting that total capital requirements with intra-group transactions (Siegel, 2013).
In Europe, the EC set the Solvency II framework in play in January, 2016. The framework is constructed in the form of three pillars, being (1) quantitative capital requirements, (2) qualitative supervisory review and (3) public disclosure. The first pillar is comparable to the RBC system in the US. Two capital requirements need to be calculated. The first magnitude is the Solvency Capital Requirement (SCR), which is the target capital requirement based on a 99.5 % value at risk concept. The second magnitude is the Minimum Capital Requirement (MCR), which constitutes a minimum level below which the capital of an insurer is not supposed to fall. If funds are below the MCR, the regulator will take action. The second pillar sets guideline for the risk management policies of insurers and the third pillar sets reporting and accounting rules for the insurers (CEIOPS, 2010). According to Klein (2012) this regulatory system is of a “principle-based” nature. This means that regulated insurers have greater freedom to manage their risk portfolio as they see fit, as long as they do it responsibly. It is the job of the regulator to make sure that this is the case. Because the regulators take an in depth interest in the way in which insurers aim to comply to the Solvency II framework, an incentive is created for the insurers to keep improving their risk management system. This could move them to implement an ERM system.
In the Solvency II framework, conglomerates are assessed by using the same formulas as solo insurers. The method used for group solvency is the “Accounting Consolidation-based Method”. The SCR and the MCR are calculated with the same formulas as solo insurers, based on the consolidated financial data. If this method is found not appropriate for a specific group, the “Deduction and Aggregation Method” should be applied. This method calculates the group solvency by subtracting the sum of the aggregated capitals of all group members and the aggregated solo SCRs. Finally, an insurer can apply for permission to calculate its group solvency with internal models (CEIOPS, 2010).
10
US, static formulas are used for their capital requirement calculations and the correlations between different types of risk are assumed to be linear. Therefore, the US does not take into account new market information regarding those correlations. If an ERM using insurer decides to hold less capital because it has discovered natural hedges, the NAIC will not see why. Even though the insurer makes a valid ERM decision, the NAIC could punish it for having too little capital. In Europe, the Solvency II framework calibrates their perception of risk dependencies based on extreme value analysis and the framework is therefore able to account for fat tails (CEIOPS, 2010). Within the Solvency II framework these interdependencies are updated regularly, enabling the regulators to see potential newly arisen natural hedges. This would make an ERM framework valuable because it could show the insurer it is able to hold less capital.
The differences in the regulatory frameworks of the US and Europe are summarized in table 2.
Table 2. Summary of institutional differences between the US and Europe
Category US RBC System Europe Solvency II Reference
1. Risk
Dependencies
Static formula with no consideration for changes in correlations
Correlations between different types of risks are calibrated on basis of extreme value analysis
(Siegel, 2013)
2. Regulation approach
“Rules-based” “Principle-based” (Klein, 2012)
3. Adjustment for group solvency
Based on the aggregation of all group members
Based on the consolidated financial statement
(Siegel, 2013)
Notes: In this table the three, for the purpose of this thesis most relevant, differences between
the regulatory frameworks of the US and Europe are summarized.
2.4. Hypotheses development
The main research question of this thesis regards the value of implementing ERM for insurers in the US and Europe and is repeated below.
RH: Enterprise Risk Management has a positive impact on the value of companies in the insurance sectors of Europe and the US.
11
looking for announcements of hiring a CRO (e.g. Beasley et al., 2008; Beasley et al., 2005; Hoyt & Liebenberg, 2011; Liebenberg & Hoyt, 2003). As Lundqvist (2014) stated in her study, companies could hire a CRO just for signaling purposes and other companies could have implemented ERM without hiring an employee with such a title. Therefore it is expected that the relation between the use of ERM and firm value is stronger in a regression with a more extensive ERM rating, that envelops more of the concept of ERM. This results in the following hypothesis.
H1: This thesis’ ERM proxy results in a stronger relation between the use of ERM and firm value than the presence of a CRO or another risk unit with a similar function.
