University of Ghana http://ugspace.ug.edu.gh UNIVERSITY OF GHANA INVESTMENT RISK TAKING BEHAVIOUR OF INSURANCE COMPANIES IN GHANA By MARGARET ASARE (10508577) A THESIS SUBMITTED TO THE SCHOOL OF GRADUATE STUDIES, UNIVERSITY OF GHANA, LEGON, IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF A MASTER OF PHILOSOPHY DEGREE IN RISK MANAGEMENT AND INSURANCE JULY, 2016 University of Ghana http://ugspace.ug.edu.gh DECLARATION I, Margaret Asare, do hereby declare that, except for references to people’s work, which I have duly acknowledged, the study herein presented is the first of its kind to be submitted to the University of Ghana Business School (UGBS) under the supervision of Dr. Charles Andoh and Dr. Eric Ofosu- Hene. I am solely responsible for any mistakes encountered in the course of writing this thesis. Signed……………………………. Data……………………………… Margaret Asare (Candidate) ii University of Ghana http://ugspace.ug.edu.gh CERTIFICATION I hereby certify that this long essay was supervised in accordance with procedures laid down by the University of Ghana. The work has been submitted for examination with our approval as supervisors. Signed…………….………………. Date…………………………. Dr. Charles Andoh (Principal Supervisor) Signed……………….…………….. Date……………………………. Dr. Eric Ofosu-Hene (Co-supervisor) iii University of Ghana http://ugspace.ug.edu.gh DEDICATION This work is dedicated to the praise and glory of the Lord Almighty, who is the light of my life, and to my senior brother Dr. Antwi-Asare. iv University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT Glory be unto God for His divine protection and provision that has seen me through this program. No person is self-made; and I am no exception. I feel really privileged to come this far leaning and standing on the shoulders of many people and I am very grateful to God for the great people He has brought into my life. I am very grateful to my supervisors, Dr. Charles Andoh and Dr. Eric Ofosu-Hene for their guidance, and immense contribution. I also thank all the lecturers at the Department of Finance whose tutelage and contribution during seminar sessions has deepened my understanding of Risk Management and Insurance. My deepest gratitude to Prof. & Dr. Mrs. Joshua Abor, for their encouragement and support in taking this step of further education. I am very thankful to Rev. & Rev. Mrs. Clement Anchebah, for their spiritual and financial support throughout the program. I deeply appreciate the mentorship and encouragement of Rev. & Pastor Mrs. Cornelius Yakung. I am exceedingly grateful to my brothers Dr. Antwi-Asare and Mr. Prince Kyei Asare for their continual support and encouragement in my entire educational life. My profound gratitude goes to Ms. M. Acquah-Swanzy, Mr. M. Sandor, Mr. Dery, Mr. Nyaaba, Mrs. Grace Onumah, Mr. & Mrs. Filson and Mrs. Theresa Aniaggei for their tremendous financial support. A special thanks to Mr. Francis Kuditcher, Mr. Courage Hodey, Mrs. Augusta Deih, and Mr. Ibraheem Jabir for proof-reading and shaping this thesis. Finally, I am very grateful to all my colleagues in Risk and Finance class for their contribution throughout the course work, thank you. v University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ................................................................................................................................. ii CERTIFICATION ..............................................................................................................................iii DEDICATION .................................................................................................................................... iv ACKNOWLEDGEMENT ................................................................................................................... v ABSTRACT ........................................................................................................................................ ix LIST OF TABLES ............................................................................................................................... x LIST OF FIGURES ............................................................................................................................ xi LIST OF ABBREVIATIONS ............................................................................................................ xii CHAPTER ONE .................................................................................................................................. 1 INTRODUCTION ............................................................................................................................... 1 1.1 BACKGROUND ................................................................................................................... 1 1.2 PROBLEM STATEMENT ................................................................................................... 5 1.3 THE PURPOSE OF STUDY ................................................................................................ 8 1.4 OBJECTIVES OF THE STUDY ............................................................................................... 8 1.5 RESEARCH HYPOTHESES ............................................................................................... 9 1.6 SIGNIFICANCE OF STUDY ............................................................................................... 9 1.7 LIMITATIONS ................................................................................................................... 10 1.8 ORGANIZATION OF STUDY .......................................................................................... 10 CHAPTER TWO ............................................................................................................................... 11 LITERATURE REVIEW .................................................................................................................. 11 2.1 INTRODUCTION ................................................................................................................... 11 2.2 THEORETICAL REVIEW ..................................................................................................... 11 2.2.1 Background Risk Theory .................................................................................................. 11 2.2.2 Portfolio theory ................................................................................................................. 13 2.3 EMPIRICAL REVIEW ........................................................................................................... 14 2.4 DETERMINANTS OF INSURANCE INVESTMENT .......................................................... 21 2.5 CATEGORIES OF INVESTMENT OF INSURANCE COMPANIES .................................. 23 2.5.1 Investment Properties........................................................................................................ 23 2.5.2 Equity Investment ............................................................................................................. 24 2.5.3 Fixed Deposit .................................................................................................................... 24 vi University of Ghana http://ugspace.ug.edu.gh 2.5.4 Treasury Bills .................................................................................................................... 25 2.5.5 Cash Accounts .................................................................................................................. 25 2.6 OVERVIEW OF THE INSURANCE INDUSTRY ................................................................ 26 METHODOLOGY ............................................................................................................................ 29 3.1 INTRODUCTION ................................................................................................................... 29 3.2 THE DATA SOURCE ............................................................................................................. 29 3.3 MODEL SPECIFICATION ..................................................................................................... 30 3.4 EXPLANATION OF VARIABLES ........................................................................................ 31 3.4.1 Capital Ratio (CR)............................................................................................................. 31 3.4.2 Leverage Ratio (LevR) ...................................................................................................... 32 3.4.3 Firm Size (Size) ................................................................................................................. 32 3.4.4 Expense Ratio (ER) ........................................................................................................... 33 3.4.5 Loss Ratio (LR) ................................................................................................................. 34 3.4.6 Investment Income Ratio (INCR) ..................................................................................... 34 3.4.7 Reinsurance Ratio (RIR) ................................................................................................... 34 3.4.8 Group Affiliation (GA) ..................................................................................................... 35 3.4.9 Inflation (Infla).................................................................................................................. 35 3.4.10 Interest rate (IntRate) ...................................................................................................... 36 3.5 MEASURE FOR INVESTMENT RISK TAKING EXPOSURE ............................................... 36 3.6 LIMITATION OF THE METHODOLOGY ............................................................................... 39 CHAPTER FOUR .............................................................................................................................. 41 PRESENTATION OF DATA AND DISCUSSION OF RESULTS ................................................. 41 4.1 INTRODUCTION ................................................................................................................... 41 4.2 DESCRIPTIVE ANALYSIS ................................................................................................... 41 4.3 CORRELATION ANALYSIS................................................................................................. 44 4.4 REGRESSION ANALYSIS .................................................................................................... 47 CHAPTER FIVE ............................................................................................................................... 57 SUMMARY OF FINDINGS, CONCLUSION AND RECOMMENDATIONS .............................. 57 5.1 INTRODUCTION ................................................................................................................... 57 5.2 SUMMARY OF FINDINGS ................................................................................................. 57 5.3 CONCLUSIONS.......................................................................................................................... 59 5.4 RECOMMENDATIONS ............................................................................................................. 60 vii University of Ghana http://ugspace.ug.edu.gh 5.5 LIMITATION AND FUTURE STUDY ..................................................................................... 61 REFERENCES .................................................................................................................................. 62 APPENDICIES .................................................................................................................................. 70 APPENDIX 1: TREND OF VARIABLES .................................................................................... 70 APPENDIX 2: HAUSMAN TEST ................................................................................................ 72 viii University of Ghana http://ugspace.ug.edu.gh ABSTRACT This study analyses the impact of firm-specific and macroeconomic factors on investment risk taking behaviour of insurers and examine how insurers’ risk taking influences the financial performance of insurance companies in Ghana. The study employs a panel regression model for the general analysis using a quantitative data sample of 16 non-life and 12 life insurance companies’, randomly selected from the National Insurance Commission database over 7 years period (2008-2014). The analysis showed that factors such as firm size, capital, claims payment, management expense, investment income, interest rate and inflation significantly affect investment risk taking behaviour of the insurer. Firm size negatively influenced life insurance high investment risk taking and non-life insurance low investment risk taking. But the impact of firm size was found to be positive under low investment risk taking measure for life insurance. The findings revealed that life insurers with higher expense ratio have less incentive to take on more risk in asset investments. Also life insurance long term investment activities had positive impact on the financial performance of the insurance industry. This study provide an insight into the investment risk taking behaviour of the insurance industry and suggest that policy makers in the insurance market need to encourage insurance managers and shareholders to give much attention to their operational activities by improving their underwriting risk activities and investment risk taking awareness. ix University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 2.1 Investment mix and categories ....................................................................................... 27 Table 3.1 Summary of variables ..................................................................................................... 37 Table 4.1 Descriptive statistics for Non-life and Life Insurance ................................................. 43 Table 4.2 Correlation matrix for Non-life (High risk taking) ...................................................... 45 Table 4.3 Correlation matrix for life (High risk) .......................................................................... 46 Table 4.4 Regression results for high and low investment risk taking behaviour ..................... 48 Table 4.5 Performance results of Insurance Business .................................................................. 54 Table A2.1 Hausman test results for Non-life Insurance - High risk .......................................... 72 Table A2.2 Hausman Test results for Life Insurance – High risk .............................................. 73 Table A3.1 CV for Life companies (High risk) ............................................................................. 75 Table A3.2 CV for Non-life Companies (High risk) ..................................................................... 76 Table A3.3 CV for Life Companies (Low risk) ............................................................................. 76 Table A3.4 CV for Non-life Companies (Low risk) ...................................................................... 77 x University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 1 Premium growth rate for life and non-life ........................................................................... 70 Figure 2: Trend of firm specific variables ......................................................................................... 71 Figure 3: Trend of firm specific variables ......................................................................................... 71 Figure 4 Trend of total assets for life and non-life ............................................................................ 