The activities of insurers involve management of a portfolio of derivative contracts and selection of investments portfolio. For multi-product insurer, the rates of return differ from one line to another and there are different risks associated with those rates of return. Insurance companies are faced with a dilemma on selection of derivative contracts mix that optimises return on equity. This dilemma complicates decision making on the types and number of derivative contracts that will be availed to a market where regulation restrictions must be adhered to (Kahane, 1977).
There is an operational problem of determining an efficient means of measuring investment returns and capital costs. An array of capital claims exists that relate to each other as well as to the portfolio of assets (Haugen & Kroncke, 1970) . Insurers generate capital by selling insurance policies. The compounded rate of cash inflows should be more than that of cash outflows for the insurer to avoid generating capital at a cost.
General insurers generate investing funds for two principle reasons; premiums paid in advance and time lag on payment of claims. The funds generating factors differ from among lines of insurance due to variation in claims settlement lag time (Cummins & Nye, 1981). Under property lines, claims are settled relatively fast thus loss reserves are comparatively lower. Inherently settlement delays under Liability lines permits the insurance company to hold and invest premium balances for longer period. For these lines, funds generating factors are higher compared to property lines.
Where the cost of policy is negative, the company earns a positive return on the instrument. The company issues a portfolio of claims on itself consisting of a number of insurance policies. A single policy may cover a number of assets like buildings, a fleet of vehicles, professional liabilities and these policyholders’ assets may or may not be dispersed in several locations. The size and composition of this portfolio is important when formulating business strategy on selection of lines of insurable risks to underwrite. The location of these assets has inherent exposure that may be specific to certain locations for instance seismic fault lines, flood or tsunami prone areas.
The capacity of an insurer is determined by among other factors the probability of ruin, the law of large numbers and the reserve funds generated from operations (Doherty, 1980). For an insurer to reach its capacity, an additional new policy would tilt the level of risk to being unacceptable. Insurer capacity threshold should always tilt ruin probability in favour of the insurer (Doherty, 1980). Risk tolerance levels and the degree of diversification of the existing insurance operations have a bearing on the acceptance of additional derivative contracts.
The overall premium to surplus ratio is consistent with the minimum level of risk for a given rate of return on the net worth (Cummins & Nye, 1981). There is a tendency of some lines of insurance to generate more investable funds than others claim occurrence and settlement mismatch. Portfolio diversification for general insurers incorporates different lines of insurance and asset categories with a goal of maximizing the return to equity for any particular value the expected return on equity (Cummins & Nye, 1981).
The financial market in Kenya is at its developing stage and unlike in other developed financial markets, there are limited financial instruments available for hedging. The Kenyan capital market lacks the legal framework to support development and trading of hedging financial instruments like derivatives. The only recourse for primary insurers in managing their insurance risk exposure is through reinsurance and product line diversification. In developed financial markets primary insurers utilises hedging instruments where transaction costs are competitive than reinsurance (Doherty, 1997).
There are various lines of insurance and this project will focus on the General insurance (property and Casualty) industry in Kenya. There are 46 licensed insurance companies as at December 2010 with 13 lines/classes of general insurance (AKI, 2010). For this project, the medical line is omitted due to lack of comparative data occasioned by its infancy in the market. For analysis, aggregate annual industry data published by the Industry Association (Association of Kenya Insurers - AKI) for the last 13 years will be used. The choice of aggregate data is to demonstrate the magnitude of portfolio effects and its influence on the underwriting and investment strategies in the insurance industry.
The Kenyan insurance industry is regulated under Insurance Act cap.487 that is currently under review to bring a framework that fits with the advanced and functionality of the industry (IRA, 2010). In consultation with IRA, insurance companies are developing products aimed at increasing the level of insurance penetration in the country as well as diversify insurance lines portfolio.
The ranking will be based on Sharpe ratio and annual data will be utilised. Whereas monthly data is not readily available, using such data would interfere with inference of the Sharpe ratio due to possibility of serial correlation. Serial correlation may be evened out through use of Ljung-box Q statistic, but such analysis will be beyond the intended scope of this project.
Statement of the problem
The rate of insurance penetration in Kenya is comparatively low with countries like South Africa accounting for more than half of non-life premium in Africa for the year 2010 (AKI, 2010). The choice of these countries is informed by the level of financial deepening in these countries. The level of development of the financial market influences the uptake Insurance policies as risk transfer financial instruments. The 46 insurance companies in Kenya are competing to increase their market share in a market with an average growth rate of 18% in the last five years (IRA, 2010).
In the last five years, insurers with product lines skewed towards motor underwriting have been struggling to survive with some in receivership and others liquidated. IRA circulars illustrate their preference of separating the short-term (general) and long-term (life) business for composite insurers. Some composite insurers have only been generating underwriting profits under the long-term business thus relying on the profitable long-term business for a return on equity. The average low levels of return on equity in property and casualty insurance have elicited a scrutiny of the business models pursued by the leading insurers in the Kenyan market. There is scanty if any relevant research on the portfolio selection criterion that insurers rely on when selecting lines of insurance. (Kinyumu, 2011) focused on innovation processes within the insurance industry in Kenya while (Karanja, 2011) has illustrated the incidence of financial innovation on insurance company growth. Most of the statistical data is availed by IRA as a regulator and AKI as the insurers’ association. The availed data is for guideline purposes to the players in the insurance industry and for post ante data mining by interested parties. There is a research gap on how industry players consume such data or whether it is relevant at all.
