A mathematical programming approach to optimise insurance premium pricing within a data mining framework

In this paper we provide evidence of the benefits of an approach which combines data mining and mathematical programming to determining the premium to charge automobile insurance policy holders in order to arrive at an optimal portfolio. An non-linear integer programming formulation is proposed to determine optimal premiums based on the insurer's need to find a balance between profitability and market share. The non-linear integer programming approach to solving this problem is used within a data mining framework which consists of three components: classifying policy holders into homogenous risk groups and predicting the claim cost of each group using k-means clustering; determining the price sensitivity (propensity to pay) of each group using neural networks; and combining the results of the first two components to determine the optimal premium to charge. We have earlier presented the results of the first two components. In this paper we present the results of the third component. Using our approach, we have been able to increase revenue without affecting termination rates and market share.     

Equilibrium relationships between insurance and self-protection activity

When an insured understakes some costly self-protection activity that reduces the probability of loss, a competitive insurer will increase the insurance coverage, given a fixed premium per dollar of coverage, to reflect the lower insurance risk.However, an imperfectly informed insurer cannot correctly adjust the coverage; while he can observe the self-protection activity of the insured, the insurer cannot determine the cost to the insured of such activity, nor can the insurer determine the reduction in the loss probability of the insured due to the self-protection activity.This paper demonstrates in an equilibrium model that insurers may be able to use the amount of self-protection activity by an insured as a screen to indicate to the insurer what the loss probability of the insured is, thus allowing the insurer to provide correctly priced insurance to all individuals. The model points out that insurers operating in a market with moral hazard may be able to overcome the adverse incentives of insureds by selectively offering certain insurance contracts contingent upon the insured meeting certain screening requirements; in the model here, self-protection activity is the screen.     

Competitive insurance pricing with complete information,loss-averse utility and finitely many policies

In a recent paper, Ramsay and Oguledo (2012) show that in a competitive insurance market with complete information about individuals’ accident probabilities and production costs, which are proportional to the amount of insurance purchased and to the premium charged, only individuals whose accident probability is in a medium range are insurable and desire insurance. The purpose of this paper is to complement the analysis of Ramsay and Oguledo by considering production costs which are proportional to the number of policies offered by an insurer. In addition to the result of Ramsay and Oguledo we show that the group of individuals who obtain insurance is partitioned into several subgroups, where each subgroup is offered the same insurance policy. To derive this result we introduce the concept of incentive compatibility which ensures that an individual has no incentive to buy another policy. Assuming that individuals have loss-averse utility, we fully characterize the boundaries of these subgroups as the result of an undercutting process in premiums between the insurers.     

Term structure of discount rates for firms in the insurance industry

There is strong evidence in the literature for the hypothesis that interest rates and the market risk premium are not constant during the business cycle. The beta risk of firms in the insurance industry is also time-varying. The major implication of these results is that discount rates for risky cash flows are time varying and must obey a term structure similar to the term structure of interest rates. The purpose of this paper is to estimate discount rates for cash flows with different time horizons for the U.S. insurance industry and for different insurance sectors. We find that the term structure cost of capital takes on different shapes depending on the business cycle. It is therefore meaningful for insurers to evaluate risky projects by selecting a discount rate most appropriate for the nature and the time horizon of each project.     

中国利率风险溢酬:时变特征与影响因素

利率风险溢酬是长期利率的组成部分,解读它所包含的信息、寻找它的来源有着重要的经济意义。本文先使用利率仿射模型,计算出先验的中国国债利率期限溢酬,然后构建VECM模型,运用脉冲响应、方差分析等技术,分析国债利率的风险溢酬和主要宏观经济变量的动态关系,发现宏观变量对溢酬的影响在当期和滞后几期有明显差异,CPI和GDP是影响最大的两个因素,但信贷供应量和M1的作用也较大。我们同时也发现银行间市场投资者比交易所市场投资者更易受到宏观经济的影响。     

平衡指数损失函数下的信度保费

在经典的信度保费模型中,得到的信度保费估计均是考虑的是纯保费,然而在保险实务中,保险公司收取的保费不可能是纯保费,必须具有正的安全负荷.在平衡指数损失函数下,研究了多合同的信度保费模型.利用正交投影方法,得到了未来保费的信度估计.最后对估计进行了数值模拟.     