72 xi University of Ghana http://ugspace.ug.edu.gh LIST OF ABBREVIATIONS NIC National Insurance Commission RBS Risk Base System GDP Gross Domestic Product CV Coefficient of Variation US United States xii University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE INTRODUCTION 1.1 BACKGROUND Globally, the insurance industry as a financial and risk management institution helps in the protection and management of variety of perils. Insurers receiving premiums from policyholders for an insured risks, pledge base on the policy contract, to settle any claims and benefits payment. However, prior to any payment of claims and benefits, the premiums received are invested to generate returns to cater for any financial cost. Apart from the financially distressed cost faced by insurers which triggers risk taking in investments activities, insurer’s ability to decide on which portfolio, either in long or short term, to invest their assets are also a major form of risk taking due to market instability. Because of the global financial crisis from 2007-2009, investment risk taking has received keen attention all over the world. The need for insurers to meet the required solvency margins has made investment of asset very important in the industry. Without any doubt investment risk taking is a major activity of insurance companies. Meanwhile two salient questions that remain unanswered are whether (I) the background operation of insurance companies does influence their investment risk taking? (II) Insurers are conscious about the interrelationship between their investment risk taking behaviour and background activities and how these influences their investment plans and expected outcomes? In Ghana, the insurance industry is striving with huge potential for growth in both life and non-life market. Gross premiums growth for the industry stood at 13% and 23% for both nonlife and life businesses in 2014 (National Insurance Commission, 2014), respectively. The trend of premium growth in the industry is falling gradually (Fig 1, Appendix 1). Basically, the regulatory instruction for the separation of life and non-life 1 University of Ghana http://ugspace.ug.edu.gh businesses in 2007 by the National Insurance Commission in addition to the increment of minimum capital requirement from GH¢5 million in 2012 to GH¢15 million in 2015 has made the insurance companies more capitalised. Once more, the low pricing of policies which has been a major challenge in the non-life insurance sector was recently addressed for motor third party policies, with an increase in premium charges by 300%. All things being equal, there are great expectations that these changes in the short and long run will sustain the equity base of companies in the insurance business. Most importantly, as policyholders seek cover for greater risk and product innovation continues with keen competition in the insurance market, with regulatory backing from NIC, insurance companies are required to be financially equipped to meet their core duty of claims payment. Ensuring continual satisfactory support for societal needs, especially for the vulnerable and enhancing persistence growth in the insurance industry, insurance investment must be well managed. Moreover, as individuals and businesses response to paying commensurate premiums to cover for the occurrence of an insured risk, it would be very unfair and illegal if insurers are not able fulfil their part of the contract should the loss of the insured risk occurs. One constraint of the insurance business is the inability to accurately determine the cost of occurrence of an insured risk since most often the pricing of insurance premiums are based on actuarial probabilities and past experience. Although professional actuaries by experience price policies taking into consideration underwriting issues regarding the insured risk and macroeconomic conditions, circumstances surrounding these factors are sometimes beyond control of professionals. For example, the cost involved in the frequent fire occurrence and the national disaster caused by a heavy rainfall in Ghana in June 2015, destroying lives and huge properties, was far beyond professional calculation, triggering huge financial burden on insurers. These uncertain costs necessitates that insurance 2 University of Ghana http://ugspace.ug.edu.gh companies’ funds are well managed and activities well scrutinized to financially cushion insurers to meet their obligation of any contractual demands. Hence, the need for investment activities which calls for strategic investment risk taking by insurers. Investment practices of an insurance firm increase profitability and add on to capital performance. Insurers do not only accrue profit from the performance of business activities, for example premium growth and product innovation but they also gain additional income from returns on investment. Making investment income really essential on insurer’s balance sheet accounts as it reduce the cost of insurance in paying policy owners, because premiums are invested to yield interest (Janowicz-lomott, 2011). The risk preference on returns cause insurers to be more selective in their investment portfolio management (Franzen, 2010). That notwithstanding, the nature of insurance business determines the choice of investment an insurers are likely to engage. For example, the long term nature of life insurance policies gives insurers the opportunity to trade in long term investments, attracting higher returns in the midst of volatility (Zou, Wen, Yang, & Wang, 2012). Considering the key activities of the insurance industry being investment and underwriting, there are several factors that influence the investment decisions of insurers which instigates a trade-off between the consequences of insurance operational activities and their investment practices. According to Zou et al. (2012), the trade-off of risk between expected return tends to increase assets and liability difference which influence future insurance benefits. Financial institutions in addition to engaging in their main business activities, often hold large investment portfolios in various segment of the financial market (Chen, Yao & Yu, 2007). For example, though the main business of insurance companies is to underwrite life or non-life policies, collect premiums and manage claim pay-outs, they also hold large investment portfolios. Investment of funds generated from a prudent underwriting activity does not only add on to 3 University of Ghana http://ugspace.ug.edu.gh company’s profitability but also to capital performance which is relevant in the financial stability of the company. The type of the insurance business will inform the sort of investment an insurer is likely to venture. The long term nature of life insurance, allows life insurers to venture into long term investments. Besides due to short term liabilities experience by non-life insurance, insurers are likely to go into short term investment. For an insurance company to achieve strategic investment goals, that is maximizes investment profit, the main task of an insurance investor will be to manage an investment portfolio that will generate maximum income with tolerable risk level. However, how insurers adjust their investment risk taking behaviour against the risk emanated from their main business operation is a paramount issue to both the insurers, regulators and policyholders (Zou et al., 2012). In recent times, risk taking behaviour has been widely studied in literature in various fields due to the fact that systems and the entire environment keeps changing, thus creating uncertain conditions beyond human control (Cheng, Elyasiani, & Jia, 2011; Fields, Gupta, & Prakash, 2012; Laeven & Levine, 2009; Lee, Mayers, & Smith, 1997; Ng, Chong, & Ismail, 2013). According to this study, investment risk taking is regarded as the capacity of an insurance firm to take on risk of investing its assets in a high risk portfolio. Now the question is, do activities of insurance operations influence the risk taking behaviour of insurance investors? Several factors arising from both internal and external activities of insurance operations affect the investment risk taking behaviour of insurers. The internal factors are insurance firm specific characteristics, whiles the external factors entails both industry features and macroeconomic variables. Bearing in mind that underwriting and investment are two essential activities of the insurance business, there is the need for insurers to have in place the right systems to minimize the risk emerging from these operations. 4 University of Ghana http://ugspace.ug.edu.gh The insurance business contributes greatly to economic development; however economic factors like GDP, interest rate, inflation, may affect the decision of an insurer’s investment risk taking behaviour. 1.2 PROBLEM STATEMENT As explained earlier, the role of insurance companies as financial institutions is not only to receive premiums and pay out claims to policyholders but also to efficiently manage their investment assets. Investment practice of insurance companies over the years has become a major concern for insurers since their activities are controlled by state regulators (Auken & Carter, 1990). Evidence of the challenges of the 2007-2009 worldwide financial crisis points out how excessive investment risk taking and poor risk management will lead to extensive insolvency risk (Schich, 2009). Insurance companies, as one of the risk management industries are exposed to significant risk from both operational activities and investment practices. One difficult task faced by insurers and regulator is how to determine the desired level of investment risk taking that optimize firms profitability and solvency (Ren, Sun, Sun, & Yu, 2011). Also balancing of insurer’s asset allocation with liabilities obligations (claims and benefits) requires that insurers have sufficient funds to meet any payment of losses. Meanwhile, expectation of higher returns has its own challenges; that is higher risk. Conditions on the capital market will reward higher returns only if higher risk are taking. Meanwhile there are several external and internal factors that influence how insurers take on risk in investing its assets. According to Gorter and Bikker (2013), the insurance industry is challenged with financially uncertain cost of claims 5 University of Ghana http://ugspace.ug.edu.gh settlement and regulatory difference, and these issues make insurers and insurance investors more sensitive in their investment risk taking. Though insurers need to generate high profit in the form of returns amidst any uncertainty, recognizing the risk association that may exist from their investment activities and the loss exposure of their operations will inform and enhance their investment strategies. The benefits from higher returns in relation to a higher investment risk taking determines how objective insurers are in examining the factors affecting their investment risk taking behaviour. Hence, the need to know the relationship that exist between these factors and to update insurers how to adjust on their risk preferences to attract good investors. Investigating these global issues of investment risk taking activities in the midst of industrial and economic challenges in a developing country like Ghana is very timing as the industry is working hard to improve on insurance penetration. Currently in Ghana, the insurance industry is experiencing fluctuations in gross premium growth, loss and expense ratios for both life and nonlife respectively (NIC, 2014). The trend to this effect is shown in Figure 1, 2 and 3 (in Appendix 1). These changes are gradually affecting investment activity in the industry. That notwithstanding, the increment in capital requirement from GH¢5 million in 2012 to GH¢15 million in 2015 for the purpose of sustaining the capital and solvency base of the insurance industry has really increased the financial burden of insurers. Meanwhile as insurers continuously receive premiums and capital is increasing, the main concern of insurers and regulators is how these assets are invested to meet the required solvency margin, securing maximum protection for the policyholder and the entire public. The insurer as an investor always want to invest more to accrue higher returns depending on the type insurance of business and the frequency of claim payment. Consequently, the desire for a higher return base on the market conditions will demand a higher risk taking. And as a financial 6 University of Ghana http://ugspace.ug.edu.gh and risk management institution, insurance companies are faced with certain firm specific and macroeconomic factors likely to influence their investment risk taking behaviour. Although, rules governing investment activities in the Ghanaian industry has gone through different phases, from the introduction of investment mix rule in 2008 to the current risk base supervisory system; background operations of the industry continuously impose a financial burden on insurers and these affects their capacity in investing their assets. Again, even though the state regulator allows insurers to manage their own investment activities, there are limits on the percentage of asset allocation in the various investment categories. With the current risk base supervision system, these limits are applied in the form of weights on the insurers assets allocated (NIC, 2014). All these regulatory issues indirectly influences the investment decision of the insurer. Insurers are more particular about how internal and external activities in their business operation affect their investment risk taking behaviour. As a financially distressed institution, investment risk taking is paramount in assessing the performance of the insurance industry. However, despite all the issues raised, insurers may have neglected the correlation between the investment risk taking and firm’s background operational factors ( Achleitner, Biebel, & Wichels, 2002) and probably managing them in silos. With these relevant issues, in the industry, there has not been a study on the investment risk taking behaviour within the Ghanaian insurance sector. Akintola-Bello (1986) investigated the investment behaviour of insurance firms in Nigeria by examining the factors governing the investment behaviour of insurance firms but did not consider the interrelation of investment risk taking and firm specific factor. 7 University of Ghana http://ugspace.ug.edu.gh Understanding the outcome of the effect between investment risk taking behaviour and firm’s operational factors will be of great interest to regulators, insurers, shareholders and policyholders. Hence, this paper sought to study the risk exposure of individual insurance firms and explore whether insurers adjust their investment risk taking behaviour based on the firm-specific and macroeconomic factors in the Ghanaian insurance industry. Based on the above discussions, these questions were raised; does firm specific factors such as firm size, claims, management expenses, reinsurance ceded, and investment income, group affiliation, GDP, inflation, and interest rate influence investment risk taking behaviour of insurers? 1.3 THE PURPOSE OF STUDY The purpose of the study is to convey to investor the investment risk taking activities of the insurance companies and also inform the regulator on how investment risk taking activities of insurers affects the performance of the insurance companies in Ghana. 1.4 OBJECTIVES OF THE STUDY The objectives of the study are to:  Determine the effect of firm specific factors on investment risk taking behaviour of insurance firms.  Investigate the impact of investment risk taking behaviour on financial performance of both life and non-life insurance.  Investigate which insurance company has the greatest risk exposure within the study period. 8 University of Ghana http://ugspace.ug.edu.gh 1.5 RESEARCH HYPOTHESES The following hypotheses are set for the study. 1. Ha: Insurers with larger capital ratio have more incentive for investment risk taking. 2. Ha: Insurers with larger firm size take on more investment risk. 3. Ha: Insurers with less claims and expense ratio take on more investment risk and vice versa. 4. Ha: Higher interest rate and inflation have no impact on investment risk taking. 5. Ha: There is no impact of investment risk taking behaviour on financial performance for life and non-life insurance companies. 1.6 SIGNIFICANCE OF STUDY Investment is an integral part of insurance activity and is also a measure of performance in the insurance industry. The issue of investment risk taking is very relevant to policy-makers and regulators. The study will inform regulators on the appropriate limit of investment risk insurance firms should take base on the level of operation risk insurers face. To maintain a financially sound and stable insurance environment, an insurer will know how to adjust the risk factors linked to their business operations which have direct influence on their investment risk taking. The interest of both academia and practitioner on managing investment risk taking will be addressed. Policy makers will be updated on strategies of insurer’s investment behaviour. 9 University of Ghana http://ugspace.ug.edu.gh 1.7 LIMITATIONS This study is limited to 28 out of 42 insurance companies due to availability of data. The influence of regulation and governance on investment risk taking behaviour was not addressed due to lack of data. 1.8 ORGANIZATION OF STUDY The research was organised under five chapters. Chapter one entailed the background of the study, the statement of the problem, the purpose of study, the objective of study, the research hypotheses, significance and limitation of the study. Chapter two presents the literature review: theoretical review, and empirical studies relevant to the objective of this research and gives an overview of the Ghanaian insurance industry. Chapter three expounded on the research methodology and provided a detailed explanation of the variables used. Chapter four involves the presentation and discussion of results. Chapter five provided a summary of the findings, the conclusion and recommendations. 10 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.1 INTRODUCTION This chapter presents the relevant theoretical and empirical literature on the various views of investment risk taking. It is to provide a theoretical and empirical basis for the discussion of the findings of the study. The chapter is divided into four sub-sections, the first part presents a discussion of the various theories on investment risk taking behaviour. The second part present summary of empirical studies and the last two sections gives information on the overview of the Ghanaian insurance industry and the investment categories of insurance companies in Ghana. 