Furthermore, this thesis studies the role of regulation on the acceptance of ERM by the insurance industry. The regulatory regime in Europe gives more attention to the way in which insurers manage their risks than does the regulatory regime in the US. This means that European insurers have more incentives to manage their risks in a well-organized manner than do the insurers in the US. Furthermore, the Solvency II framework appreciates the outcomes of an ERM framework more than does the RBC system in the US, because it incorporates the interdependencies of different sources of risk better. Therefore, it is expected that European insurers benefit more from an ERM framework than insurers in the US. This leads to the next hypothesis:
H2: ERM usage adds more value to insurers situated in Europe than to insurers situated in the US.
3. Research Method
This thesis follows the methodology of McShane et al. (2011). Both their study and this thesis are focused on the insurance sector and on the impact of ERM implementation on firm value. The difference between the studies is the proxy used for ERM and the inclusion of the European insurance market in this thesis. First, a description of the sample used for the study is given. Secondly, the empirical method and its variables are explained.
3.1. Data collection and sample
12
Of these 59 insurers, 30 insurers have their headquarters in the US and 29 insurers have their headquarters in Europe. There were thirteen insurers that missed data of the previous ten years and those were excluded from the initial dataset. The two insurers that are not primarily in the insurance branch were excluded as well. This resulted in a final dataset of 44 insurers, of which 24 are situated in the US and 20 in Europe, and 440 observations over a period of ten years. The minimum accounting size of the insurers at the end of 2016 was 11.560.871.000,- US dollars.
The data required for the assessment of the ERM rating were taken from the annual reports and SEC filings of the insurers. The stock information was collected using Orbis and the financial data of the insurers were generated using the Datastream database. The index level of the previous ten years of the MSCI world index was gathered from its official website.
3.2. Empirical method
To answer the research question and test the hypotheses, a Weighted Least Squares specification with panel corrected standard errors and cross-sectional fixed effects was made. The cross-sectional fixed effects were added to the estimation because US insurers have other characteristics than European insurers. Next to this, this thesis used Weighted Least Squares of residuals in the regression analysis. An Ordinary Least Squares regression resulted in a Durbin-Watson statistic of 0.69, which meant that there was autocorrelation in the data. By weighting the variables per insurer, this thesis controlled for this autocorrelation. Finally, cross-sectional dependencies were found in the residuals of the data. O’Connell (1998) demonstrated that these dependencies, when not controlled for, can result in falsely rejecting null hypotheses that should not have been. To account for these dependencies this thesis employed panel corrected standard errors.
The specification consists of firm value, the newly constructed ERM rating and six control variables. The specification that is used to answer the research question is as follows:
𝐹𝑖𝑟𝑚 𝑉𝑎𝑙𝑢𝑒 = 𝛽0 + 𝛽1 ∗ 𝐸𝑅𝑀 𝑟𝑎𝑡𝑖𝑛𝑔 + 𝛽2 ∗ 𝐶𝑜𝑛𝑡𝑟𝑜𝑙 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠
13 Dependent variable
The dependent variable of the regression is the value of the firm. As Bromiley et al. (2014) conclude in their review of ERM literature, Tobin’s Q is used most frequently in ERM research as a proxy for firm value. Tobin’s Q is defined as follows:
𝑇𝑜𝑏𝑖𝑛’𝑠 𝑄 = 𝑀𝑎𝑟𝑘𝑒𝑡 𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝑒𝑞𝑢𝑖𝑡𝑦 + 𝑏𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝐵𝑜𝑜𝑘 𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝐴𝑠𝑠𝑒𝑡𝑠
This version of Tobin’s Q is appropriate for the insurance industry as the book value of assets is a good proxy for the replacements costs (Cummings, Lewis, & Wei, 2006; Hoyt & Liebenberg, 2011; McShane et al., 2011). If the Tobin’s Q ratio is larger than 1, the stock is overvalued and this means that the company in question earns a higher rate than its replacement cost. In turn, this indicates that the firm is performing above average. A Tobin’s Q ratio lower than 1 means that the company earns less than its replacement cost which indicates a performance below average. Another reason to use Tobin’s Q is that this measure reflects the expectations of investors. This is important because the benefits of ERM implementation are expected to be lagged (Hoyt & Liebenberg, 2011). The Tobin’s Qs in the dataset are winsorized at 1% to ensure that extreme values do not unduly influence the results of the estimation.