There is need of a study that can empirically test if portfolio approach leads to selection of the insurance lines that yield above average underwriting profits and investment income. This project aims at investigating if such a relationship exists and if its exist, to what extent does it influence decision making on selection of insurance lines.
Objectives of the study
An insurance contract is a derivative contract where the underlying asset is the value of losses experienced by the insured. Insurers in Kenya have to diversify their insurance lines to capture more market share as they satisfy the needs of different consumer segments. Stiff competition in the market influences insurers to resort to rates undercutting to win over competition. This under-pricing of risk reduces the rates of return on product lines making insurers vulnerable to insolvency. The law of large numbers works against an insurer that is unable to earn positive returns on the insurance lines underwritten.
The industry regulator sanctions the products insurers are marketing. This makes product development cycle long and expensive resulting in few new products launches in the Kenyan market. The problem is complicated by the existence of interrelationships between the sale of policies in various line, the regulatory restrictions on the amount and type of obligations that an insurer may take (Kahane, 1977). The interests of the insurance consumers must be protected from rogue insurers that collapse leaving policyholders without cover hence the stringent regulatory regime. To ascertain growth in revenues, insurance companies are faced with the challenge of selecting a portfolio of existing product lines that will maximise return on capital employed. Repackaging Products lines is common to extend their market life cycle.
Hedging of insurance risk through the Kenyan capital market is not practical due to its undeveloped nature and lack of a legal framework. There is no trading of derivatives in the capital market despite availability of capable technological platform thus unavailability of insurance linked securities. Reinsurance provides a hedge that insurance companies utilise in reducing their exposure to mega risks portfolios they have underwritten. Its mandatory under the current regulations that 33% of the total reinsurance cessation must be local (IRA, 2010).This translates to Kenya Re having an influence on the product lines on offer.
The capacity of an insurer limits the magnitude of lines it can underwrite and thus its ability to diversify and generate optimal revenue given its capital base. The product-mix dilemma may be formulated by using tools of portfolio selection theory. The aim of this project is to evaluate the success of insurers in Kenya in adopting portfolio selection approach when choosing which insurance lines to underwrite. The measure of success will be the levels of underwriting profits and investment income with the assumption that the management intention is to maximize the rate of return on equity for any given level of risk.
The main objective is to evaluate the relationship between the insurance risk underwriting activities and the underwriting profitability and investment income.
Significance of the study
The study fills a gap in that there is few studies that have attempted to study the extent to which the ex-ante annual data provided by industry regulator has an impact on the future performance of insurance companies.
They will find this study relevant when choosing their insurers as it will illustrate the extent on professionalism in management of insurance companies.
The study would be the starting point for other researchers who want to shed more light on relevance of portfolio theory in selection of insurance product lines.
Industry service providers
As competition intensify, players in the industry will benefit from this study, as they will have an empirical understanding of the going concern status of insurers in addition to general market sentiments.
They will have statistical benchmarks to gauge their performance based on choice of their product lines and the industry performance. They will also test the practical relevance of the use of Sharpe-single-index technique in analysing the profitability of the insurance industry.
The viability of many an insurance company in Kenya lingers on after the collapse of some insurers denting the insurance industry credibility. In analysing the structure and performance of short-term insurance industry, a number of measures and approaches have been used with each having its own merits (Kahane & Nye, 1975). Some studies have looked at the competitive and structural aspects of the industry. Much of these studies have contributed towards development of the conceptual framework of analysis with limited empirical testing.
In their paper,(Haugen & Kroncke, 1970) developed a theoretical framework on the problem of optimizing the structure of assets and capital claims of an insurance company. Both sides of the statement of financial position were optimised in reference to each other using portfolio techniques with the objective of minimizing risk to the insured. With the practical limitations of such theoretical analysis (Haugen & Kroncke, 1970) illustrated how portfolio analysis certainly proved itself a valuable tool.
The possibility that the rates of profit from different activities may be correlated enforces negation of isolated decision making giving rise to risk reduction effects of diversification (Kahane, 1977). The choice of insurance product lines should be determined simultaneously with the investment portfolio due to possible correlation between underwriting and investment incomes (Kahane, 1977). Implicitly there is the assumption under the portfolio approach of the possibility of variation in volume of activities in each insurance line without changing the rate of profit or risk characteristics. Insurance regulations also restricts amount of premium to be charged thus removing the freedom of choice in the optimal insurance mix(Kahane, 1977).