Competition among non-life insurers under solvency constraints: A game-theoretic approach

We formulate a noncooperative game to model competition for policyholders among non-life insurance companies, taking into account market premium, solvency level, market share and underwriting results. We study Nash equilibria and Stackelberg equilibria for the premium levels, and give numerical illustrations.     

Pricing general insurance in a reactive and competitive market

A simple parameterisation is introduced which represents the insurance market’s response to an insurer adopting a pricing strategy determined via optimal control theory. Claims are modelled using a lognormally distributed mean claim size rate, and the market average premium is determined via the expected value principle. If the insurer maximises its expected wealth then the resulting Bellman equation has a moving boundary in state space that determines when it is optimal to stop selling insurance. This stochastic optimisation problem is simplified by the introduction of a stopping time that prevents an insurer leaving and then re-entering the insurance market. Three finite difference schemes are used to verify the existence of a solution to the resulting Bellman equation when there is market reaction. All of the schemes use a front-fixing transformation. If the market reacts, then it is found that the optimal strategy is altered, in that premiums are raised if the strategy is of loss-leading type and lowered if it is optimal for the insurer to set a relatively high premium and sell little insurance.     

Pricing general insurance in a reactive and competitive market

A simple parameterisation is introduced which represents the insurance market’s response to an insurer adopting a pricing strategy determined via optimal control theory. Claims are modelled using a lognormally distributed mean claim size rate, and the market average premium is determined via the expected value principle. If the insurer maximises its expected wealth then the resulting Bellman equation has a moving boundary in state space that determines when it is optimal to stop selling insurance. This stochastic optimisation problem is simplified by the introduction of a stopping time that prevents an insurer leaving and then re-entering the insurance market. Three finite difference schemes are used to verify the existence of a solution to the resulting Bellman equation when there is market reaction. All of the schemes use a front-fixing transformation. If the market reacts, then it is found that the optimal strategy is altered, in that premiums are raised if the strategy is of loss-leading type and lowered if it is optimal for the insurer to set a relatively high premium and sell little insurance.     

A new method for valuing underwriting agreements for rights issues

The problem of valuing underwriting agreements for rights issues is addressed. In the recent literature the theory of put options has been used to find the premiums for these services. It has been a puzzle to date that the values obtained by this method have not been in agreement with observed premiums paid by the firms to the underwriters. In particular, one has concluded that firms have overpaid for underwriting. In this paper we use the theory of contingent claims analysis to find an equation on which the premium computation must be based. We derive an upper and a lower bound for the net premium, and we compare the values obtained by our method to the premiums computed by the traditional technique for 22 rights issues listed on the stock exchange in Oslo (Oslo Børs). It turns out that in many cases the discrepancies between the two methods can be substantial. In particular, from the upper bounds and from the risk-adjusted probabilities that the issues are successful, we infer that the traditional technique may sometimes undervalue these premiums. Thus we argue that the market for underwriting agreements is not inefficient after all.     

A risk-based premium: What does it mean for DB plan sponsors?

This paper develops a risked-based premium calculation model for the insurance provided by the Pension Benefit Guaranty Corporation (PBGC). It takes account of the pension fund’s and the plan sponsor’s investment policy and extends Chen (2011) by considering distress termination triggered by the sponsor’s underfunding. We empirically illustrate our theoretical pricing formula for the 100 biggest American DB sponsoring companies. Our result clearly casts doubt on the current practice where about 70% of the PBGC premiums charged are flat. We observe that the funding ratio and the leverage are the main risk factors in a risk-based premium calculation.     

Log-supermodularity of weight functions, ordering weighted losses, and the loading monotonicity of weighted premiums

The paper is motivated by a problem concerning the monotonicity of insurance premiums with respect to their loading parameter: the larger the parameter, the larger the insurance premium is expected to be. This property, usually called the loading monotonicity, is satisfied by premiums that appear in the literature. The increased interest in constructing new insurance premiums has raised a question as to what weight functions would produce loading-monotonic premiums. In this paper, we demonstrate a decisive role of log-supermodularity or, equivalently, of total positivity of order 2 (TP₂) in answering this question. As a consequence, we establish-at a stroke-the loading monotonicity of a number of well-known insurance premiums, and offer a host of further weight functions, and consequently of premiums, thus illustrating the power of the herein suggested methodology for constructing loading-monotonic insurance premiums.     