The ERM rating
14
The first aspect of the ERM rating is the presence of a CRO. As aforementioned, the appointment of a CRO has been found to be related with the adoption of an ERM framework, as hiring a CRO indicates that a firm aims to manage all its risk in a coordinated fashion through that single risk director (Kleffner, Lee, & Mcgannon, 2003; Liebenberg & Hoyt, 2003). Since then, multiple studies have used this variable as their proxy for the use of ERM (e.g. Beasley et al., 2008; Florio & Leoni, 2016; Hoyt & Liebenberg, 2011). This indicator of ERM is a binary variable and is defined as “Element1”. Lundqvist (2014) determined that solely the hiring of a CRO is not a conclusive measure for the implementation of an ERM framework. Some firms hire a CRO just for signaling purposes and do not manage their risks in a holistic fashion at all. Others did not hire a CRO, but still have an integrated risk management program in place that is directed by for example the CFO. Although it does not capture the whole concept of ERM, the hiring of a CRO still is the best proxy for the first core element of ERM.
The second ERM indicator captures the second core element of ERM as shown in table 1. In order to have an successful ERM framework in place, all risks, including non-financial risks, need to be quantified and included in the total risk portfolio of the company. Otherwise, an incomplete risk picture is hedged by the managers of the firm. This ERM indicator is a binary variable as well and is defined as “Element2”. If an insurer has protocols for assessing non-financial risks like compliance or reputational risks, it gets a 1. If an insurer fails to have those protocols, it gets a 0 on this aspect of the ERM rating.
The third and final aspect of the ERM rating is based on the third core element of ERM. This ERM indicator is defined as “Element3” and it aims to assess whether insurers merely mitigate their risk or incorporate it into their decision making. An insurer could decide to sell advisory services for certain types of risk where they have an information advantage. They could also decide to steer their risk by redistributing capital from one to another risky activity. If there is evidence that an insurer incorporates risk into its business decisions rather than just eliminating it, the value of the ERM indicator gets a 1, otherwise it gets a 0. Examples of evidence of these indicators are given in appendix A.
15
must be clearly communicated to all employees. Otherwise, the information asymmetry between the CRO and other managers can distort the decision making of the CRO which results in inefficient actions regarding the risk of the company. In this sample, consisting of companies in the business of managing risks, this aspect is redundant. Nearly all insurers in the sample have a written definition of their risk management strategy and risk appetite in their annual reports and therefore, this aspect is not included in the ERM rating. For other industries, this aspect of the ERM rating could be of significance for evaluating the use and efficiency of the use of an ERM framework.
The final ERM rating is as follows:
𝐸𝑅𝑀 𝑟𝑎𝑡𝑖𝑛𝑔 = 𝐸𝑙𝑒𝑚𝑒𝑛𝑡1 + 𝐸𝑙𝑒𝑚𝑒𝑛𝑡2 + 𝐸𝑙𝑒𝑚𝑒𝑛𝑡3
Every indicator has a binary scale, which means that the ERM rating can vary between values of 0 and 3.
Control Variables
Next to the dependent variable and the ERM rating, some control variables are added to the estimation. This thesis follows the method of McShane et al. (2011). These control variables are size, financial leverage, systematic risk, profitability, income volatility and growth opportunities. This thesis controls for size because larger insurers achieve higher return as they have greater market power and lower insolvency risk. The size of insurers is calculated by taking the natural logarithm of total assets. This thesis controls for financial leverage as higher leverage implies higher default risk. Policyholders of a highly leveraged insurer should have to pay a lower premium than less leveraged insurers. This results in lower income for highly leveraged insurers. Like McShane et al. (2011), the measure of financial leverage in this thesis is the financial leverage index and is calculated as the Return On Average Assets (ROAA) divided by the Return On Average Equity (ROAE) of the previous 2 years with a moving average.