2.2 THEORETICAL REVIEW 2.2.1 Background Risk Theory An economic theory relevant in explaining the relationship between investment risk taking and the risk faced by the operations of insurers is the background risk. Background risk is an uninsurable risk in the insurance demand. It is the risk emanating from the insurers operational activities which are not controllable. Given that insurance companies are financial institutions and therefore make vital decisions on portfolio choices, background risk originates mainly from insurers underwriting activities and the uncertainty associated with insurance claims. Earlier studies in developed countries on background risk by Gollier and Pratt (1996), Kimball (1993), Heaton and Lucas (2000), Huang and Wang (2013) and Jiang, Ma, An (2010) and Meyer an Meyer (1998), explained 11 University of Ghana http://ugspace.ug.edu.gh that investors experiencing higher background risk behave as if they were more conservative or risk averse, preferring safer assets. Gollier and Pratt (1996) in their study of risk vulnerability and tempering effect of background risk found that investors turns to be more risk averse when they are faced with several forms of background risk. Heaton and Lucas (2000) disclosed that due to high backgruond risk, the positive relationship between labour income and risky asset return reduces investment in risky assets. Jiang, Ma, and An (2010) using the mean - variance framework analyse the impact of background risk on investors porfolio chioce. Their findings showed that the existence of background risk changes the position of the efficient frontier and that both the composition of the portfolio and the efficient frontier are affected by several backgroung risk factors. A study by Huang and Wang (2013) also investigated the individual portfolio selection in the presence of background risk and found that when background risk exist, the portfolio variance difference between ignoring or not ignoring background risk equals the variance of a particular portfolio. Studies above gives evidence of how background risk affect invesments. According to Chen et al. (2007) the uncertainty of insurance business operation is very difficult to trade away and beyond the control of portfolio managers, hence these uncertainties can be considered as a type of background risk for insurers portfolio investment decisions. In this study, background risk is considered as the firm’s specific factors in the insurance settings that influence the decision of insurer’s investment risk taking. In relation to insurers underwriting risk and financial risk bearing capacity as a description of background risk, the factors considered are underwriting risk (loss ratio (LR) and expense ratio (ER)), capital ratio (CR), firm size (Size), leverage (LevR), investment income (INCR), and reinsurance ceded (RIR). As the theory explains, 12 University of Ghana http://ugspace.ug.edu.gh insurers with less background risk, that is insurance companies with large firm size, having more access to capital, less leverage, less underwriting risk (less claims and expenses), ceding more premiums to reinsurers have the capacity to take on more risk in investment securities. 2.2.2 Portfolio theory Portfolio theory by Markowitz (1952) have been extensively used in literature (Balogun, 2013; Elton & Gruber, 1997; Hagin, 1979; Hutchinson, Seamer & Chapple, 2015; Mangram, 2013). The choice of portfolio theory is based on the concern of investors and economic agents who acts under uncertainty. The basic concepts of financial instruments are based on the returns of an investment and the risk associated with the investment portfolio. Practically, in finance the higher the risk, the higher the returns. Investment returns are basically the financial outcome for investors whiles the risk is the uncertainty about the outcome the investment will yield. Two main assumptions on portfolio theory are; the normal distribution of expected return and the condition that individual investors are risk averse. The first assumption means that the expected returns and the risk associated with it are sufficient to explain the distribution of the returns. Whiles the second assumption clarifies that investors expects higher returns relative to a given risk taking and vice versa. Amidst insurance expectation to receive higher returns per the risk taking, their choice of investing in a particular portfolio is affected by certain background risk from their operations. The expected return of an investor’s portfolio is calculated by 𝑛 𝐸(𝑅𝑝) = ∑ 𝑤𝑖𝐸(𝑅𝑖) 𝑖=1 13 University of Ghana http://ugspace.ug.edu.gh where 𝐸(𝑅𝑖) is the expected return of the 𝑖 𝑡ℎ security and 𝐸(𝑅𝑝) is the expected return of the portfolio securities. 𝑤𝑖 is the proportion of weights attached to the 𝑖 𝑡ℎsecurity and 𝑛 is the number of securities in the portfolio. The risk on the return is computed by 𝑛 𝜎 2𝑝 = 𝑉𝑎𝑟 (𝑅𝑝) = 𝑉𝑎𝑟 (∑ 𝑤𝑖𝑅𝑖) 𝑖=1 𝑛 𝑛 = ∑ 𝑤 2𝑖 𝑉𝑎𝑟(𝑅𝑖) + ∑ 𝑤𝑖𝑤𝑗 𝐶𝑜𝑣(𝑅𝑖 , 𝑅𝑗) 𝑖=1 𝑖,𝑗=1,𝑖≠𝑗 where, 𝑉𝑎𝑟 (𝑅𝑝) is the variance of the portfolio, 𝑉𝑎𝑟(𝑅𝑖) is the variance of the 𝑖 𝑡ℎ security, 𝑤𝑖 ,𝑤𝑗 is the proportions of weights attached to the 𝑖 𝑡ℎ, 𝑗𝑡ℎ securities respectively and 𝐶𝑜𝑣(𝑅 , 𝑅 ) is the covariance between the 𝑖𝑡ℎ𝑖 𝑗 , 𝑗 𝑡ℎ securities. 2.3 EMPIRICAL REVIEW Extant literatures have mostly reported a positive interaction between investment risk and capital. The estimation of capital-to-asset ratio and risk by Cummins and Sommer 1996 confirms a positive relationship between risk and capital. Meaning insurers attain targeted solvency risk through the choice of capitalization and risk. They found a positive and statistically significant relationship between capital and portfolio risk. Their explanation was that higher risk firms holds more capital and better capitalized firms are able to take more risk. Baranoff and Sager (2002) explore the 14 University of Ghana http://ugspace.ug.edu.gh relation between capital and risk of life insurance firms by adopting a simultaneous-equation partial adjustment model. Their findings also depicted a significantly positive relation between capital and risk. Following the same positive trend, in West Africa, Bokpin (2016) analyse bank governance, regulation and risk taking in Ghana and also found a positive and significant relationship between capital adequacy and risk taking. He explained that an increase in capital leads to higher risk taking, resulting in a lower insolvency risk, which means that well capitalized banks are associated with lower risk of insolvency, hence, availability of more funds give banks the incentive to shoulder more risky investment. Another study by Manka and Belgacem (2015) on the interaction between risk taking, capital and reinsurance of U.S. property and casualty insurance recorded a positive and significant relationship between capital and risk taking. Similarly, Zou et al. ( 2012) found a positive relationship between capital ratio and investment risk taking. They measured investment risk taking as proportion of investment assets in common stocks. Their findings is consistent with Gorter and Bikker (2013) on investment risk taking by institutional investors of pension funds and insurance firms in the Netherlands, showing capital ratio highly significant at 1%, which means investors with larger capital buffers take on more risk in equity investment. They explained that insurers are more risk sensitive in their asset allocation because of the exposure to regulatory differences and financial distress costs. On the contrary, in the banking sector Berger (1995) found a negative link of capital-asset-ratio with portfolio risk. He argued that existence of a negative association is as a result of guaranteed fund for insurers and deposit insurance for the banking sector which induces a higher risk taking at a lower capital. 15 University of Ghana http://ugspace.ug.edu.gh In literature, underwriting and investment are viewed as the two core related activities in the insurance industry (Adams & Buckle, 2003). These activities are associated with specific risk called underwriting and investment risk respectively which influence the strategic financial decision of the industry. Meanwhile there are few studies in literature on interrelation between these two risks. Hammond, Melander and Shilling (1976) investigated the linkage between underwriting and investment risk of U.S property and casualty insurance industry by regressing underwriting risk on investment risk. They measured underwriting risk as a ratio of net premium to equity and investment risk as the standard deviation of return on investment or proportion of investment in equity, and recorded a negative relationship between investment and underwriting risks. Hammond et al. (1976) established that insurance managers’ trade-off between the two forms of risks in other to get the desired level of overall risk to achieve a giving level of profitability. Still in the U.S Zou et al. (2012) in a similar study by Hammond et al. (1976) on the relationship between underwriting and investment risks found different results between the two forms of risk. They measured underwriting risk as a combined ratio of loss ratio and underwriting expense; and investment risk as a proportion of investment in common stock. Their results showed a positive but not significant relationship between underwriting and investment risk, contrary to the findings of Hammond et al. (1976). Their explanation was much alike with earlier study by Hammond et al. (1976) that insurers’ trade-off between the underwriting and investment risk to achieve a desired level of risk for a higher rate of return. In the health insurance business Cheng, Elyasiani, and Jia (2011) studying the institutional ownership and stability on U.S life-health insurance industry documented a positive relation between investment and underwriting risk in their study. They measured underwriting risk as the 16 University of Ghana http://ugspace.ug.edu.gh ratio of written premiums in health business to total net premium while investment risk was measured by risk based capital factor comprising weighted investment proportion of stocks, bond, mortgages, real estates and loans scaled by total assets. In a prior study, Chen et al. (2007) examines how background risk from insurance operation affects the risk taking behaviour of insurers in their corporate bond investment. They recognized that insurers underwriting risk measured by standard deviation of insurers underwriting income and underwriting cash flows, have a strong impact on their risk taking in corporate bond investment. They recorded a positive relationship and explained that insurers with higher underwriting risk tend to hold better rating bonds. The implication is that insurers are more resilient to risky investment when they are faced with higher underwriting losses and tighter capital constraint. Ren et al. (2011) found a negative and positive association between investment risk taking and underwriting performance depending on the alternate investment risk measure used. In their study, insurers’ loss ratio representing underwriting risk was measured as a combined ratio of expense and claim ratio and the results was consistent with their hypothesis of an inverse relation between investment risk taking and underwriting performance. The negative relationship was significant to stock insurers but insignificant to mutual insurers. Moreover, on a robustness check, they alternate the investment risk measure by using proportion of invested asset in cash, bonds and short term investment, representing low investment risk. With this low investment risk measure, they recorded a positive relation between loss ratio and investment risk, which was significant to stock insurers and insignificant to mutual insurers. Their study suggests that insurers in the property or casualty industry in the United States of America would take less risk in risky investment assets to offset the stress from greater underwriting losses, hence investors should consider the trade-off 17 University of Ghana http://ugspace.ug.edu.gh between underwriting and investment risk when evaluating the rate of return on investment. Consistent with Cummins and Sommer (1996) and Staking and Babbel (1995) insurers maintain a balance between underwriting risk and leverage and investment risk to achieve a desired insolvency risk level. Similarly, a study by Andersson, Lindmark, Adams and Upreti (2013) in Northern Europe, showed a negative and statistically significant link between underwriting risk and investment returns for mutual insurers but not for stock insurers. They attribute this to the fact that greater amount of investment risk taking by Swedish fire insurers generated lower return on their invested asset than fire insurers that assumed less risky business. Lu (2011) investigates the extent to life insures characteristic and the characteristic of bond affects investment decisions. The results shows life insurers of higher underwriting risk take lower investment risk. Strong evidence in literature expounds that size affects risk taking of insurers. Firm size is an important factor in the interaction of financial institutions risk taking and investment decisions. In the banking sector, Bhagat, Bolton and Lu (2015) investigated the link between firm size and risk taking of financial institutions and reported a positive association between firm size and risk taking. Kadapakkam, Kumar and Riddick (1998) also examines the degree to which firm size and cash flows influences firms investment by invstigating six OECD countries (Canada, France, Germany, Great Britain, Japan and the United States). Their aim was to investigate how internal funds is affected by firm size since there is much evidence that financially constraint firms have limited internal funds available for investment. They first examined all firms regardless of size and found that the amount of corpoate investment is affected by internal resourses and these facts confirm earlier studies recording a positive relationship between cash flow and investment. 18 University of Ghana http://ugspace.ug.edu.gh Repeatimg the same anaysis using three different measures of firm size, they found that the cash flow investment sensitivity is generally highest in larger firm size group and smallest in smaller firm size group. In the insurance sector, several research outcomes record a positive relationship between firm size and investment activities. Audretsch and Elston ( 2002), Chen et al. (2007), Gorter and Bikker (2013), Ren et al. (2011), and Samina (2012) found a positive relationsip between firm size and investment risk taking and concluded larger insurers have incentives to shoulder riskier investment. Larger firms have the capacity to finance capital expenditures through internal resource and issuance of equities while smaller firms are limited in the extent of their internal earnings and the potential for issuing equities (Audretsch & Elston, 2002). Leverage is another essential factor in the investment decision of a firm. In Pakistan, Ahmad, Zulfiqar, Bilal and Ahmad (2013) studied the impact of financial leverage on firms investment. They found that financial leverage has significantly negative impact on firms investment. This means the highly leverage firms invest less. Khan and Naeemullah (2015) also recorded a negative relationship between leverage and investment. They further extablished that the relationship is much stronger for firms with lower opportinities for growth than those with higher opportunities. The above findings are consistent with several studies in literature which also established a negative relationship between leverage and investment in developed countries like Canada, USA, United Kingdom (Aivazian, Ge, & Qiu, 2005; Andersson et al., 2013; Dang, 2011; Lang, Ofek, & Stulz, 1996). Group affilation, a strategic incentive on investment risk taking has significant record in literature. Chen et al. (2007) and Gorter and Bikker (2013), examine the effect of group affilation on investment risk taking of insurance companies and pension investors and found that group 19 University of Ghana http://ugspace.ug.edu.gh affilaition related significantly positive with investment risk taking. The reason being that insurers belonging to another financially viabrant group holds less capital and channel larger portion of their assets into strategic investment to recoup profitable returns. Reinsurance is a required practice by all insurance companies whereby an insurer transfer portion of its risk to another party called the reinsurer, by ceding part of its premiums received to the reinsurer. The intent of reinsurance as a risk management tool is to reduce the risks associated with underwriting policies by spreading the risk across alternative institutions. A study by Manka and Belgacem, (2015) explained that insurers ceding more to reinsurers have minimal burden for claim settlement since the underwriting risk borne from rapid claim payments and benefits are absorbed by reinsurers. Hence, insurers with greater ability to cede more have the security to shoulder risky investments (Adams Mike, 1996; Baltas, Frangos, & Yannacopoulos, 2012; Chen et al., 2007; Gorter & Bikker, 2013; Peng, Chen, & Hu, 2014). Infltion and interest rate among other variables affect investment decisions of financial institutions. Taking into consideration the various investment categories, higher inflation is set to have a negative effect on firms investment. A temporal increase in nomimal interest rate causes a short term increase in output and investment. Under a passive interest rate rule, financial authorities often respond to inflation defects by lowering interest rate, explaining the uniqueness of equilibrium (Dupor, 2001). Interest rate on the other hand, have positive effect on investment (Brewer, Carson, Elyasiani, Mansur, & Scott, 2007; Muhammad, Lakhan, Zafar & Noman, 2013). According to Berdin, Elia and Helmut Gründl (2014) low interest rate affects the stability of life insurance industry particularly countries where products of high guaranteed returns sold in the past still represent a share in the total portfolio. Their results in the study on the effect of low interest 20 University of Ghana http://ugspace.ug.edu.gh rate environment on life insurance, suggest that long period of low interest rates affect the solvency of life insurers, leading to likelihood of default in less capitalized insurance companies. In an earlier study, Antolin, Schich, and Yermo (2011) emphasized longer period of low interest rates affect investment opportunities and this can have adverse on insurers whose liabilities consist of fixed investment returns. 