The utility of deductibles from the insurer's point of view

The relationship between the premiums for deductible cover and for full cover are analyzed with respect to the utility for the insurer. Theorem 1 shows that within natural bounds for the premiums neither form of coverage is generally preferred by all insurers. Theorems 2–4 indicate that in many realistic cases a deductible cover requires a higher risk loading factor by the insurer than a full insurance cover. Theorem 2(b) also contains an exception to this rule.     

Threshold control of mutual insurance with limited commitment

We study the optimal premium policy of mutual insurance when the charged premium cannot be higher than a preset rate. We provide a complete solution to the problem and use numerical simulations to illustrate how the optimal premium policy responds to changes of outside factors. The results are useful for mutual insurance firms to design premium policies and can be used to test the behavior of these firms in empirical studies.     

存款保险费率的敏感性分析及实证检验

资本充足率、存款保险费率是发达国家普遍采用的维持银行稳定的基本措施,一般来说资本充足率越高,则所需的存款保险费率越低,银行为减小成本,都希望在合理控制风险的前提下减少所缴纳的存款保险.本文使用根据期权思想建立的存款保险定价模型,推导了存款保险费率对资本充足率的敏感性系数;其次根据中国上市银行的数据进行测算,分别计算了14家银行每增加一个单位的资本充足率可降低存款保险费率的数额,并对实证的结果进行比较;最后给出相关结论.     

保险公司的最优投资策略选择

保险公司传统的投资模型只允许保险公司在保费收取与赔付之间的时滞范围内投资,即投资期间不收取保费也不允许任何赔付发生。本文研究的模型克服了传统模型的不足,投资期间可以收取保费也可以接受索赔。模型在保证保险公司实现目标收益的条件下,使得公司面临的风险最小。另外在模型中引进一个安全投资比例,即保险公司以此比例的财富用于风险投资是相对安全的。通过求解模型,得到保险公司的最优投资策略和风险最小情况下用于投资的财富的比率,并讨论了保费、索赔对投资的影响;另外还得到保险公司投资组合的有效边界,并讨论了有效边界的动态性质;最后用实际数据对保险公司如何选择安全投资比例、如何分配投资资金进行了模拟。     

A Class of Ruin Probability Mo del with Dep endent Structure

In this paper, we study a class of ruin problems, in which premiums and claims are dependent. Under the assumption that premium income is a stochastic process, we raise the model that premiums and claims are dependent, give its numerical characteristics and the ruin probability of the individual risk model in the surplus process. In addition, we promote the number of insurance policies to a Poisson process with parameter λ, using martingale methods to obtain the upper bound of the ultimate ruin probability.     

关于停止损失再保险的调节系数最大化问题

停止损失再保险作为一种再保险方式,在具有相同保费的前提下,能使保险人的期望效用最大,并能使其自留风险方差最小.另外在保费和费率相等的前提下,停止损失再保险的调节系数不可能比其他再保险方式的调节系数小.本论文在此基础上作了相应推广,讨论了在保费相等的前提下,停止损失再保险的费率满足时,其调节系数不小于其他再保险方式的调节系数.     

Investment portfolio behavior of non-life insurers: A utility analysis

Non-life insurers often claim that they are lead to reach out for extremely risky assets in the composition of their investment portfolio when their underwriting results deteriorate. In this paper we develop a model which attempts to explain the real behavior of investment portfolios of non-life insurance companies. The model is extended to include several variables over which the analysis is completed, namely the underwriting result, the premiums, the funds generating coefficient. The model is then subjected to a comparative statics analysis in order to examine the behavior of the portfolio composition in response to changes in the above mentioned variables. We show that the arguments developed by non-life insurers are at least questionable and critically dependent upon the increase and/or decrease of risk aversion measures.     

Reexamining the feasibility of diversification and transfer instruments on smoothing catastrophe risk

The present study discusses the effects of diversification and transfer of risk by global insurers on smoothing the peak of catastrophic claims. Empirical experiments indicate that the occurrence frequency of natural catastrophes (NatCat) has a serially dependent trend and that the Cox–Ingersoll–Ross square-root model for global insured losses is best fit than any other static distributions. The results are used to develop a NatCat risk insurance model that sets up a NatCat premium formula, uses the serially dependent dynamics of insured loss and establishes the cash flow of all involved parties while considering corporate income tax and no additional risk premium. The simulation results based on this model shows that fluctuation reserves, catastrophe bonds and catastrophe funds with payback schemes are feasible options for smoothing risk because they can benefit all long-term involved parties, including insurance company shareholders, the insured, bondholders, the fund and the government (i.e. taxpayers).