16
Growth opportunities, as these influence the perception of investors on firm value. Profitability and growth opportunities are expected to have a positive influence on firm value. Income volatility is expected to negatively influence the value of a firm. All variables of the estimation are summarized in table 3.
Table 3. Variable definitions and expected signs
Dependent variable Definition
Firm Value Tobin’s Q = Market value of Equity+Book value of LiabilitiesBook value of Assets Independent variables Expected Sign Definition
ERM rating + Element1 + Element2 + Element3
Size + Natural logarithm of Total Assets
Financial Leverage - ROAA/ROAE based on a 5 year moving average of total assets and equity (in percentage)
Beta - β (based on monthly returns with a 2 year
moving average)
Profitability + ROA (in percentage)
Income volatility - Standard deviation of monthly FCF scaled by the moving average of FCF of the previous 2 years. Income growth + The average of the monthly revenues growth of
the previous 2 years (in percentage).
Notes: ERM = Enterprise Risk Management; ROAA = Return on Average Assets; ROAE = Return on
Average Equity; β = beta; ROA = Return on Assets. This table shows the definitions and expected signs of all independent variables in this thesis’ estimation. The dependent variable is Tobin’s Q, a measure for firm value. All point data are lagged 1 year before an observation. All averages are determined from the 2 year before an observation.
4. Results
The descriptive statistics of the sample are shown in table 4.
Table 4. Descriptive statistics organized by ERM rating.
ERM rating
No. of Observations
Firm
Value Size Leverage Beta Profitability
Income volatility Income growth 0 21 1,03 17,169 0,21 1,15 1,60 7,40 -0,94 1 55 1,03 17,633 0,18 1,13 2,02 7,28 1,86 2 148 1,04 18,222 0,16 0,99 1,79 7,98 2,30 3 135 1,08 18,065 0,18 1,20 2,35 8,34 1,85
Notes: The definitions of the variables are stated in table 3. This table shows the mean values of
17
The mean values of the variables (table 4) indicate that firms with a higher ERM rating are on average valued better than insurers with a lower ERM rating. Their beta and income volatility are on average the highest as well. This indicates that insurers with an ERM rating of 3 acknowledge that risk management should not only mitigate risk, but also try to profit from it. They could do this by allocating more capital to profitable, although risky ventures in field where they have an information advantage. Furthermore, the results from table 4 indicate that larger insurers tend to be further on their way to an ERM framework that fulfills all three core elements of ERM.
The correlations of the variables in the specification are summarized in table 5.
Table 5. The Pearson correlations coefficients.
Firm value
ERM
rating Size Leverage Beta Profitability
Income volatility Income growth Firm value 1 ERM rating 0.15 .01 1 Size -0.28 .00 0.18 .00 1 Leverage 0.39 .00 -0.01 .80 -0.63 .00 1 Beta -0.16 .00 0.03 .54 0.39 .00 -0.32 .00 1 Profitability 0.54 .00 0.11 .04 -0.61 .00 0.60 .00 -0.29 .00 1 Income volatility -0.12 .02 0.04 .44 0.16 .00 -0.13 .02 0.12 .02 -0.18 .00 1 Income growth 0.01 .87 0.06 .24 0.06 .26 -0.05 .31 -0.10 .05 -0.03 .58 0.34 .00 1
Notes: The definitions of the variables are stated in table 3. The p-values of the correlations are
italics.
18
applied the Variance Inflation Factors (VIFs) of Belsley, Kuh and Welsch (1980). All VIFs are under 2.5, indicating that multicollinearity most likely will not be a problem for the estimation.
4.1. Main results
Next are the results of the main regression of firm value and the ERM rating, as defined in this thesis. The results are summarized in table 6.
Table 6. The results of the regression of the ERM rating on firm value.