2.4 DETERMINANTS OF INSURANCE INVESTMENT The premiums insurance companies received from the policyholders are invested in securities to generate returns on their investments. A fundamental concept in finance records that investments which carries a higher risk has the potential to generate higher returns. For instance a zero-risk investment such as T-bills has low returns whiles stock or equity investments has the potential to make investors very wealthy. The nature of insurance business determines the types of investment category insurers are likely to invest their assets. However this is generally not the case in Ghana where government borrowing keep rising plus high volatility in our markets. Thus making risk free investments more attractive with better rates to Ghanaian investors because it is safer and returns are predetermined. Investment choices are also driven by major macroeconomic factors, like interest rate and inflation which has a direct influence on the investment decision of insurers. A major characteristic of the insurance business is that premium pay out always precedes payment of claims or any benefit. As a result, insurance companies are required to have sufficient liquid assets to meet the obligation of claim payment and also have extra cash available for investment to improve financial performance. 21 University of Ghana http://ugspace.ug.edu.gh The insurance industry as a financial institution is made up of two distinct markets; we have the life insurance and the general or nonlife insurance business. The activities of the two market differ in terms of operation, vulnerability, duration of liability and investments (Brockett, Cooper, Golden, & Pitaktong, 2013). These differences in activities have significant effects on their investment behaviour. General insurance businesses or non-life insurance providing financial protection cover on physical damages to life and properties are rapidly confronted with uncertainty of loss occurrence and are required to settle policyholders claim payment if the loss of an insured risk occurs. This means funds, in the form of liquid assets must be available to meet the contingency of any claim payment. Hence, general insurers most likely invest larger proportion of funds in short term investments due to the short term nature of rapid loss occurrence of the insured risk. Classes of non-life business include among other things, general liability, motor insurance, marine and aviation, fire burglary and property insurance, and accident policies. Life insurance businesses are mostly long term in nature and pay benefit of a certain amount to the insured or their nominated beneficially upon a certain event such as death of the individual who is insured. Life insurance includes insurance business of the following classes; whole life, universal life insurance, funeral policies, endowment, term policies and unit linked product. Life insurance involves the undertaking of liability to pay money on death or the happening of any contingency dependent on the termination of human life. In life insurance, premiums become revenues when they become payable by the policyholder. The long term nature of life business makes insurers channel their funds into long term investment, such as corporate bonds and equity securities (Sunder & Monita, 2014). Life insurers have generally been conservative in their investment activities. Other factors the influences investment risk taking behaviour are corporate 22 University of Ghana http://ugspace.ug.edu.gh governance variables, example, ownership structure, board size, underwriting risk, regulation, product composition, size and leverage. 2.5 CATEGORIES OF INVESTMENT OF INSURANCE COMPANIES Every type of investment has characteristics which are unique. Some investments have high risk and others low risk. Due to the nature of non-life business, insurers are rapidly associated with high risk, hence, majority of non-life insurance investment assets are invested in stable and liquid investment categories while life insurers invest more of their assets in equity securities. The investment asset categories of insurance companies are outlined below. 2.5.1 Investment Properties Property investment involves real estate investment through the purchase of buildings or part of building either on freehold or held under financial lease with an ultimate goal of earning returns in the form of income on the property purchased. Property investment can also be the acquisition of lands under operating lease or can also be freehold held for long term rental yields which are not occupied by the company. This type of investment can be either long term, for instance the purchase of buildings or short term, such as the renovating of properties and selling them out and it is carried out at fair value, representing the market value. 23 University of Ghana http://ugspace.ug.edu.gh 2.5.2 Equity Investment Investments in equities can also mean investments in stocks or shares either listed or unlisted. This is usually accounted for at fair value with unrealized gains and losses. It comprises trading security or available for sale securities. A firm’s intention to buy securities and resell in a short term before the maturity period is termed as trading security. Equity investments involves quoted market price in an active market which are best evidence of fair value (Denis, 1994). Stock or shares are investment made to a company with the aim of gaining returns in the form of dividends. In Ghana, the stock exchange regulate transaction of shares securities. Shares are exposed to high volatility causing greater investment risk hence investors have more expectation of higher returns though uncertain and quite risky. 2.5.3 Fixed Deposit Fixed deposit is an investment account where money is deposited for a fixed period. This form of investment is risk free financial instrument issued by banks with certain returns at a fixed interest rate and fixed maturity period usually ranging from 30 days to two years. Interest rate ranges from bank to bank and are negotiable depending on the amount invested and the period chosen. An advantage of this investment is that cash can be redeemed before the maturity date. Most insurance companies in Ghana per their balance sheet accounts invest more of their funds in fixed deposit investments. This is because by regulation insurance investors in Ghana are required to invest 0- 35% in T-bills, cash and fixed deposit. Moreover, most insurers are very comfortable with fixed deposit investment because its risk free and interest rates are good, often negotiable across banks. 24 University of Ghana http://ugspace.ug.edu.gh The payoffs are also predetermined, making it easier for insurers to channel their investment funds to fixed deposit investment categories that will yield higher returns. 2.5.4 Treasury Bills Treasury bill is a short term investment issued by the government with fixed maturity dates and interest rates. In Ghana, treasury bills maturity dates fall under three categories; 91 days, 182 days, one year and two years notes. High interest rate recorded by the Bank of Ghana over the past few years encourages financial institutions like insurance companies in Ghana to channel larger sums of their investment into treasury bills. Treasury bills seems to be a desirable investment by insurance companies because of the safety and liquidity it provides. 2.5.5 Cash Accounts The cash at bank is a savings investment account which is more liquid but generates very low returns. There is nothing like fixed interest rate or fixed maturity period on savings account. The decision of insurers regarding the amount of capital to hold is very critical to its operations and financial stability. Although the choice of insurers holding large amount in cash makes them financially flexible, the trade-off will be the lower gains on cash as compared to the higher returns on less liquid investments forgone (Colquitt, Sommer, & Godwin, 1999). Cash instrument offers complete capital security subject to credit risk. The return on cash accounts varies according to the levels of short term interest rates, and does not offer income security. 25 University of Ghana http://ugspace.ug.edu.gh 2.6 OVERVIEW OF THE INSURANCE INDUSTRY The insurance industry in Ghana is regulated through constant supervision and monitoring, by the National Insurance Commission (NIC) under the insurance Act, 2006 (Act 724) to ensure effective administration, regulation, and control of insurance business to protect the policyholder and the industry. As at 2015, the insurance industry recorded 26 non-life, 23 life, 3 reinsurance, and 69 brokers, 1 loss adjuster, 1 reinsurance broker, 6000 insurance agents as intermediaries (NIC, 2015). Innovation, competition and new product development by insurers enhance premium growth in the industry, providing enough funds for investments. Investment activities in the industry are classified into two; long and short term investment. Long term investments are categorized into quoted and unquoted shares, investment properties, corporate bonds and others while short term investments are treasury bills, fixed deposits, call accounts, unit trust and bonds. Investment assets are classified based on insurance investors risk appetite and regulatory rules. The investment practice directives of insurance companies in Ghana have gone through three regimes: prior to 2008 the insurance companies were allowed by the commission to manage their investments without any restriction on their asset categories. In 2008, the National Insurance Commission order for the separation of life and non-life business was implemented. During this era, insurers in the industry were directed to follow an investment mix to monitor their investment assets allocation required to meet solvency margin. However, the commission currently is not very strict on these investment mixes because of the immediate plan of introducing the risk base supervisory system which is able to check the imbalance of the 26 University of Ghana http://ugspace.ug.edu.gh investment allocations required for solvency calculations, and issue instant sanctions when appropriate. That notwithstanding, the commission requires all insurers, both life and non-life to establish and maintain an investment strategy deem appropriate by their boards, for the nature, size and complexity of the business. Insurers are expected to have procedures and controls sufficient to ensure that the investment strategies and policies are effectively implemented. The investment mix categories required by Ghanaian insurance companies are: Table 2.1: Investment mix and categories Investment Proportion of Investment Portfolio Life Nonlife Government securities, cash At least 35% At least 35% and deposits (excluding statutory deposit) Statutory deposit At least 10% of minimum At least 10% of minimum capital, capital Listed Stocks 0 – 30% 0 – 30% Unlisted Stocks 0 – 20% 0 – 10% Mutual Funds 0 – 20% 0 – 20% Investment Properties 10% - 30% 0 – 20% Other investments approved 0 – 10% 0 – 10% by NIC Source: National Insurance Commission annual reports, 2008. 27 University of Ghana http://ugspace.ug.edu.gh Without limiting the asset allocation of insurers, the investment strategy and policies of insurers address the risk profile of the insurer; mixture and diversification of the investment type, including long term asset mix; the extent to which the holding of some asset is restricted or disallowed, for example illiquid and volatile assets. Again the investment policy shall address the concern on insurers being made accountable for all their assets transactions and the associated risk. Recently, the Risk Base Supervisory system (RBS), a world leading standard which require all insurers to focus on managing all the risk facing their organization has been introduced. As at 2015, the NIC is strictly supervising the use of the new system by all registered insurance companies in Ghana. As recommended by Akotey, Sackey, Amoah, and Manso (2013) the introduction of the RBS in the Ghanaian insurance industry is appropriate, hence its implementation now is a prompt step in the right direction. With the current system, compliance of the investment mix on various asset categories which was introduced earlier is barely minimal. Moreover, insurers are allowed to invest any of its surplus assets, which are those assets that are not required to meet the capital adequacy of 150%, outside Ghana. The new system monitors investment activities by assigning weights on the various investment categories and automatically issuing sanctions to insurer’s asset allocation below the benchmark. The RBS helps to check the imbalance in the solvency calculation. 28 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE METHODOLOGY 3.1 INTRODUCTION This chapter explains the research methodology. It describes the type of study carried out, the research model and the rationale for selecting such a methodology. This chapter seeks to justify the methods employed in achieving the research objectives for the study. The chapter outline begins with the data source, followed by model specification, explanation of variables and measures of insurer’s investment risk taking exposures. It also gives details of the data source and data analysis tools adopted. 3.2 THE DATA SOURCE The study utilizes annual data of 28 insurance companies in Ghana from 2008-2014, specifically 12 life and 16 nonlife insurance companies. Secondary data on investment activities, capital, firm size, liability, management expenses, claims, reinsurance ceded, investment income, were obtained from the National Insurance Commission annual reports and selected insurance company’s annual financial statement. Macroeconomic data was obtained from Bank of Ghana quarterly statistical bulletins. 29 University of Ghana http://ugspace.ug.edu.gh 3.3 MODEL SPECIFICATION The research investigated the effect of firm specific factors on investment risk taking behaviour of life and non-life insurers in Ghana over the period of 2008-2014. The study employed panel data analysis technique to achieve its set objectives. The choice of panel data analysis was necessitated by the financial modelling where data used comprise both time series and cross-sectional elements. In recent times, the relevance of time-series variation of investment activities in the financial sector calls for an empirical search to justify the variation of specific factors that influences the investment risk taking across insurance firms over time. Employing a panel data analysis gives evidence of the effect of firm’s specific factors on insurers risk taking behaviour. The estimates generates a more conclusive results than using only time series or cross-sectional. Moreover, it is easy to address a broader and complex issues using panel analysis than time series or cross- sectional. Panel analysis removes omitted variable bias and tackles the problem of adding on variables that are not relevant to the model. A panel data is analysed either under fixed or random effects. The fixed effect model analyses the impact of the predictor and the explanatory variables within an entity and vary over time. However, under random effect model the variation across entities is assumed to be random and uncorrelated with the predictor or explanatory variables in the model (Oscar, 2010). A Hausman test at 5% significance level is conducted to ascertain either to use fixed or random effect for the regression analysis. The fixed effect model is used if the p- value is less than 0.05, otherwise the random effect model stands as best for the analysis. From the results of the Hausman test we settled on the random effect model for all the empirical analysis. The study adopts two alternate measures for insurer’s investment risk taking. The proportion of invested asset in equity (stock or shares) plus property investment represents high investment risk taking and proportion of invested asset in short term securities, denotes low investment risk taking 30 University of Ghana http://ugspace.ug.edu.gh (Gorter & Bikker, 2013; Ren et al., 2011). We considered firm specific variables namely, capital ratio, leverage ratio, size, underwriting risk (loss and expense ratio), investment income ratio, reinsurance ceded and group affiliation. Since macroeconomic have significant effect on investment activities, we added inflation, real GDP and interest rate as control variables to the model. The main variables of interest are size, capital ratio, investment income and underwriting risk (proxy by loss and expense ratio) for non-life and life insurance companies respectively. 