Risk Management and Insurance Review
Bingzheng Chen,Sharon Tennyson,Maoqi Wang,Haizhen Zhou
2014-01-01
Abstract:The theory of adverse selection predicts that high-risk individuals are more likely to buy insurance than low-risk individuals if asymmetric information regarding individuals’ risk type is present in the market. The theory of advantageous selection predicts the opposite—a negative relationship between insurance coverage and risk type can be obtained when hidden knowledge in other dimensions (e.g., the degree of risk aversion) is present in addition to the risk type. Using the heterogeneity of insurance buyers in either risk type or risk aversion, we first introduce a classroom-based insurance market simulation game to show that adverse selection and advantageous selection can coexist. We then explain the underlying concepts using two methods: a mathematical framework based on expected utility theory and an empirical framework based on the results of the game itself. The game is easy to implement, reinforces textbook concepts by providing students a hands-on experience, and supplements current textbooks by bringing their content up to date with current research. INTRODUCTION Theoretical and empirical research in the field of asymmetric information has made significant progress in the past few decades. In an influential paper, Rothschild and Stiglitz (1976) (RS hereafter) show that in the presence of asymmetric information between the insurance buyers and sellers, high (low) risk buyers will purchase more (less) coverage. Their model of adverse selection predicts a positive correlation between insurance coverage and the ex post realization of a buyer’s risk, measured by the number of claims, total payouts, etc. In the RS model, there may be no equilibrium in competitive insurance markets. Empirical tests of their model find a positive correlation in some insurance markets such as health insurance (Cardon and Hendel, 2001) and annuities (Finkelstein and Poterba, 2004), and a negative correlation in other markets such as life insurance (Cawley and Philipson, 1999), long-term care (Finkelstein and McGarry, 2006), reverse Leon Chen is Associate Professor of Finance, Minnesota State University Mankato, phone: 507389-5336; e-mail: yilin.chen@mnsu.edu. Puneet Jaiprakash is Associate Professor of Finance, Minnesota State University Mankato, phone: 507-389-1826; e-mail: puneet.jaiprakash@mnsu.edu. 133 134 RISK MANAGEMENT AND INSURANCE REVIEW mortgage (Davidoff and Welke, 2007), Medigap (Fang et al., 2008), and commercial fire insurance (Wang et al., 2009).1 Subsequent extensions of the RS model have provided explanations to reconcile the original RS prediction and the mixed empirical evidence. These extensions can generally be grouped in two categories—the first category examines the heterogeneity of insurance buyers in one dimension (other than risk type), whereas the second category studies the heterogeneity in two dimensions.2 In the first category, the single dimensions of insurance buyers studied include the degree of risk aversion (de Meza and Webb, 2001), patience (Sonnenholzner and Wambach, 2009), and overconfidence (Huang et al., 2010). In general, these one-dimensional studies assume that risk type or hidden knowledge is linked with a certain unobservable dimension or hidden action; thus, equilibria can exist when some individuals with certain characteristics (such as high degree of risk aversion, high degree of patience, and low degree of confidence) purchase more insurance and spend more effort on mitigating risk, and therefore become low-risk types. In such cases, we will observe a negative relationship between insurance coverage and risk type, which is opposite to the RS prediction. Further, this negative relationship may be large enough to more than offset the positive relation predicted by the RS model. It benefits the insurers and is therefore aptly named advantageous or propitious selection. In the second category, the two dimensions of insurance buyers examined include risk type and risk aversion (Smart, 2000), transaction costs and risk aversion or wealth (Liu and Browne, 2007), wealth and risk aversion arising from endogenous choice of labor supply/savings (Netzer and Scheuer, 2010), and risk type and regret preferences (Huang et al., 2015). These two-dimensional studies provide a rich supplement to the RS model by finding different types of new equilibria under asymmetric information.3 Unfortunately, the field of insurance education has lagged behind. Popular textbooks such as Reija and McNamara (2013), Dorfman and Cather (2013), and Vaughan and Vaughan (2014) introduce the concept of adverse selection but do not discuss the 1 For brevity, we do not cite all of the papers that have tested the adverse selection model. See Cohen and Siegelman (2010) for an excellent and comprehensive review of these empirical studies. 2 We thank an anonymous reviewer for providing this valuable comment. 