Firm value
(1) ERM rating (2) CRO
Coefficient p-value Coefficient p-value
Intercept 0,191 0,135 0,292 0,009 *** ERM rating 0,006 0,024 ** -0,010 0,062 * Size 0,046 0,000 *** 0,043 0,000 *** Leverage 0,021 0,070 * 0,016 0,169 Beta - 0,005 0,004 *** -0,006 0,002 *** Profitability 0,005 0,004 *** 0,008 0,000 *** Income Volatility < 0,001 0,134 0,000 0,828 Income Growth <-0,001 0,946 0,000 0,683 Adjusted R2 0,742 0,789
Cross-section fixed (dummy variables)
Notes: The definitions of the variables are stated in table 3. In the regression analyses, weighted
least squares of residuals with cross-sectional fixed effects are used. To control for cross-sectional dependencies, panel corrective standard errors are employed. Regression 1 is the main result. The estimation is based on the ERM rating as defined in table 3 of US and European insurers. Regression 2 is based on solely the presence of a Chief Risk Officer as a proxy for the use of ERM. ***, **, * indicate significance at the 1%, 5%, 10% level, respectively.
This thesis concludes that at least one core element of ERM creates value to the insurer (table 6). The coefficient is positive and significant at the 5% level. This result verifies the research hypothesis that ERM has a significant and positive impact on firm value for insurers.
The results of the regression provide the following estimation of the regression of the ERM rating on firm value:
19
Three of the control variables have the expected strong impact on firm value (Table 6). Both the size and the profitability have a positive relation with firm value at the 1% significance level. The systematic risk has a negative relation with firm value, significant at the 1% level. There is some evidence of a positive relation between leverage and firm value (table 6). This indicates that when firms start using ERM, they discover that they can take on more debt while still being within their own risk appetite (Hoyt & Liebenberg, 2011; Pagach & Warr, 2010). This way, their leverage reaches its optimal level and value is created.
The income volatility and revenue growth on firm value has no significant impact on firm value (table 6). According to McShane et al. (2011) this could be because the two year average sales growth is a rather weak proxy. Most risk management literature have focused on non-financial firms and use ratios such as capital expenditures or research and development expenditures on total sales as a proxy for growth opportunities. These ratios are not applicable to the insurance sector. A reason for the absence of a relation between both income volatility and growth and firm value could be the income structure of an insurer. Next to its operational income from premiums from their policyholders, an insurer also earns substantial financial returns from their investments. Therefore, investors evaluate insurers based on their financial returns instead of on their income from their policy premiums.
Hypothesis 1 stated the expectation that the results of a regression with the ERM rating as a proxy for the use of ERM are more conclusive in comparison with using the presence of a CRO. In support of this expectation, the results of the regressions are more significant when using the extensive ERM rating than using only the presence of a CRO in the company as the proxy for the use of ERM (table 6). Hypothesis 1 is therefore accepted after consideration of the data.
20
a CRO is merely a prerequisite for an ERM framework to work. In itself it does not add value, but without a CRO, the other core elements of ERM are hard to implement. This would explain why the ERM rating has a positive and only the presence of a CRO a negative influence on firm value. This is in accordance with Gordon et al. (2009) who state that the value of ERM depends on its implementation. If ERM is not utilized completely and customized to the company, it will not create value. Meanwhile, ERM does have costs. Therefore, when not properly implemented, ERM could reduce firm value.
Next to these results, this thesis investigates the differences in the use of ERM by insurers in the US and Europe. The results of that comparison are summarized in table 7.
Table 7. The differences between the US and the European insurance industry.
Firm Value
(3) US (4) Europe
Coefficient p-value Coefficient p-value
Intercept 0,687 0,000 *** 0,411 0,017 ** ERM rating 0,003 0,242 0,006 0,081 * Size 0,021 0,015 ** 0,031 0,001 *** Leverage 0,024 0,556 0,088 0,165 Systematic Risk - 0,004 0,102 - 0,011 0,002 *** Profitability 0,017 0,000 *** 0,010 0,003 *** Income Volatility < 0,001 0,252 < 0,001 0,020 ** Income Growth < 0,001 0,649 < 0,001 0,501 Adjusted R2 0,869 0,651
Cross-section fixed (dummy variables)
Notes: ERM = Enterprise Risk Management. This table states the results of the regressions of the
ERM rating on firm value. The ERM rating and the other variables are defined in table 3. Regression 3 states the results of the estimation based on the US insurers in the sample. Regression 4 states the results of the estimation based on the European insurers. ***, **, * indicate significance at the 1%, 5%, 10% level, respectively.