3.4 EXPLANATION OF VARIABLES 3.4.1 Capital Ratio (CR) Capital adequacy explains the financial capacity of a firm to take on more risk. It is a widely used measure in the banking sector in determining the amount of capital a firm should hold compared to its risk taking. The insurance company as one of the risk management institution, capital ratio indicate insurer’s ability to withstand risk. More often, insurers with high capital ratio are considered very strong and stable. The role of capital to regulators and stakeholders cannot be over emphasized as it ensures that there is enough assets to cover all liabilities under adverse markets conditions. The recorded capital on an insurance company’s balance sheet in relation to its liabilities to policyholders determines the level of insolvency. Capital ratio is measured by firm’s equity to total asset, where firm’s equity is measured as total asset minus liability. Firms with more capital can shoulder riskier portfolio assets which intern will generate higher returns. Availability of more capital gives insurers the incentive to take on more risk. Meanwhile, there are cost involve in holding more capital and the pressure from stakeholders push insurers to efficiently use their capital. Intuitively, financially constraint insurers have less capital to venture risky investment. 31 University of Ghana http://ugspace.ug.edu.gh 3.4.2 Leverage Ratio (LevR) Several forms of regulation are applied to protect the solvency of insurers and to minimize the low level of insolvency. The most popular way of regulating insurer’s financial stability are constraint on its leverage. Other studies measures leverage as the ratio between premiums and equity; assets divided by liability; reserves to surplus; reserves to equity (Öner, 2015; Lang, Ofek, & Stulz, 1996; Lu, 2011; Ho, Lai, & Lee, 2013; Chen & Wong, 2004). Earlier studies measured the relationship between leverage and growth using pooling regressions and ignored the individual firm effects. In this study, leverage ratio is measured as total liability divided by total capital (equity) and uses panel data methodology to examine the effect of leverage among individual firms investment risk taking. Leverage determines the amount of debt of an insurance company relative to its equity. A higher proportion of an insurance company’s leverage ratio may have a negative effect on investment activities (Andersson et al., 2013). Financial risk is the risk that a company defaults on the repayment of its liability. When liability to equity ratio is high, it increases the tendency that the company defaults, causing financial risk to the firm. Clearly, this is not good for investors as it increases the risk associated with their investment activities which causes them to require a higher rate of return to compensate for the additional risk. Consequently, the increase in the required return on investment and leverage means an increase in the cost of capital for the company. 3.4.3 Firm Size (Size) Size plays an essential role in determining the risk appetites of insurers on their investment activities. Larger insurers seems to be more diversified and have better access to capital, giving them the potential to shoulder riskier investment than smaller firms (Gorter & Bikker, 2013). 32 University of Ghana http://ugspace.ug.edu.gh Insurers with larger firm size have limited information asymmetry between managers and potential investors which reduces the cost of capital to take on risk in investing their asset. A study by Touny and Shusha, (2014) documented that the benefits of economy of scale and good set up of risk management organizations by larger insurers enhances their risk bearing capacity. For example, insurance companies with larger size can avoid the volatility on the investment returns by engaging qualified investment managers to handle their large and well diversified asset portfolios. Contrary, smaller insurance firms can diversify part of the risk on their liabilities through reinsurance, though reinsurance can be very expensive imposing more financial tension on smaller firms. Size is considered to be a significant factor of investment risk taking, though the impact is less on smaller insurers. Earlier studies measured firm size as the natural log of investment portfolio size, log of total assets, and log of gross premium and number of employees (Bhagat, Bolton, & Lu, 2015; Chen, Yao, & Yu, 2007 ; Lin, Lin, & Wang, 2012). This study measures firm’s size as the natural log of total assets on the balance sheet. 3.4.4 Expense Ratio (ER) Expense ratio is one of the proxies for underwriting risk. This ratio is expressed as total management expense to gross premium. The expense ratio determines the efficiency of management to write premiums given the cost of operation; a lower expense ratio shows a more efficient management. This implies that insurance companies with lower expense ratio can diversify more their assets into riskier investment. Contrary, a higher expense ratio can affect the proportion of investment risk taking. Notwithstanding, the increase in expenses can result from recruiting expertise in managing investment funds which intern will generate higher returns. 33 University of Ghana http://ugspace.ug.edu.gh 3.4.5 Loss Ratio (LR) The loss ratio is measured by the total claims incurred to gross premium, and it measure how efficient the insurance company performs under its underwriting functions. The claim consists of both outstanding and paid claims. A lower ratio means a better underwriting function. Underwriting risk are proxies by expense ratio, loss ratio or combined ratio, which is a combination of the two. A good underwriting performance, will increase insurers desire to pursue profitable investment with high risk. Non-life insurance claims history are often frequent due to the nature of businesses, hence a higher ratio can put much pressure on the investment activities of insurance companies. 3.4.6 Investment Income Ratio (INCR) This ratio is measured by generated investment income to total investment asset. The ratio explains the quality and adequacy of the returns on the investment assets. Investment of assets involves risk taking, which means insurers will require a return on their investment to cover the risk and earn rewards. 3.4.7 Reinsurance Ratio (RIR) This ratio is computed by premiums ceded to reinsurers to total premium of the insurance company (Garven & Lamm-Tennant, 2003). A strategic way to minimize volatility, insolvency and underwriting risk is through reinsurance (Adams, 1996). Reinsurance is used to solve problems of excessive risk taking of insurers which reduces their ability to maintain projects with positive net present value. As insurers are able to transfer part if not all of their losses to reinsurers, they are motivated to take on more risk in assets investment (Zhu, Deng, Yue, & Deng, 2015). Non-life 34 University of Ghana http://ugspace.ug.edu.gh insurers cede more premiums to reinsurers due to the rapid nature of their loss occurrence and as a result have less underwriting risk which gives them the capacity to take more investment risk. 3.4.8 Group Affiliation (GA) Extant literature proofs that insurers affiliated to a group can shoulder more risk since such insurers have access to the internal funds of the groups they are affiliated (Gorter & Bikker, 2013). Again it is much easier for affiliated insurers to get financial support from parent groups when they are cash trap. A studied by Powell & Sommer (2007) gives evidence that capital market within insurance groups play a significant role in the investment behaviour of affiliated insurers. Following the measure of other studies, group affiliation is assigned 1, for affiliated insurance company or zero otherwise. 3.4.9 Inflation (Infla) Inflation measures insurer’s ability to sustain the increasing pricing of premiums from year to year. The rate of inflation is important as it represent the rate at which the real value of an investment is achieved and the loss in spending power over time. Inflation tells investors how much return of their investment will be more profitable. In other words, as inflation rises and falls insurers get the motivation to adjust their investment decisions. Inflation rate is measured by (𝐶𝑃𝐼𝑡 − 𝐶𝑃𝐼𝑡−1)/𝐶𝑃𝐼𝑡−1 , where CPI represent the consumer price index. This figure is yearly reported by the Bank of Ghana annual report. 35 University of Ghana http://ugspace.ug.edu.gh 3.4.10 Interest rate (IntRate) Interest rate has notable significant impact on investment activities of financial institutions (Balogun, 2013; Begenau, Piazzesi, & Schneider, 2015; Chiu & Wong, 2014; Dell’Ariccia, Laeven, & Marquez, 2014; Dupor, 2001; Staking & Babbel, 1995). All insurers are affected by interest rate but some categories of insurance business are more vulnerable than others. Life insurance business experience greater shocks on changes in interest rate since their major source of earning is from their investment income (Brewer et al., 2007). For example, as interest rate reduces, insurers become desperate for higher returns and shift towards riskier investments which is a major threat of life insurers relative to the long term nature of insurance category and investment. Life insurance exposure to low interest rate risk affects their investment on fixed income securities and their unique liability. This is due to the fact that their asset and liability are significantly influenced by interest rate movement. Non-life insurers on the other hand benefits from increase in short term interest rate since larger portion of their funds are in short term investment. Interest rate in this research is measured as the annual interest rate on treasury bills. For the study, interest rate is expected to have a positive relationship with investment risk taking. 3.5 MEASURE FOR INVESTMENT RISK TAKING EXPOSURE The coefficient of variation (CV) was used as a proxy for investment risk taking exposure and the company with greatest CV is taking as having the greatest risk exposure. The CV is measured as the ratio of standard deviation to the mean of the investment risk taking estimate of each company. 36 University of Ghana http://ugspace.ug.edu.gh The table below outline the variable descriptions and measures for both the dependent and independent variables. Table 3.1: Summary of variables Variable name Variable description Measure Expected sign High Investment risk 𝐸𝑞𝑢𝑖𝑡𝑦 + 𝑝𝑟𝑜𝑝𝑒𝑟𝑡𝑦 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 𝑖𝑡 𝐻𝑖𝑔ℎ𝑅𝑖𝑠𝑘𝑖𝑡 taking behaviour of 𝑇𝑜𝑡𝑎𝑙 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 company 𝑖 at time 𝑡 𝑖𝑡 Low investment risk 𝑆ℎ𝑜𝑟𝑡 𝑡𝑒𝑟𝑚 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡𝑖𝑡 𝐿𝑜𝑤𝑅𝑖𝑠𝑘𝑖𝑡 taking behaviour of 𝑇𝑜𝑡𝑎𝑙 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡𝑖𝑡 company 𝑖 at time 𝑡 Capital ratio of an 𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝑖𝑡−1 𝐶𝑅𝑖𝑡−1 insurance company 𝑖 at 𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡 + 𝑖𝑡−1 time 𝑡 − 1 Leverage ratio of an 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖𝑡−1 𝐿𝑒𝑣𝑅 𝑖𝑡−1 insurance company 𝑖 at 𝐶𝑎𝑝𝑖𝑡𝑎𝑙 - 𝑖𝑡−1 time 𝑡 − 1 Natural log of size an 𝑆𝑖𝑧𝑒𝑖𝑡−1 insurance company 𝑖 at Natural log of total assets + time 𝑡 − 1 Loss ratio of an 𝐶𝑙𝑎𝑖𝑚𝑠𝑖𝑡−1 𝐿𝑅 𝑖𝑡−1 insurance company 𝑖 at 𝑁𝑒𝑡 𝑝𝑟𝑒𝑚𝑖𝑢𝑚 - 𝑖𝑡−1 time 𝑡 − 1 High Investment risk 𝐸𝑞𝑢𝑖𝑡𝑦 + 𝑝𝑟𝑜𝑝𝑒𝑟𝑡𝑦 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 𝑖𝑡 𝐻𝐼𝑛𝑣𝑅𝑖𝑠𝑘𝑖𝑡 taking behaviour of + 𝑡 𝑇𝑜𝑡𝑎𝑙 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡company 𝑖 at time 𝑖𝑡 Low investment risk 𝑆ℎ𝑜𝑟𝑡 𝑡𝑒𝑟𝑚 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡𝑖𝑡 𝐿𝐼𝑛𝑣𝑅𝑖𝑠𝑘𝑖𝑡 taking behaviour of 𝑇𝑜𝑡𝑎𝑙 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 + 𝑖𝑡 company 𝑖 at time 𝑡 Investment income ratio 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 𝑖𝑛𝑐𝑜𝑚𝑒 𝑖𝑡−1 𝐼𝑁𝐶𝑅 of an insurance 𝑖𝑡−1 + 𝑇𝑜𝑡𝑎𝑙 𝑖𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 company 𝑖 at time 𝑡 − 1 𝑖𝑡−1 Expense ratio of an 𝐸𝑥𝑝𝑒𝑛𝑠𝑒𝑠𝑖𝑡−1 - 𝐸𝑅𝑖𝑡−1 insurance company 𝑖 at 𝑁𝑒𝑡 𝑃𝑟𝑒𝑚𝑖𝑢𝑚𝑖𝑡−1 time 𝑡 − 1 Reinsurance ratio of an 𝑅𝑒𝑖𝑛𝑠𝑢𝑟𝑎𝑛𝑐𝑒 𝑐𝑒𝑑𝑒𝑑 + 𝑖𝑡−1 𝑅𝐼𝑅 𝑖𝑡−1 insurance company 𝑖 at 𝑇𝑜𝑡𝑎𝑙 𝑝𝑟𝑒𝑚𝑖𝑢𝑚 time 𝑡 − 1 𝑖𝑡−1 37 University of Ghana http://ugspace.ug.edu.gh Group affiliation of an Dummy: 1 for an insurance 𝐺𝐴𝑖 insurance company 𝑖 company affiliated to a group or 0 + otherwise (𝐶𝑃𝐼𝑡 − 𝐶𝑃𝐼𝑡−1)/𝐶𝑃𝐼𝑡−1, CPI 𝐼𝑛𝑓𝑙𝑎𝑡 Inflation rate at time 𝑡 represent consumer price index + Annual recorded inflation 𝐼𝑛𝑡𝑅𝑎𝑡𝑒𝑡 Interest rate at time 𝑡 One year annual T bill rate + 𝑅𝑒𝑎𝑙𝐺𝐷𝑃𝑡 GDP at time 𝑡 Annual recorded GDP + The following specified models were used for the analysis: 𝐻𝑖𝑔ℎ𝑅𝑖𝑠𝑘𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐶𝑅𝑖𝑡−1 + 𝛽2𝐿𝑒𝑣𝑅𝑖𝑡−1 + 𝛽3 𝑙𝑛𝑆𝑖𝑧𝑒𝑖𝑡−1 + 𝛽4𝐿𝑅𝑖𝑡−1 + 𝛽5𝐸𝑅𝑖𝑡−1 + 𝛽6𝐼𝑁𝐶𝑅𝑖𝑡−1 + 𝛽7𝑅𝐼𝑅𝑖𝑡−1 + 𝛽8𝐺𝐴𝑖 + 𝛽9𝐼𝑛𝑓𝑙𝑎𝑡 + 𝛽10𝐼𝑛𝑡𝑅𝑎𝑡𝑒𝑡 + 𝛽11𝑅𝑒𝑎𝑙 𝐺𝐷𝑃𝑡 + 𝜆 𝑡 + µ𝑖 𝑡 (1) 𝐿𝑜𝑤𝑅𝑖𝑠𝑘𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐶𝑅𝑖𝑡−1 + 𝛽2𝐿𝑒𝑣𝑅𝑖𝑡−1 + 𝛽3 𝑙𝑛𝑆𝑖𝑧𝑒𝑖𝑡−1 + 𝛽4𝐿𝑅𝑖𝑡−1 + 𝛽5𝐸𝑅𝑖𝑡−1 + 𝛽6𝐼𝑁𝐶𝑅𝑖𝑡−1 + 𝛽7𝑅𝐼𝑅𝑖𝑡−1 + 𝛽8𝐺𝐴𝑖 + 𝛽9𝐼𝑛𝑓𝑙𝑎𝑡 + 𝛽10𝐼𝑛𝑡𝑅𝑎𝑡𝑒𝑡 + 𝛽11𝑅𝑒𝑎𝑙 𝐺𝐷𝑃𝑡 + 𝜆 𝑡 + µ𝑖 𝑡 (2) 𝐼𝑁𝐶𝑅𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐻𝐼𝑛𝑣𝑅𝑖𝑠𝑘𝑖𝑡 + 𝛽2𝐶𝑅𝑖𝑡−1 + 𝛽3𝐿𝑒𝑣𝑅𝑖𝑡−1 + 𝛽4 𝑙𝑛𝑆𝑖𝑧𝑒𝑖𝑡−1 + 𝛽5𝐿𝑅𝑖𝑡−1 + 𝛽6𝐸𝑅𝑖𝑡−1 + 𝛽7𝑅𝐼𝑅𝑖𝑡−1 + 𝛽8𝐺𝐴𝑖 + 𝛽9𝐼𝑛𝑓𝑙𝑎𝑡 + 𝛽10𝐼𝑛𝑡𝑅𝑎𝑡𝑒𝑡 + 𝛽11𝑅𝑒𝑎𝑙𝐺𝐷𝑃𝑡 + 𝜆 𝑡 + µ𝑖 𝑡 (3) 𝐼𝑁𝐶𝑅𝑖𝑡 = 𝛼𝑖 + 𝛽1𝐿𝐼𝑛𝑣𝑅𝑖𝑠𝑘𝑖𝑡 + 𝛽2𝐶𝑅𝑖𝑡−1 + 𝛽3𝐿𝑒𝑣𝑅𝑖𝑡−1 + 𝛽4 𝑙𝑛𝑆𝑖𝑧𝑒𝑖𝑡−1 + 𝛽5𝐿𝑅𝑖𝑡−1 + 𝛽6𝐸𝑅𝑖𝑡−1 + 𝛽7𝑅𝐼𝑅𝑖𝑡−1 + 𝛽8𝐺𝐴𝑖 + 𝛽9𝐼𝑛𝑓𝑙𝑎𝑡 + 𝛽10𝐼𝑛𝑡𝑅𝑎𝑡𝑒𝑡 + 𝛽11𝑅𝑒𝑎𝑙𝐺𝐷𝑃𝑡 + 𝜆 𝑡 + µ𝑖 𝑡 (4) 38 University of Ghana http://ugspace.ug.edu.gh Equations (1) and (2) represent the models for high and low investment risk taking behaviour respectively, where, 𝐻𝑖𝑔ℎ𝑅𝑖𝑠𝑘𝑖𝑡 𝑎𝑛𝑑 𝐿𝑜𝑤𝑅𝑖𝑠𝑘𝑖𝑡 are alternate measures of investment risk taking behaviour of insurance company 𝑖 at time 𝑡. 𝐻𝑖𝑔ℎ𝑅𝑖𝑠𝑘𝑖𝑡 , is the proxy for high investment risk taking and is measure by the proportion of equity (stocks or shares) investments plus property investments to total investment assets. 𝐿𝑜𝑤𝑅𝑖𝑠𝑘𝑖𝑡, represents the proxy for low investment risk taking and is also measured by the proportion of invested assets in short term securities to total investment. The choice of variables measures is based on earlier studies (Chen et al., 2007; Gorter & Bikker, 2013; Ren et al., 2011). Chen et al. (2007) measured investment risk taking as the percentage of invested assets in common stock of property and casualty insurers. Equation (3) and (4) represents the models for measuring the effect of investment risk taking behaviour on the financial performance of insurance companies. 𝐼𝑁𝐶𝑅𝑖𝑡 is the proxy for financial performance, which is the ratio of investment income (return on investment) to total investment asset of company 𝑖 at time 𝑡 (Akotey et al., 2013). The two models explains the effect of the alternate measures of investment risk taking on financial performance for both life and non-life insurance companies. It is expected insurance companies with very HighRisk scores will score low on LowRisk measure, and vice versa, hence their impact on financial performance will vary. 3.6 LIMITATION OF THE METHODOLOGY The regression output of the panel analysis is limited due to the short period used in the analysis. Ideally, it would have been better if the number of years considered was above 10 years but data was not available. Perhaps the Hausman test was in favour of random effect because the number of years was quite short. Moreover, with panel analysis, the results does not make specific 39 University of Ghana http://ugspace.ug.edu.gh recommendation for specific insurance companies. The results are general for all the insurance companies and it is even possible that the significant results may be influenced by just a few insurance companies. Accessing data from NIC office was very challenging. There was no proper database for easy accessible of information. Most of the records are in hardcopies and there is no system in place for updating information from individual insurance firms for research purposes. It took several attempts to manually retrieve data from the hardcopies of insurance company’s financial statements in the NIC office. Despite these challenges, the collation of data was carried out for all the variables employed in the analysis through a painstaking process to avoid any error of omissions. Random effect model was more effective and our results very authentic. Coefficient of variation does not exist when the mean is zero and if the data set contains both positive and negative numbers, the result of the coefficient of variation can be misleading because the mean can be zero or close to zero. However, the data set used in this study did not have any of the challenges stated above, all the data points were positive and none of the means computed was zero. 40 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR PRESENTATION OF DATA AND DISCUSSION OF RESULTS 4.1 INTRODUCTION This chapter gives detailed summary statistics of dependent and explanatory variables for both non-life and life insurance business in Ghana. The findings of the study and the results are presented and discussed linking to both theoretical and empirical literature. The discussion of the results are also linked to the objectives of the study. 