In their decision making model (Cummins & Nye, 1981) examined the mean-variance efficient diversification methodology in developing efficient frontiers for property-insurance companies. Points within the efficient frontiers indicate the optimal product and investment mix for each target rate of return. There is tendency of some insurance lines to generate more investible funds comparatively due existence of lags between claim occurrence and settlement. (Cummins & Nye, 1981) argued that the overall premium to surplus ratio, the distribution among insurance lines and the proportion of assets in each major investment class is consistent with the risk aversion for a given rate of return of net worth.
Property –liability companies have limitation on the acceptable risk levels for each additional new policy/risk. This capacity has largely been determined by three factors namely: the probability of ruin, the law of large numbers and reserve funds generated from operations (Doherty, 1980). Capacity is reached when the ruin probability is unacceptable by company managers though the level of underwriting risk attached to a new policy and degree of diversification of existing insurance operations also matters (Doherty, 1980).
In their evaluation on whether financial portfolio theory could be applied, (Cardozo & Smith, 1983) found that risk and return measurement of product- market investments demonstrate high positive covariance. Therefore can be used in consistent constrained optimisation approach similar to that developed in modern portfolio theory. Modified financial portfolio theory offer promise as a tool for design and management of product portfolios (Cardozo & Smith, 1983).
Portfolio problems arises in nearly all facets of decision making such as companies choosing shares and bond portfolios to invest in. (Mcentire, 1984) introduced the concept of generalized harmonic mean emphasising that the expected return of an asset included in an optimal portfolio exceed the expected value of any asset not chosen for the optimal portfolio. Insurance companies should thus invest their pooled cash inflows in assets with the largest mean values.
The insurance and reinsurance industries are constrained with resources capable of covering major catastrophes. The imbalance in supply and demand bearing in mind the elevated levels of risk and reward potential , has prompted participants in the insurance industry to tap the huge capacity that capital market has to offer (Canter, Cole, & Sandor, 1996). This capacity can be tapped through non-traditional forms of capital such hedge funds, commodity funds and pension funds.
The exposure to single event or multiple major event within a short duration, has triggered the need to insurance companies to look for additional sources of funds to finance or spread the risk (American Institute of Actuaries, 1999). The mega catastrophes that can impair the capital of the insurance industry can have minimal impact if spread through the capital markets. These mega catastrophe risks are viewed and uncorrelated with other financial risks (American Institute of Actuaries, 1999).
In the multifactor world of financial securities , the mantra is that different risk factors are associated with their own risk premiums and no single investment strategy can span the entire risk factor space (Agarwal & Naik, 2003). Therefore, different investment strategies need to deployed to earn the risk premia associated with different risk factors. The various lines of insurance have differing risk exposures thus dictating the preference in portfolio selection. Insurance companies fund their investment using premium collection and the time lag between premium collection and settlement of claim limits the investment duration. Shorter time lags may translate to shorter investment maturity period hence probable lower returns and limited investment opportunities.
On the supply side, the insurers may not offer as many insurance lines because of significant long-term insolvency cost even if the underwriting returns are positive and insurers are risk neutral (Lin, 2005). Under the multi-period formulation, the demand and supply curves may differ substantially from those under single period formulation. It is imperative that insurers make optimal choices of the insurance lines that will offer short-term returns bearing in mind the demand side preferences. The cover duration of most lines of property and casualty insurance is twelve months hence the short-term outlook.
The information asymmetry is more pronounced in insurance markets than in financial market. The severity and frequency of insurance losses is influenced by insured behaviour impeding pareto efficiency making the risk transfer mechanism inefficient (Lin & Lu, 2007). To achieve pareto-optimal resource allocation private information need to be absent given that its costly to monitor. In designing appropriate insurance contract, a menu of multiple price-quantity policies is provided or assembling a multi-period contract in accordance with insured’s underwriting experiences (Lin & Lu, 2007). The intention is to induce the insured to reveal their risk type. Segmenting insured by risk types assist in designing insurance contracts that factor in behavioural tendencies.
At times regulation limits the choice of insurers in determination of pricing of some risks. For instance , in USA crop insurers are price takers as the premium rates and underwriting guidelines are set by the Risk Management Agency (RMA)which is an agency established by Federal Crop Insurance Corporation (FCIC) (Vedenov, Miranda, Dismukes, & Glauber, 2006). Crop insurers assume large potential risk exposure without recourse to raising premium rates or declining to cover high-risk individuals for them to participate under the federal program. The federal government assumes most of the risks while less risky business can be placed in funds where the insurers pays more for underwriting losses and keeps more of underwriting gains. How well an insurer classifies their risks and manages their portfolio determines the underwriting returns (Vedenov et al., 2006).
Underwriting cycle is a business cycle operating in the insurance industry. It is characterised by premium rates falling or stabilising during soft markets and premium rates rising, competition intensity declining , underwriting profits increasing during hard markets (Elango, 2009). The cause of the underwriting cycles is multifaceted and complex elements being the driver with evidenced impact on insurers’ profitability. Insurance lines variety influences the overall performance and insurers with more variety improving their odds of outperforming insurers with less variety (Elango, 2009). The expectations are the larger the premium income the higher the performance relative to small insurers.