21
5. Conclusion
This thesis investigates the impact that the use of ERM has on the value of insurers in the US and Europe in the timeframe 2007 until 2016. In doing so, this thesis acknowledged the lack of a broadly accepted proxy for the use of ERM and constructed an ERM rating that includes the whole concept of ERM. Furthermore, this thesis examined the differences between the insurance regulation regimes in the US and Europe.
A limitation of this study is the relatively small sample size. Only 44 of the insurers on the MSCI index were viable for this thesis. Next to this, the sample was constructed of the biggest insurers in the US and Europe. The sample is therefore subject to a possible selection bias. Further research on other samples is needed on all implications of this study in order to be able to broadly apply the implications of this thesis.
In determining the extent of the use of the ERM framework by the insurers in the sample, this thesis theorized an ERM rating, based on the three core elements of ERM. These core elements are:
(1) managing all kinds of risk in a holistic fashion,
(2) the inclusion of non-financial risks, such as operational and reputational risks, in the risk portfolio and
(3) the inclusion of risk management in the decision making of the insurer, making risk management about taking advantage of risks and not just mitigating them.
22
ERM. The presence of a CRO by itself is not a guarantee for the creation of value. The relation with firm value was significant, but negative. Therefore, it is recommended to insurers to implement a complete ERM framework as soon as possible. Simply hiring a CRO without taking further actions to improve the ERM could harm firm value. Once insurers implement more core elements of ERM, they can create value with their risk management framework. If insurers do not fully commit to ERM and only implement part of the aspects of the concept, they might be counterproductive.
23
Bibliography
Altuntas, M., Berry-Stölzle, T. R., & Hoyt, R. E. (2011). Implementation of Enterprise Risk Management: Evidence from the German Property-Liability Insurance Industry. The Geneva
Papers, 36(2004), 414 – 439.
Beasley, M. S., Clune, R., & Hermanson, D. R. (2005). Enterprise risk management: An empirical analysis of factors associated with the extent of implementation. Journal of Accounting and
Public Policy, 24(6), 521–531.
Beasley, M. S., Pagach, D., & Warr, R. (2008). Information conveyed in hiring announcements of senior executives overseeing enterprise-wide risk management processes. Journal of
Accounting, Auditing & Finance, 23(3), 311–332.
Belsley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression diagnostics: Identifying influential data and
sources of collinearity. New York, NY: John Wiley.
Bromiley, P., Mcshane, M., Nair, A., & Rustambekov, E. (2014). Enterprise Risk Management: Review, Critique, and Research Directions. Long Range Planning, 48(4), 265–276.
Committee of European Insurance and Occupational Pension Supervisors (CEIOPS). (2010). Solvency II – Calibration Paper. Retrieved from https://eiopa.europa.eu
Cummings, J. D., Lewis, C., & Wei, R. (2006). The Market Impact of Operational Risk Events for US Banks and Insurers. Journal of Banking and Finance, 30, 2605–2634.
Eling, M., & Holzmüller, I. (2008). An Overview and Comparison of Risk-Based Capital Standards.
Journal of Insurance Regulation, 26(4), 31–60.
Florio, C., & Leoni, G. (2016). Enterprise risk management and firm performance : The Italian case.
The British Accounting Review, 1–19.
Gordon, L. A., Loeb, M. P., & Tseng, C. Y. (2009). Enterprise risk management and firm performance: A contingency perspective. Journal of Accounting and Public Policy, 28(4), 301–327.
Grace, M. F., Leverty, J. T., Phillips, R. D., & Shimpi, P. (2015). The value of investing in enterprise risk management. Journal of Risk and Insurance, 82(2), 289–316.
Hoyt, R. E., & Liebenberg, A. P. (2011). The Value of Enterprise Risk Management. The Journal of Risk
24
Kleffner, A. E., Lee, R. B., & Mcgannon, B. (2003). The Effect of Corporate Governance on the Use of Enterprise Risk Management: Evidence from Canada. Risk Management and Insurance Review,
6(1), 53–73.