4.2 DESCRIPTIVE ANALYSIS The investment risk taking behaviour of insurers differs in terms of the type of insurance business, as such it is very necessary to describe the variables used for each insurance type. The study used two alternate proxies to measure investment risk taking behaviour as the dependent variable; we have high and low risk denoted by HighRisk and LowRisk. In relation to our study, investment risk taking is defined as the capacity for an insurer to take the risk in investing its assets, either in the long or short term. High and low risk taking are measured by the ratio of equity securities plus property investments to total investment assets and the ratio of short term securities to total investment assets respectively. The independent variables are capital ratio (CR), leverage ratio (LevR), size, loss ratio (LR), expense ratio (ER), investment income ratio (INCR), reinsurance ceded (RIR), group affiliation (GA), inflation (Infla), real GDP and interest rate (IntRate). Non-life insurance companies represent about 57% of the total sample used while life insurance companies were 43%. 41 University of Ghana http://ugspace.ug.edu.gh Table 4.1 below gives a detailed descriptive statistics of all the variables used for both life and non-life insurance operations. The average of the two measures of insurer’s investment risk taking behaviour taken as high risk (low risk) - HighRisk (LowRisk) are 0.2249 (0.6296) for non-life insurance and 0.6745 (0.2769) for life insurance. This is an indication that on the average, due to the short term nature of non-life insurance businesses, insurers invested 63% of their assets in short term investments (low risk) and 22% in long term investments (high risk). Life insurers on the other hand invested 67% in long term investment (high risk) and 28% in short term investment since the nature of life insurance contract involves more long term activities. Capital ratio (CR) which measure financial capacity of the insurance firm on average stood at 41% for non-life and 26% for life. This means that on average, non-life businesses have more access to capital than life insurers. Leverage recorded a mean of 1.4996 and 2.8023 for both non-life and life respectively. On average approximately 150% and 280% leverage shows that both non-life and life insurance liabilities far exceeds their equity. This points to the fact that the insurance company’s activities are financed by other funds, for example premiums received from clients rather than from the company’s own equity (capital). Size, measured by the natural logarithm of firm’s total asset had a mean of GH¢29.5 million for non-life and GH¢47.8 million for life insurance firms. Thus, it can be deduced that life insurance on average has larger assets than non-life (see Fig 4 in Appendix 1). This confirms the industry annual report in 2014 (NIC, 2014) that life insurance assets keep growing more over the years under the study than non-life. Loss and expense ratios are proxies for underwriting risk and the means stood at 41% and 124% for non-life, 40% and 51% for life. And on average investment income ratio was 21% and 13% for non-life and life insurance respectively. Perhaps due to the frequent payment of losses by nonlife insurance, reinsurance ceded on average was 31% for nonlife 42 University of Ghana http://ugspace.ug.edu.gh and 3% for life insurance. Refer to Table 4.1 for the descriptive values of other variables captured in the study. Table 4.1: Descriptive statistics for Nonlife and Life Insurance Variable Ins. Type Obs. Mean Std. Dev. Min Max Nonlife 112 0.2249 0.2352 0.0000 0.9386 HighRisk Life 84 0.6745 0.2253 0.0974 0.9990 Nonlife 112 0.6297 0.2614 0.0358 1.0000 LowRisk Life 84 0.2769 0.4582 0.0000 3.9816 Nonlife 112 0.4091 0.2021 -0.9367 0.8322 CR Life 84 0.2582 0.2806 -0.5883 0.9148 Nonlife 112 1.4997 1.0184 -3.1944 4.0983 LevR Life 84 2.8023 5.3546 -22.448 26.8034 Nonlife 112 29500000 34300000 114543 175000000 Size Life 84 47800000 66100000 169971 323000000 Nonlife 112 0.4088 0.3252 0.0370 2.5830 LR Life 84 0.4045 0.3042 0.0003 2.1016 Nonlife 112 1.2391 5.2419 0.0527 55.234 ER Life 84 0.5127 0.3570 0.1639 1.5666 Nonlife 112 0.2051 0.5870 0.0023 5.1291 INCR Life 84 0.1302 0.0659 0.0007 0.2833 Nonlife 112 0.3140 0.1721 0.0012 0.9409 RIR Life 84 0.0296 0.0620 0.0005 0.4035 Nonlife 112 0.4375 0.4983 0.0000 1.0000 GA Life 84 0.3333 0.4742 0.0000 1.0000 43 University of Ghana http://ugspace.ug.edu.gh Nonlife 112 13.0671 3.7542 8.7300 19.2500 Infla Life 84 13.0671 3.7542 8.7300 19.2500 Nonlife 112 18.1357 4.2648 11.300 22.9000 IntRate Life 84 18.1357 4.2648 11.300 22.9000 RealGDP Nonlife 112 8.0771 3.0898 3.990 14.050 Life 84 8.0771 3.0898 3.990 14.050 SOURCE: RESEARCH DATA, 2016. 4.3 CORRELATION ANALYSIS Table 4.2 shows the correlation matrix for nonlife insurance to test whether there exist any multicollinearity between the variables. The correlation was run separately for nonlife and life insurance since we believe their operation are far different. The results show that the pair of GA (group affiliation) and size is highly correlated. Group affiliation, which is the only dummy variable was dropped on the basis of multicollinearity with firm size (Wooldridge, 2003). Firm size was maintained because it was one of the main variable of interest. Again real GDP was also dropped because it was highly correlated with interest rate and inflation. From the correlation matrix, leverage ratio (LevR) and expense ratio (ER) negatively correlates with the dependent variable HighRisk whiles capital ratio (CR), Size, loss ratio (LR), investment income ratio (INCR), reinsurance ratio (RIR), inflation (Infla) and interest rate (IntRate) positively correlated with investment risk taking. 44 University of Ghana http://ugspace.ug.edu.gh Table 4.2: Correlation matrix for Nonlife Insurance HighRisk LowRisk CR LevR Size LR ER INCR RIR GA Infla IntRate Real GDP HighRisk 1.00 LowRisk -0.60 1.00 CR 0.11 -0.04 1.00 LevR -0.20 0.10 -0.11 1.00 Size 0.23 -0.23 0.20 0.05 1.00 LR 0.12 -0.01 0.01 0.05 -0.11 1.00 ER -0.06 0.11 0.13 -0.08 -0.27 0.50 1.00 INCR 0.08 -0.04 0.03 -0.03 0.08 0.03 -0.03 1.00 RIR 0.20 -0.11 0.07 -0.15 0.16 0.45 0.38 0.04 1.00 GA 0.20 -0.07 0.18 -0.03 0.69 -0.11 -0.12 0.14 0.42 1.00 Infla 0.03 0.09 -0.07 0.09 0.16 0.13 0.07 -0.05 0.07 0.00 1.00 IntRate 0.09 -0.05 0.08 0.01 -0.08 -0.12 0.08 0.04 -0.02 0.00 0.48 1.00 Real GDP -0.01 0.00 0.01 -0.08 0.07 0.04 -0.12 0.03 0.01 0.00 -0.69 -0.61 1.00 45 University of Ghana http://ugspace.ug.edu.gh Table 4.3: Correlation matrix for life Insurance HighRisk LowRisk CR LevR Size LR ER INCR RIR GA Infla IntRate Real GDP HighRisk 1.00 LowRisk -0.52 1.00 CR -0.19 0.19 1.00 LevR -0.09 -0.03 0.01 1.00 Size -0.45 0.28 0.09 0.17 1.00 LR 0.09 0.01 -0.29 -0.04 -0.08 1.00 ER 0.24 -0.24 -0.15 -0.05 -0.29 -0.14 1.00 INCR 0.51 -0.33 -0.17 0.08 -0.35 0.32 0.17 1.00 RIR 0.14 -0.17 0.23 -0.07 -0.48 -0.08 0.04 0.26 1.00 GA -0.38 0.32 0.21 0.03 0.68 -0.23 -0.04 -0.24 -0.22 1.00 Infla -0.10 0.10 -0.02 0.12 -0.17 0.10 0.04 0.16 0.11 0.00 1.00 IntRate -0.06 0.03 -0.11 0.24 0.09 0.11 -0.02 0.27 -0.02 0.00 0.48 1.00 Real GDP -0.02 0.04 0.03 -0.21 -0.08 -0.24 -0.02 -0.33 0.01 0.00 -0.69 -0.61 1.00 46 University of Ghana http://ugspace.ug.edu.gh The correlation matrix for life insurance from Table 4.3 was quite different from that of non-life. Capital, leverage, size, inflation and interest rate negatively correlated with investment risk taking. The correlation was surprisingly positive between the underwriting risk variables (expense ratio and capital ratio) and the dependent variable. The result for the life insurance is not consistent with some prior literature (Andersson et al., 2013; Ren et al., 2011; Zou et al., 2012). In measuring investment risk taking by LowRisk, representing low risk taking, size, capital ratio, investment income ratio, reinsurance ratio, interest rate and loss ratio negatively correlated with the dependent variable for non-life. Also for life insurance, leverage, expense ratio, investment income ratio and reinsurance ratio negatively correlated with the investment risk taking variable whiles the rest of the other independent variables were positively correlated. It is necessary to note that the correlation between the variables under the study is merely association and not causality. However, in other to establish that the independent variables actually impact on investment risk taking behaviour, we performed the econometric estimation for the model stated in the previous chapter. 4.4 REGRESSION ANALYSIS A panel regression model was estimated for investment risk taking behaviour. In order for a decision to be made on whether to use fixed effects or random effects model, the Hausman test (for the exogeneity of the unobserved error component) was conducted (Baltagi, 1998). The results failed to accept the hypothesis for the fixed effects model to be used (see Table A2.1, Appendix 2, for Hausman test results). Outcome of the empirical model using random effect are presented in Table 4.4. 47 University of Ghana http://ugspace.ug.edu.gh Table 4.4 Regression results for high and low investment risk taking behaviour (1) (2) (3) (4) Nonlife Life Nonlife Life VARIABLES HighRisk HighRisk LowRisk LowRisk CRi,t-1 -0.0141 -0.1018 -0.2980 0.0818* (0.1342) (0.0669) (0.2357) (0.0437) LevRi,t-1 -0.0021 -0.0004 -0.0318 -0.0003 (0.0177) (0.0037) (0.0320) (0.0027) Sizei,t-1 0.0528 -0.0672** -0.0635* 0.0354** (0.0352) (0.0262) (0.0385) (0.0179) LRi,t-1 -0.0477 -0.0687 -0.1402 0.0902** (0.0634) (0.0654) (0.0930) (0.0428) ERi,t-1 0.0008 0.0555 -0.0321 -0.1236** (0.0036) (0.0754) (0.0291) (0.0533) INCRi,t-1 0.0708*** 0.7985** -0.0736** -0.3148 (0.0261) (0.3129) (0.0316) (0.2035) RIRi,t-1 0.1168 0.438 -0.0865 -0.3485 (0.1303) (0.3764) (0.1179) (0.2433) Inflai,t -0.0027 -0.0107** 0.0094 0.0063** (0.0053) (0.0044) (0.0061) (0.0030) IntRatei,t 0.0075** -0.0019 -0.0097** -0.0051** (0.0038) (0.0034) (0.0048) (0.0026) Constant -0.7644 1.8989*** 2.0211*** -0.3045 (0.5521) (0.4735) (0.6708) (0.2943) Observations 96 72 96 72 No. of Codes 16 12 16 12 48 University of Ghana http://ugspace.ug.edu.gh R2 within 0.2373 0.1142 0.1438 0.1489 R2 between 0.0728 0.6844 0.1266 0.6398 R2 overall 0.1867 0.5292 0.1829 0.5633 Adjusted R2 0.1016 0.4609 0.0974 0.4999 Note: standard errors in parentheses, *** denotes 1% level of significance, ** denotes 5% level of significance, * denotes 10% level of significance. Firm Size and investment risk taking behaviour The amount of firm size, measured by the natural logarithm of firm’s assets had significant impact on insurer’s investment risk taking behaviour. The coefficient was found to be negative and significant on life insurance high investment risk taking and nonlife insurance low investment risk taking. This means that as the firm’s asset is increasing less funds are put into long term investment for life insurance and short term investments for non-life insurance. However, the impact was positive for the low investment risk taking by life insurance, consistent with Gorter and Bikker (2013) and Ren et al. (2011), reporting a positive relationship between size and investment risk taking. The later result of a positive impact is consistent the second hypothesis hence we accept the hypothesis that insurers with larger firm size can shoulder more investment risk. Technically as firm size increases, life insurance long term investment should increase due to the long term nature of life insurance business, but the results proved contrary. One reason could be because as firms size increases insurer’s turns to hold more of their assets in other investment portfolios which are doing much better even if in the short term. Long term investments are favoured in a more stable economic environment but in Ghana, the economy is currently highly volatile (Bekoe & Adom, 2013), therefore investors prefer to hold more short term investments rather than long term as their size gets larger. Thus this may be the case for the life insurance investments. 49 University of Ghana http://ugspace.ug.edu.gh Another explanation to the negative effect might be due to the fact that in our part of the world and for that matter a developing country, majority of the insurance companies do not have larger firm size. Evidence from NIC annual reports SIC life, Enterprise life assurance, Star Assurance Vanguard, and Glico are the largest insurance companies (NIC, 2014) with total assets of GH¢323,378,964, GH¢298,264,195, GH¢175,327,492, GH¢16,210,947, GH¢140,881,000 as compared with 5star life insurance company in US with a total asset of $268,587,888 (AM Best Rating Services, 2014). Even though new companies are sprouting up, it will take some time for their assets to grow to meet the expected demand in the mist of the existence of market imperfection. Moreover, the increase in unexpected large claims and benefits payment by insurance companies requires that insurer’s channel more of their investment activities in short term securities rather than long term investment, to meet the demand of policyholder when due. Thus, evidence of less investment in the long term as assets increases. The negative impact may also be that the insurance companies do not engage qualified investment personnel for their investment activities and are busy with marketing their insurance products to grow the assets base without considering proper management of their investment portfolios (Akotey et al., 2013). Underwriting risk and investment risk taking behaviour Underwriting risk in this study was measured by loss and expense ratios. In life insurance the level of loss ratio has positive influence on low investment risk taking, meanwhile the relationship between expense ratio and low risk investment risk taking was negative. The results are both significant at 5%. The later confirms earlier effects in literature of higher management expense resulting in less 50 University of Ghana http://ugspace.ug.edu.gh money available for investment (Andersson et al., 2013; Ren et al., 2011; Zou et al., 2012). The reason is attributed to the high cost of life insurance management activities especially in employee benefit expenses and directors emoluments. The positive impact of loss ratio with life insurance low investment risk taking is consistent with Cheng, Elyasiani and Jia (2011) for the health insurance industry. Intuitively life insurance being a long term contract, for every premium paid, there is a risk portion and an investment portion. Thus claims emanating from life insurance would have been catered for in the long term arrangement, hence we expect claims to negatively affect the long term investments risk taking and not necessarily short term investments. This may account for the positive correlation for low investment risk taking and a negative correlation with high investment risk taking (though the results for the latter is not significant). Reasonably, insurers challenged with frequent claims and benefit payment hold more short term accessible investment to cater for the immediate cost of any loss exposures (Chen et al., 2007). Another explanation of the positive impact of claim ratio (loss ratio) on investment risk taking is attributed to the fact that the life insurance companies engage qualified actuaries in the pricing of their product; providing enough funds for any financial cost of claims payment and at the same time increase investment in the short term. Therefore, as claims increases life insurer’s investment in the short term also rises. Capital and investment risk taking behaviour The regression results for low investment risk measure showed that capital ratio (proxy for capital adequacy) is positive and significant at 10%. This means that a cedi increase in life insurance capital 51 University of Ghana http://ugspace.ug.edu.gh will cause short term investment to increase by GH¢0.082, suggesting, life insurance firms having larger capital can take on more investment risk in the short term. This confirms the acceptance of the first hypothesis that insurers with greater capital have higher incentive for investment risk taking. The outcome is consistent with our expectation and similar to the studies in literature (Gorter & Bikker, 2013; Manka & Belgacem, 2015). The issue of life insurers (due to the long term nature of their business) investing more in equity securities which are generally riskier than short term securities, requires they have more capital to invest in short term securities rather than the long term. This will help to cover not only expected claims but also generally large unexpected losses. Investment income and investment risk taking behaviour Return on investment is a good measure of investment performance. However, before an insurance firm can boast of increased in returns, it requires that a certain amount of risk must be taking. In financial market conditions, the higher the risk, the higher the expected returns; as explained under the theory section in the literature review. In this study we are expecting investment income to have positive relationship with investment risk taking. Investment income ratio is the measure used for return on investment. From the results, investment income has a significant positive impact on high investment risk taking. The effect is at 1% and 5% significant level for high investment risk taking of nonlife and life insurance respectively. Implying, a one cedi increase in yearly returns on investment, increases high investment risk taking for nonlife by GH¢0.071 and life GH¢0.799. 