3 Smart (2000) separates the insurance buyers into four types based on high/low accident risk and high/low risk aversion, and the important finding is that “different risk classes may be pooled at a single contract in equilibrium; thus, cross-subsidization may occur.” Liu and Browne (2007) consider both transaction costs and the heterogeneity in insurance buyers’ risk aversion, and find “a separating equilibrium in which neither high-risk nor low-risk individuals are penalized due to information asymmetry.” Netzer and Scheuer (2010) consider the endogenous wealth accumulation (by supplying more labor and saving more) and the level of risk aversion, and show that equilibria in their model exhibit a nonmonotone relation between risk and coverage, thus altering the nature of equilibrium under the RS model. Huang et al. (2015) examine both risk type and insurance buyers’ regret preferences, and show some interesting natures of new equilibria in contrast to the RS model. They find that there can exist separating equilibria that predict both positive and negative correlation between insurance coverage and risk type, depending on the insurance loading, and whether low-risk types or high-risk types are regret averse. AN INSURANCE MARKET SIMULATION 135 conflicting empirical evidence and factors that might offset the positive coverage–risk correlation. Recognizing this fact, Eisenhauer (2004) provides a numerical example and graphical illustration to explain advantageous selection. In a valuable contribution to the field of insurance education, Eckles and Halek (2007) develop a simulation game that illustrates the concept and consequences of adverse selection. The simulation requires students to buy (fictitious) insurance policies from other students under different scenarios (symmetric and asymmetric information); it reinforces textbook concepts by providing students an opportunity to experience the workings of an insurance market under both ideal and nonideal conditions. In this article, we first redesign the simulation of Eckles and Halek (2007) to allow for the possibility of both adverse selection and advantageous selection. For simplicity purpose, we consider only the one-dimensional heterogeneity in either risk type or risk aversion.4 Our simulation has three periods, with each period representing a particular scenario. Period I represents the scenario of symmetric information between buyers and sellers; though probably unrealistic, it provides a benchmark for subsequent periods. In Period II, information is asymmetric, thereby allowing for the possibility of adverse selection. In both Periods I and II, buyers differ only in their risk types (i.e., probability of potential losses). In Period III, buyers differ across two dimensions— risk type and risk aversion—and at the same time information is asymmetric. Thus, it allows for both adverse selection and advantageous selection in that period. Next, we utilize the results from the game to illustrate the theoretical predictions for each of these scenarios. This helps students not only see the effect of each phenomenon, but also understand the extent to which advantageous selection may offset adverse selection. The next section describes the details of our redesigned game. We then discuss the results from the game, which was played in two sections of our Insurance and Risk Management class, and explain how they can be used to illustrate and extend the discussion in existing textbooks. The last section concludes. THE SIMULATION GAME We begin by separating students into insurance buyers (insureds) and insurance sellers (insurers). To keep the record-keeping manageable for the instructor, and to ensure that each student participates, each insurer consists of a group of three students and each insured consists of a team of two students, for a class of around 35 students. Other instructors may form groups for buyers and sellers differently based on the class size. In addition, to simulate conditions of perfect competition, we ensure that the supply is close to the demand. Also, we try to balance the numbers of insureds of different risk types (i.e., the number of high-risk buyers in the market is close to the number of lowrisk buyers in each round). Finally, for simplicity we do not allow partial loss or policies 4 This not only makes the simulation manageable in the classroom, but also achieves the purpose of introducing the concept of advantageous selection to students. Furthermore, adding risk aversion as a single factor in detecting advantageous selection makes intuitive sense, since it directly simulates the insurance market under the model of de Meza and Webb (2001), the pioneer theoretical research on advantageous selection and probably the most cited RS extension work so far. Risk aversion is also included as a dimension of buyer heterogeneity in many studies related to RS extensions. 136 RISK MANAGEMENT AND INSURANCE REVIEW with partial coverage. The game is played in multiple rounds, and in each round we have 4 insurers who can each sell no more than three policies, and