Klein, R. W. (2012). Principles for Insurance Regulation: An Evaluation of Current Practices and Potential Reforms. The Geneva Papers on Risk and Insurance - Issues and Practice, 37(1), 175– 199.
Klein, R. W., & Wang, S. (2009). Catastrophe risk financing in the United States and the European Union: A comparative analysis of alternative regulatory approaches. Journal of Risk and
Insurance, 76(3), 607–637.
Liebenberg, A. P., & Hoyt, R. E. (2003). The Determinants of Enterprise Risk Management : Evidence From the Appointment of Chief Risk Officers. Risk Management and Insurance Review, 6(1), 37– 52.
Lundqvist, S. A. (2014). An exploratory study of Enterprise Risk Management: pillars of ERM. Journal
of Accounting, Auditing & Finance, 29(3), 393–429.
McShane, M. K., Cox, L. A., & Butler, R. J. (2010). Regulatory competition and forbearance: Evidence from the life insurance industry. Journal of Banking and Finance, 34, 522–532.
McShane, M. K., Nair, A., & Rustambekov, E. (2011). Does Enterprise Risk Management Increase Firm Value? Journal of Accounting, Auditing & Finance, 26(4), 641–658.
Modigliani, F., & Miller, M. H. (1958). The Cost of Capital, Corporation Finance and the Theory of Investment. The American Economic Review, 48(3), 261–297.
National Association of Insurance Commissioners (NAIC). (2009). Risk-Based Capital –General Overview. Retrieved from www.naic.org
Nocco, B. W., & Stulz, R. M. (2006). Enterprise Risk Management: Theory and Practice. Journal of
Applied Corporate Finance, 18(4), 8–20.
O’Connell, P. G. J. (1998). The Overvaluation of Purchasing Power Parity. Journal of International
Economics, 44, 1–20.
Pagach, D., & Warr, R. (2010). The Effects of Enterprise Risk Management on Firm Performance. The
Geneva Papers on Risk and Insurance - Issues and Practice, 45(April), 65–66.
25
Comparison of Regulatory Frameworks. The Geneva Papers on Risk and Insurance - Issues and
Practice, 38(2), 308–331.
Smithson, C., Associates, R., & Simkins, B. J. (2005). Does Risk Management Add Value? A Survey of the Evidence. Journal of Applied Corporate Finance, 17(3), 8–17.
Stulz, R. M. (1996). Rethinking risk management. Journal of Applied Corporate Finance, 9(3), 8–24. Stulz, R. M. (2003). Rethinking risk management, The Revolution in Corporate Finance (4th edition).
26
Appendix A
Examples of the indicators of ERM in the annual reports of insurers. Element1 is the presence of a CRO in the company. Element2 is the inclusion of non-financial risks, such as operational and strategic risks, in the risk management framework. Element3 is the consideration of risks in the decision making of the insurer and not just mitigating as much risk as possible. If an insurer tries to profit from its risk management system, it also gets a score on Element3.
For instance, Element1 gets scored a 1 if an annual report or SEC filing reports: • ‘The risk committee reports to our group chief risk officer.’
• ‘The Executive Board and the Group Chief Risk Officer report to the Risk Committee of AEGON’s Supervisory Board, which is responsible for overseeing AEGON’s enterprise risk management framework.’
For instance, Element2 gets scored a 1 if an annual report or SEC filing reports:
• ‘The group risk map, approved by the company’s board of directors within the risk management policy, identifies the following main risks faced by the company: financial risks, credit risks, insurance risks, operational risks and other risks.’ • ‘The Group is currently developing a strategic risk management system which will
increase our understanding of the operational risks in the business and facilitate the improvement in the controls to reduce losses.’
For instance, Element3 gets scored a 1 if an annual report or SEC filing reports:
• ‘All risk factors affecting the ordinary business are taken into consideration in the decision making process: a risk based approach is applied in particular in the processes related to capital management, reinsurance, asset allocation and new product development.’