52 University of Ghana http://ugspace.ug.edu.gh Contrarily, for nonlife low risk taking investments, the influence was negative and significant at 5% and the consequence lead to a GH¢0.074 fall in short term investments. The negative impact is attributed to frequent payment of claims by nonlife insures, which might cause the insurers to call for the amount of money invested to pay out claims, even before the investment maturity period. Macroeconomic variables and investment risk taking behaviour The two macroeconomic variables considered under the study are inflation and interest rate. Reporting for life insurance, inflation had negative effect on high investment risk taking at 5% significant level but positive for low investment risk taking at 5%. As inflation rose, life insurance long term investment fell by 0.011 and increased by 0.006 under short term investment. Higher interest rate is better for investors since it will accrue higher return on their investments. For nonlife insurance, higher interest rate lead to an increase in high investment risk taking. Meaning as the interest rate goes up, long term investments increased by GH¢0.008. Giving nonlife insurers the incentive to channel a little more of their investments into long term securities to generate higher returns. Under short term investment, interest rate had a negative significance effect on investment risk taking. As interest rate falls in the short run, insurer’s turns to increase their investments in short term securities to be able to make up with their returns on investment. Since there exist a positive association between interest rate and investment income, as interest rate falls in the short run, insurance investor need to increase their investment in the short term to balance off any cost that will arise. The results gives significant evidence that inflation and interest rate have impact on insurer’s 53 University of Ghana http://ugspace.ug.edu.gh investment risk taking. Hence, we reject hypothesis that higher interest rate and inflation have no impact on investment risk taking behaviour of insurers and accepts decision that macroeconomic factors like inflation and interest rate have impact on investment risk taking behaviour. Investment risk taking and insurance performance Investment return is a widely used measure for financial performance (Akotey et al., 2013; Chen- Ying Lee, 2014; Muhaizam, 2013). In this study investment income was used as a proxy for financial performance for both nonlife and life insurance. The results in Table 4.5, indicates that long term investments have a significantly positive impact on the performance of life insurance companies, consistent with Chen-Ying Lee (2014). This is an indication that life insurance firms investments in the long term generates higher returns because of the longer span roll over. Nonlife insurance investment risk taking, though positive had no effect on the performance of the general insurance companies. The reason is that nonlife insurance companies assets is currently declining (see Fig 4 in Appendix 1), hence the insurers are not able to shoulder more investment in the long term. Meanwhile, the low investment risk taking measure was not significant and had no impact on any of the two businesses financial performance and so the results was not shown. Table 4.5 Performance results of Insurance Business (1) (2) Nonlife Life VARIABLES INCR INCR HighRisk 0.181 0.109*** (0.294) (0.031) CRi,t-1 -0.058 -0.003 (0.310) (0.023) LevRi,t-1 -0.011 0.002 54 University of Ghana http://ugspace.ug.edu.gh (0.061) (0.001) Sizei,t-1 0.044 -0.002 (0.072) (0.007) LRi,t-1 0.133 0.060*** (0.239) (0.020) ERi,t-1 -0.004 0.006 (0.014) (0.022) RIRi,t-1 -0.040 0.175 (0.434) (0.199) Inflai,t -0.019 0.000 (0.019) (0.002) IntRatei,t 0.015 0.004** (0.017) (0.001) Constant -0.591 -0.010 (1.176) (0.128) Observations 112 84 No. of Code 16 12 R2 within 0.0218 0.3012 R2 between 0.0380 0.7298 R2 overall 0.0234 0.4684 Adjusted R2 -0.0673 0.4037 Note: standard errors in parentheses, *** denotes 1% level of significance, ** denotes 5% level of significance, * denotes 10% level of significance. Investment risk taking exposure The findings of investment risk taking exposure for all insurance companies under both life and nonlife are presented in Appendix 3. Coefficient of variation technique was used as a proxy for investment risk taking exposure and the insurance company with the highest CV was described as the company with the greatest exposure in investment risk taking. The analysis from table A3.1 (Appendix 3) showed that for long term investment risk taking (high risk) under life insurance companies, Glico life insurance had the greatest risk taking exposure, followed by Ghana union and Donewell Assurance with CV of 0.593, 0.412, 0.319 respectively. 55 University of Ghana http://ugspace.ug.edu.gh And under long term investment risk taking for nonlife insurance companies, Metropolitan insurance, followed by Quality insurance, Vanguard insurance and Enterprise insurance recorded the highest CV in the order of 1.111, 1.105, 1.093, and 1.054 respectively (refer to Appendix 3, Table A3.2). The findings for short term investment risk taking (low risk) showed that Glico insurance and Ghana union assurance records the highest CV of 1.117 and 1.038 for both life and nonlife insurance companies respectively (refer to Appendix 3, Table A3.3 and A3.4). In general, from the period of 2008-2014, the analysis revealed that Glico life recorded the greatest risk exposure under high and low investment risk taking. 56 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE SUMMARY OF FINDINGS, CONCLUSION AND RECOMMENDATIONS 5.1 INTRODUCTION This chapter presents the summary of the findings and conclusions relative to the objectives of the study. Based on the findings, recommendations have been made to the effect of firm specific and macroeconomic factors on investment risk taking behaviour. The chapter also outline the limitation of the study and suggested areas for future research. 5.2 SUMMARY OF FINDINGS The study investigates the effects of firm specific and macroeconomic factors on investment risk taking behaviour of life and nonlife insurance companies in Ghana. The investment activities consist of high or low risk taking measures as stated earlier in the methodology. The research was conducted on 16 nonlife and 12 life insurance companies, using random effect model from a panel data analysis. Firm specific and macroeconomic variables from the insurance operations were used as measures for background risk. Background risk is an uncontrollable risk emanating from the operational activities of insurance companies, for example risk emerging from insurance underwriting activities. The variables employed for the background risk were firm size, leverage, capital, claims payment, management expense, reinsurance ceded, interest rate and inflation. The theoretical review brought to the fore, views of appreciating the fact that insurance firms with large background risk have less incentive to take on more risk asset investment, either in the long or short term, depending on the nature of insurance business. Investment risk taking behaviour in this study was defined as the capacity for an insurer to take the risk of investing its asset. 57 University of Ghana http://ugspace.ug.edu.gh Relative to the first objective, the analysis pointed out that firm specific factors like firm size, return on investment (investment income), capital, management expense, interest rate and inflation have significant effect on investment risk taking behaviour. Firm size had a negative impact on life insurance high investment risk taking. For nonlife insurance, firm size had a negative impact on low investment risk taking. The implication to this is that as firm size gets larger, the insurer’s capacity to invest in the long term decreases but increases in the short term for life insurance companies. The results indicates that the insurance industries in Ghana, especially, nonlife insurers do not have larger firm size that can shoulder more long term investments and most often insurer divert their long term investments into short term investments. For both life and nonlife insurance, investment income recorded a positive relationship with high investment risk taking, which is a better indication of more returns, the higher the risk taking. The macroeconomic variable, inflation also had positive effect on high investment risk taking for nonlife insurance. This means that as inflation increases in the long term, insurance prices (premiums) increases, hence, insurers are able to adjust their premiums to get more funds for investments in the long run. From the findings, interest rate was positively related to long term investment risk taking for life insurance companies. The effect was negative on short term investments for both life and nonlife insurance firms. Thus as interest rate increases, returns on investment increases and insurers are able to get additional income to roll again on their investment portfolios. Meanwhile, under low investment risk taking measure (short term investments), capital and firm size had positive effect on the dependent variable for life insurance companies; this result is consistent with literature. 58 University of Ghana http://ugspace.ug.edu.gh Claim payment was positively related with short investment risk taking measure for life insurance companies. This implies that as life insurance claims and benefits rises, insurers at any point in time should have accessible investments to meet immediate policy obligations. Again higher management spending by life insurance companies negatively influenced insurer’s investment in the short term. The second objective aimed at determining the effect of investment risk taking on insurance companies’ performance. The study found that returns on investments (investment income) in the long term have positive significant impact on the financial performance of life insurance companies. The result, however, was not significant for nonlife insurance companies. The results shows that insurance company’s managers’ misalignment of investment activities and firm specific factors in the industries operation can be disadvantageous. Finally, Glico Life Insurance Company stood out as the insurance company with the greatest investment risk taking exposure for the period under the study. On the other hand, for nonlife insurance companies, Ghana union assurance and Metropolitan insurance were found to have the highest risk taking exposure for short and long term investment, respectively. 5.3 CONCLUSIONS Background risk which was explained by firm specific factors of insurance companies significantly affects the investment risk taking behaviour. This is evidenced by the fact that insurance companies with less background risk have more incentive to take on extra risk in investing their assets. Firm specific factor such as size, expense, claims, capital and investment income significantly affect the investment risk taking behaviour of insurance companies in Ghana. These indicates that more capitalised insurance firms have the capacity to shoulder short or long term investments. The impact 59 University of Ghana http://ugspace.ug.edu.gh of expenses and claims on investment depicts that insurance underwriting activities significantly influenced their investment risk taking behaviour. Long term investment risk taking behaviour of life insurance companies has significant positive impact on the financial performance of the insurance industry. Finally, the investment risk taking exposures of insurers addresses the risk appetite levels of the various insurance company’s investment activities, informing either risk averse or risk lover investors to decide on which insurance firm they wish to invest in. 5.4 RECOMMENDATIONS The study clearly establishes that insurance companies’ investment activities are affected by the availability of internal finance and insurance firm’s needs to be motivated to increase their asset base for more investment in the both short and long term. As insurers, through product innovation increases their asset base, the study recommends that life insurances companies should increase their investment capacity in the long term for better asset management. Insurer must also bear in mind that investors are keen about investing in companies with more appetite to take on risky investments; hence, insurance managers should have in place better risk management systems to minimize the risk from their operation to attract both local and foreign investors. Insurers should critically reduce their spending and engage qualified experts in their investment and underwriting activities. We advocate that the regulator introduce a threshold with sanctions on management expenditure to reduce the rate at which companies spend. The requirement by NIC to ensure that every insurance company engages a qualified actuary or a risk management professional is very timely and essential to the industry. 60 University of Ghana http://ugspace.ug.edu.gh Policy makers should constantly encourage a deeper partnership of asset managers with all insurance companies and emphasise the consequences of interaction firm specific factors on the investment strategies policies of insurance companies. Investment managers must also consider the influence of firm specific and macroeconomic factor as they explore and exploit new opportunities to improve their investment and business strategies. The introduction of the current risk base supervision in Ghana is timely since it will help control the unbalances in the system and allow for easy accessibility of data. 5.5 LIMITATION AND FUTURE STUDY Data on insurance companies’ investment activities in earlier years was not available. Prior to 2007 most of the insurance companies did not have detailed information on their investment activities and submission of 2015 financial statements information was too late to be captured into the study. Hence, the research was limited to only seven years. Further studies should aim at increasing the sample size for a better and consistent panel estimate. In addition, though the current regulatory system of risk base supervision is likely to cause some changes in the investment practices of insurers, this study did not consider the effect of regulatory rules on investment activities since data to these effects are not readily available and the sample study was before the enacting of the RBS. We anticipate that changes in regulation is likely to impose binding constraint on investment portfolios; and with the RBS system now in full operation, a future study can be the effect of regulation and corporate governance on the investment activities of insurance companies in Ghana. 61 University of Ghana http://ugspace.ug.edu.gh REFERENCES Achleitner, P. M., Biebel, J. H., Wichels, D. (2002). Does WTC Matter for the Investment Policy of P/C Insurance Companies? Geneva Papers on Risk And Insurance, 27(2), 275–282. Adams, M. (1996). The Reinsurance Decision in Life Insurance Firms: An Empirical Test of the Risk-Bearing Hypothesis. Accounting and Finance, 36(1), 15–30. Adams, M., & Buckle, M. (2003). The Determinants of Corporate Financial Performance in the Bermuda Insurance Market. Applied Financial Economics, 13(2), 133–143. Ahmad, K., Zulfiqar, S., Bilal, H., & Ahmad, H. (2013). Impact of Financial Leverage on Firm Investment, 7(14), 1115–1125. Aivazian, V. A., Ge, Y., & Qiu, J. (2005). The Impact of Leverage on Firm Investment: Canadian Evidence. Journal of Corporate Finance, 11(1-2), 277–291. Akintola-Bello, O. (1986). Investment Behaviour of Insurance Companies in Nigeria. Savings and Development, 10(4), 379–407. Akotey, J. O. , Sackey, F. G. , Amoah, L. , & Manso, R. F. (2013). The Financial Performance of Life Insurance Companies in Ghana. The Journal of Risk Finance, 14(3), 286–302. AM Best Rating Services. (2014). Global Insurance Credit Ratings & Information Services. Retrieved from www.AMBest .com Andersson, L. F., Lindmark, M., Adams, M., & Upreti, V. (2013). The Determinants of Investment Returns in the Fire Insurance Industry: The Case of Sweden, 1903-1939. Financial History Review, 20(1), 73–89. 62 University of Ghana http://ugspace.ug.edu.gh Antolin, P., Schich, S., &, Yermo, J. (2011). The Economic impact of protracted Low Interest Rates on Pension Funds and Insurance Companies. OECD Journal: Financial Market Trends, 2011(1), 237-256. Audretsch, D. B., & Elston, J. A. (2002). Does Firm Size Matter ? Evidence on the Impact of Liquidity Constraints on Firm Investment Behavior in Germany, 20, 1–17. Auken, Richard B. Carter, H. E. Van. (1990). Investment Practices of Life Insurance Companies. The Journal of Insurance Issues and Practices, 13(2), 24–38. Balogun, I. O. (2013). Portfolio Management : An Appraisal of Insurance Industry’s Investment Profile Under Interest Rate Deregulation in Nigeria ( 1985 – 2007 ), International Journal of Business and Social Science, 4(11), 287–292. Baltagi, B. (1998). Panel Data Methods. Handbook of Applied Economic Statistics, (409), 291–324. Baltas, I. D., Frangos, N. E., & Yannacopoulos, A. N. (2012). Optimal Investment and Reinsurance Policies in Insurance Markets under the Effect of inside Information. Applied Stochastic Models in Business and Industry, 28(6), 506–528. Baranoff, E. G., & Sager, T. W. (2002). The Relations among Asset Risk, Product Risk, and Capital in the Life Insurance Industry. Journal of Banking and Finance, 26(6), 1181–1197. Begenau, J., Piazzesi, M., & Schneider, M. (2015). Banks’ Risk Exposures. National Bureau of Economic Research. Working Paper, (No. w21334). Bekoe, W., & Adom, P. K. (2013). Macroeconomic Uncertainty and Private Investment in Ghana: An Empirical Investigation. International Journal of Economics & Financial Issues (IJEFI), 3(2), 276–293. 63 University of Ghana http://ugspace.ug.edu.gh Berdin, Elia & Helmut Gründl. The effect of a Low Interest Rate Environment on Life Insurers. No. 65. SAFE Working Paper Series, 2014. Berger, A. (1995). The Relationship between Capital and Earnings in Banking. Journal of Money, Credit & Banking, 27(2), 432. Bhagat, S., Bolton, B., & Lu, J. (2015). Size, Leverage, and Risk-taking of Financial Institutions. Journal of Banking and Finance, 59, 520–537. Bokpin, G. A. (2016). Bank Governance, Regulation and Risk-Taking in Ghana. Journal of African Business, 17(1), 52–68. Brewer, E., Carson, J. M., Elyasiani, E., Mansur, I., & Scott, W. L. (2007). Interest Rate Risk and Equity Values of Life Insurance Companies: A GARCH-M model. Journal of Risk and Insurance, 74(2), 401–423. Brockett, P. L., Cooper, W. W., Golden, L. L., & Pitaktong, U. (2013). A Neural Network Method for Obtaining an Early Warning of Insurer Insolvency. The Journal of Risk and Insurance, 61(3), 402–424. Chen, R., & Wong, K. A. (2004). The Determinants of Financial Health of Asian Insurance Companies. The Journal of Risk and Insurance, 71(3), 469–499. Chen, X., Yao, T., & Yu, T. (2007). How does Background Risks affect Investment Risk taking? Evidence from Insurers’ Corporate Bond Portfolios. mimeo, University of Iowa. Cheng, J., Elyasiani, E., & Jia, J. J. (2011). Institutional Ownership Stability and Risk Taking: Evidence from the Life-Health Insurance Industry. Journal of Risk and Insurance, 78(3), 609– 641. 64 University of Ghana http://ugspace.ug.edu.gh Chen-Ying, L. (2014). The effects of Firm Specific Factors and Macroeconomics on Profitability of Property-Liability Insurance Industry in Taiwan. Asian Economic and Financial Review, 4(5), 681–691. Chiu, M. C., & Wong, H. Y. (2014). Optimal investment for Insurers with the extended CIR Interest Rate Model. Abstract and Applied Analysis. Online available at htt://doi.org/10.1155/2014/129474 Colquitt, L., Sommer, D., & Godwin, N. (1999). Determinants of Cash holdings by Property-Liability Insurers. Journal of Risk and Insurance, 66(3), 401–415. Cummins, D. J. &, & Sommer, D. W. (1996). Capital and Risk in Property-Liability Insurance Markets. Journal of Banking and Finance, 20(6), 1069–1092. Dang, V. A. (2011). Leverage, Debt Maturity and Firm Investment: An Empirical Analysis. Journal of Business Finance and Accounting, 38(1-2), 225–258. Dell’Ariccia, G., Laeven, L., & Marquez, R. (2014). Real Interest Rates, Leverage, and Bank Risk- Taking. Journal of Economic Theory, 149(1), 65–99. Denis, D. J. (1994). Investment Opportunities and the Market Reaction to Equity Offerings. Journal of Financial and Quantitative Analysis, 29(2), 159–177. Dupor, B. (2001). Investment and Interest Rate Policy. Journal of Economic Theory, 98, 85–113. Elton, E. J., & Gruber, M. J. (1997). Modern Portfolio Theory, 1950 to date. Journal of Banking & Finance, 21(11-12), 1743–1759. Fields, L. P., Gupta, M., & Prakash, P. (2012). Risk Taking and Performance of Public Insurers: An International Comparison. Journal of Risk and Insurance, 79(4), 931–962. 65 University of Ghana http://ugspace.ug.edu.gh Franzen, D. (2010). Managing Investment Risk in Defined Benefit Pension Funds. OECD publishing, (38). available online at http://doi.org/10.1787/5kmjnr3sr2f3-en Garven, J. R., & Lamm-Tennant, J. (2003). The Demand for Reinsurance: Theory and Empirical Tests. Insurance and Risk Management, 7(3), 217–237. Gollier, C., & Pratt, J. W. (1996). Risk Vulnerability and the Tempering Effect of Background Risk. Econometrica, 64(5), 1109–1123. Online available at http://doi.org/10.2307/2171958 Gorter, J., & Bikker, J. a. (2013). Investment Risk taking by Institutional Investors. Applied Economics, 45(33), 4629–4640. Hagin, R. (1979). Modern Portfolio Theory. Modern Portfolio Theory, 11–13 & 89–91. Online available at http://doi.org/10.2143/AST.19.3.2014899 Hammond, J. D., Melander, E. R., & Shilling, N. (1976). Risk, Return, and the Capital Market: the Insurer Case. Journal of Financial and Quantitative Analysis, 11(01), 115–131. Heaton, J., & Lucas, D. (2000). Portfolio Choice in the Presence of Background Risk. The Economic Journal, 110(460), 1–26. Ho, C. L., Lai, G. C., & Lee, J. P. (2013). Organizational Structure, Board Composition, and Risk Taking in the U.S. Property Casualty Insurance Industry. Journal of Risk and Insurance, 80(1), 169–203. Huang, H. H., & Wang, C. P. (2013). Portfolio Selection and Portfolio Frontier with Background Risk. North American Journal of Economics and Finance, 26, 177–196. Hutchinson, M., Seamer, M., & Chapple, L. E. (2015). Institutional Investors, Risk/Performance and Corporate Governance. International Journal of Accounting, 50(1), 31–52. 66 University of Ghana http://ugspace.ug.edu.gh Janowicz-lomott, M. (2011). Investment activity of Non-life Insurance Companies in Poland. In MIBE International Conference. Thessaloniki, Greece, September. Jiang, C., Ma, Y., & An, Y. (2010). An analysis of Portfolio Selection with Background Risk. Journal of Banking and Finance, 34(12), 3055–3060. Kadapakkam, P. R., Kumar, P., & Riddick, L. A. (1998). The Impact of Cash flows and Firm Size on Investment: The International Evidence. Journal of Banking & Finance, 22(3), 293–320. Khan, J., & Naeemullah. (2015). The Impact of Leverage on Firm’s Investment. Research Journal of Recent Sciences, 4(5), 67–70. Kimball, M. S. (1993). Standard Risk-Aversion. Econometrica, 61(3), 589–611. Retrieved from ://A1993KZ70800005 Laeven, L., & Levine, R. (2009). Bank Governance, Regulation and Risk taking. Journal of Financial Economics, 93(2), 259–275. Lang, L., Ofek, E., & Stulz, R. M. (1996). Leverage, Investment, and Firm Growth. Journal of Financial Economics, 40(1), 3–29. Lee, S. J., Mayers, D., & Smith, C. W. (1997). Guaranty Funds and Risk-taking evidence from the Insurance Industry. Journal of Financial Economics, 44(1), 3–24. Lin, C. H., Lin, C. H., & Wang, Y. Y. (2012). The impacts of firm size on the interactions between investment, financing and hedging decisions. Journal of Statistics and Management Systems, 15(6), 663–683. Lu, P. J. (2011). Investment Risk taking of U.S. Life Insurers. ProQuest Dissertations and Theses. Retrieved from http://search.proquest.com/docview/903253208?accountid=39703 67 University of Ghana http://ugspace.ug.edu.gh Mangram, M. E. (2013). A Simplified Perspective of The Markowitz Portfolio Theory. Global Journal of Business Research, 7(1), 59–70. Manka, S., & Belgacem, A. (2015). Interactions between risk taking, capital, and reinsurance for property-liability insurance firms. Journal of Risk and Insurance. Markowitz, H. (1952). Portfolio Selection. Journal of Finance, 7(1), 77–91. Meyer, D. J., & Meyer, J. (1998). Changes in Background Risk and the Demand for Insurance. The Geneva Papers on Risk and Insurance Theory, 23(1), 29–40. Muhaizam, I. (2013). Determinants of financial performance: The case of General Takaful and Insurance Companies in Malaysia. International Review of Business Research Papers Issue, 9(6), 111–130. Muhammad, S. D., Lakhan, G. R., Zafar, S., & Noman, M. (2013). Rate of Interest and its Impact on Investment to the Extent of Pakistan. Pakistan Journal of Commerce & Social Sciences, 7(1), 91–99. Ng, T. H., Chong, L. L., & Ismail, H. (2013). Firm Size and Risk taking in Malaysia’s Insurance Industry. The Journal of Risk Finance, 14(4), 378–391. NIC. (2014). Annual Report, National Insurance Commission. Retrieved from www.nic.gh.org NIC. (2015). Annual Report ,National Insurance Commission. Retrieved from www.nic.gh.org Öner, E. (2015). The Effects of Firm-Specific Factors on the Profitability of Non-Life Insurance Companies in Turkey. International Journal of Financial Studies, 3(4), 510–529. Oscar, T. (2010). Panel data analysis fixed and random effects using Stata. Data and Statistical Services, 3(December), 1–40. 68 University of Ghana http://ugspace.ug.edu.gh Peng, X., Chen, F., & Hu, Y. (2014). Optimal Investment, Consumption and Proportional Reinsurance under Model Uncertainty. Insurance: Mathematics and Economics, 59, 222–234. Powell, L. S. L. L. S., & Sommer, D. W. (2007). Internal Versus External Capital Markets in the Insurance Industry: The Role of Reinsurance. Journal of Financial Services Research, 31(2-3), 173–188. Ren, Y., Sun, Q., Sun, Z., & Yu, T. (2011). Do Underwriting Cycles Affect Property / Casualty Insurer Investment. Journal of Insurance Regulation, 30(1), 5. Samina, Q. S. (2012). Investment Portfolio of Insurance Companies in Bangladesh : A Study on Selected Insurance Companies of Bangladesh Samina, 2(7), 37 – 47. Schich, S. (2009). Insurance Companies and the Financial Crisis. Financial Market Trends, 2009(2), 1–31. Staking, K. B., & Babbel, D. F. (1995). The Relation Between Capital Structure, Interest Rate Sensitivity, and Market Value in the Property-Liability Insurance Industry. Journal of Risk and Insurance (1986-1998), 62(4), 690. Sunder, R. K. & Monita, J. (2014). Investment practices of life insurance companies in India : the quest for a compliant portfolio. Insurance Markets and Companies: Analyses and Actuarial Computation, 5(1), 35–46. Touny, M. A., & Shusha, A. A. (2014). The Determinants of the Return of Investment : An Empirical Study of Egyptian Listed Corporations, 4(4), 127–139. Wooldridge, J. M. (2003). Introductory Econometrics: A Modern Approach. Economic Analysis, 2nd, 824. Available online at http://doi.org/10.1198/jasa.2006.s154 69 University of Ghana http://ugspace.ug.edu.gh Zhu, H., Deng, C., Yue, S., & Deng, Y. (2015). Optimal Reinsurance and Investment problem for an Insurer with counterparty Risk. Insurance: Mathematics and Economics, 61, 242–254. Zou, H., Wen, M. M., Yang, C. C., & Wang, M. (2012). Underwriting and Investment Risks in the Property-Liability Insurance Industry: Evidence prior to the 9-11 event. Review of Quantitative Finance and Accounting, 38(1), 25–46. APPENDICIES APPENDIX 1: TREND OF VARIABLES 60 50 40 30 20 10 0 2008 2009 2010 2011 2012 2013 2014 Years Premium Growth Rate-Nonlife Premium Growth Rate-Life Figure 1: Premium growth rate for life and nonlife 70 Growth rate University of Ghana http://ugspace.ug.edu.gh 100 90 80 70 60 50 40 30 20 10 0 2008 2009 2010 2011 2012 2013 2014 Years Expense Ratio Investment Loss Ratio Figure 2: Trend of firm specific variables for nonlife insurance 90 80 70 60 50 40 30 20 10 0 2008 2009 2010 2011 2012 2013 2014 Years Loss Ratio Expense Ratio Investment Figure 3: Trend of firm specific variables for life insurance 71 Percentages Percentages University of Ghana http://ugspace.ug.edu.gh 1.6E+09 1.4E+09 1.2E+09 1E+09 800000000 600000000 400000000 200000000 0 2008 2009 2010 2011 2012 2013 2014 Year Nonlife Life Figure 4: Trend of total assets for life and nonlife APPENDIX 2: HAUSMAN TEST Table A2.1 Hausman test results for Nonlife Insurance - High risk Coefficients (b) (B) ( b-B) Sqrt (diag(V_b - V_B)) Fixed Random Difference S.E. CRi,t-1 -0.0252473 -0.0141043 -0.011143 0.0424976 LevRi.t-1 0.0023662 -0.0020557 0.004422 0.0042191 LnAsseti,t-1 0.0586418 0.0527855 0.0058562 0.0252651 LRi,t-1 -0.0650404 -0.0477078 -0.0173326 0.0159324 ERi,t-1 0.0015909 0.0008451 0.0007458 0.000886 INCRi,t-1 0.0730363 0.0707891 0.0022472 0.0050798 RIRi,t-1 0.1247612 0.1168413 0.0079199 0.035801 72 Total Assets University of Ghana http://ugspace.ug.edu.gh Inflat -0.0029665 -0.00272 -0.0002464 0.002315 IntsRate 0.0074644 0.0075103 -0.0000459 0.0008434 b = consistent under Ho and Ha B = inconsistent under Ha, efficient under Ho Test: Ho: difference in coefficients not systematic chi2 (11) = 2.59 Prob. > chi2 = 0.9782 Table A2.2: Hausman Test results for Life Insurance – High risk Coefficients (b) (B) ( b-B) Sqrt (diag(V_b - V_B)) Fixed Random Difference S.E. CRi,t-1 -0.1007927 -0.1018033 0.0010106 0.0277109 LevRi.t-1 0.0005895 -0.0004105 0.0010001 0.0017684 LnAsseti,t-1 -0.0463492 -0.0671902 0.020841 0.0321141 LRi,t-1 -0.053395 -0.0686879 0.0152929 0.031276 ERi,t-1 0.032026 0.0554656 -0.0234397 0.0662043 INCRi,t-1 0.5845031 0.7984519 -0.2139488 0.2046226 RIRi,t-1 0.4973401 0.4380131 0.059327 0.2425601 Inflat -0.0087417 -0.0106803 0.0019386 0.0033301 IntsRate -0.0021602 -0.0018711 -0.0002891 0.0010635 b = consistent under Ho and Ha B = inconsistent under Ha, efficient under Ho 73 University of Ghana http://ugspace.ug.edu.gh Test: Ho: difference in coefficients not systematic chi2 (11) = 1.48 Prob. > chi2 = 0.9973 74 University of Ghana http://ugspace.ug.edu.gh APPENDIX 3: INSURANCE COMPANY WITH THE GREATEST RISK EXPOSURE Table A3.1: CV for Life companies (High risk) Long term investment risk taking Life Insurance Companies Mean Stdev CV Glico Life Insurance 0.332 0.197 0.593 Ghana Union Assurance 0.784 0.323 0.412 Donewell Life Insurance 0.604 0.193 0.319 Provident Life Insurance 0.588 0.180 0.307 Metropolitan Life Insurance 0.643 0.139 0.217 SIC Life Insurance 0.594 0.105 0.177 Vanguard Life Assurance 0.676 0.116 0.172 Star Life Assurance 0.483 0.058 0.121 Unique Life Assurance 0.898 0.098 0.109 Enterprise Life Assurance 0.649 0.057 0.087 Phoenix Life Insurance 0.880 0.074 0.085 Quality Life Insurance 0.963 0.034 0.035 Stdev = Standard deviation, CV= Coefficient of variation 75 University of Ghana http://ugspace.ug.edu.gh Table A3.2: CV for Nonlife Companies (High risk) Long term investment risk taking Nonlife Insurance Company Mean Stdev CV Metropolitan Insurance 0.017 0.019 1.111 Quality Insurance 0.433 0.479 1.105 Vanguard Insurance 0.187 0.205 1.093 Enterprise Insurance 0.296 0.312 1.054 Glico Insurance 0.077 0.067 0.867 Prime Insurance 0.225 0.187 0.830 Donewell Insurance 0.248 0.163 0.659 Equity Assurance 0.185 0.115 0.619 Phoenix Insurance 0.252 0.150 0.597 Regency Alliance Insurance 0.019 0.011 0.568 Star Assurance 0.194 0.102 0.526 Ghana Union Insurance 0.742 0.235 0.317 CDH Insurance 0.111 0.033 0.297 Provident Insurance 0.042 0.053 0.268 SIC Insurance 0.431 0.113 0.263 Unique Insurance 0.512 0.099 0.192 Stdev = Standard deviation, CV= Coefficient of variation Table A3.3: CV for Life Companies (Low risk) Short term investment risk taking Life Insurance Company Mean Stdev CV Glico Life Insurance 1.133 1.266 1.117 Quality Life Insurance 0.012 0.012 1.061 Phoenix Life Insurance 0.068 0.062 0.912 Unique Life Assurance 0.049 0.041 0.838 Ghana Union Assurance 0.880 0.269 0.306 Donewell Life Insurance 0.279 0.174 0.626 Metropolitan Life Ins. 0.139 0.050 0.359 Vanguard Assurance 0.261 0.088 0.336 Provident Life Insurance 0.191 0.055 0.291 SIC Life Insurance 0.348 0.099 0.285 Enterprise Life Assurance 0.271 0.056 0.207 Star Assurance 0.491 0.060 0.122 Stdev = Standard deviation, CV= Coefficient of variation 76 University of Ghana http://ugspace.ug.edu.gh Table A3.4: CV for Nonlife Companies (Low risk) Short term investment risktaking Nonlife Insurance Company Mean Stdev CV Ghana Union Insurance 0.222 0.231 1.038 Enterprise Insurance 0.637 0.348 0.547 Equity Assurance 0.611 0.302 0.494 Unique Insurance 0.314 0.145 0.461 SIC Insurance 0.373 0.142 0.379 Provident Insurance 0.449 0.169 0.376 Vanguard Insurance 0.661 0.194 0.293 Star Assurance 0.549 0.153 0.278 Prime Insurance 0.775 0.187 0.241 Glico Insurance 0.664 0.157 0.236 Donewell Insurance 0.591 0.120 0.203 CDH Insurance 0.634 0.097 0.153 Phoenix Insurance 0.638 0.086 0.135 Metropolitan Insurance 0.894 0.112 0.125 Quality Insurance 0.937 0.093 0.099 Regency Alliance Insurance 0.981 0.011 0.011 Stdev = Standard deviation, CV= Coefficient of variation 77