Criteria for the Stochastic Ordering of Random Sums,with Actuarial Applications

It is shown that vectors ( S M 1 , … , S Mn ) and ( S' M'1 , …, S' M'n ) of random sums of positive random variables are stochastically ordered by upper orthant dependence, lower orthant dependence, concordance or by the supermodular ordering whenever their corresponding random numbers of terms ( M 1 , … , M n ) and ( M' 1 , … , M' n ) are themselves ordered in this fashion. Actuarial applications of these results are given to different dependence structures for the collective risk model with several classes of business.     

A note on generalized distortion risk measures

A generalized distortion risk measure is introduced as power of the mean absolute deviation power of a distorted random variable with respect to a location parameter. This class of risk measures extends both the distortion risk measure by Wang and Denneberg and the class of financial risk measures by Pedersen and Satchell, which itself contains the class of Stone. Integral representations and a stop–loss order preserving property of a special up-side risk measure are derived.     

ON THE THEORETICAL RELATIONSHIP BETWEEN SYSTEMATIC RISK AND PRICE ELASTICITY OF DEMAND

The linkage between financial valuation and the product and factor markets faced by the firm is an area that is rapidly developing in the literature. The purpose of this paper is to demonstrate the theoretical bridge between systematic risk and price elasticity of demand for a firm's output. The primary theoretical result indicates that price elasticity of demand, the certainty equivalents of the random demand parameter and variable cost, and their respective covariances with the cash flow of the market portfolio are key determinants of systematic risk and thus a firms's cost of capital.     

Market Timing And Model Uncertainty: An Exploratory Study For The Swiss Stock Market

We use statistical model selection criteria and Avramov's (2002) Bayesian model averaging approach to analyze the sample evidence of stock market predictability in the presence of model uncertainty. The empirical analysis for the Swiss stock market is based on a number of predictive variables found important in previous studies of return predictability. We find that it is difficult to discard any predictive variable as completely worthless, but that the posterior probabilities of the individual forecasting models as well as the cumulative posterior probabilities of the predictive variables are time-varying. Moreover, the estimates of the posterior probabilities are not robust to whether the predictive variables are stochastically detrended or not. The decomposition of the variance of predicted future returns into the components parameter uncertainty, model uncertainty, and the uncertainty attributed to forecast errors indicates that the respective contributions strongly depend on the time period under consideration and the initial values of the predictive variables. In contrast to AVRAMOV (2002), model uncertainty is generally not more important than parameter uncertainty. Finally, we demonstrate the implications of model uncertainty for market timing strategies. In general, our results do not indicate any reliable out-of-sample return predictability. Among the predictive variables, the dividend-price ratio exhibits the worst external validation on average. Again in contrast to AVRAMOV (2002), our analysis suggests that the out-of-sample performance of the Bayesian model averaging approach is not superior to the statistical model selection criteria. Consequently, model averaging does not seem to help improve the performance of the resulting short-term market timing strategies.     

Bounds for Actuarial Present Values Under the Fractional Independence Assumption

We consider the fractional independence (FI) survival model, studied by Willmot (1997), for which the curtate future lifetime and the fractional part of it satisfy the statistical independence assumption, called the fractional independence assumption.

The ordering of risks of the FI survival model is analyzed, and its consequences for the evaluation of actuarial present values in life insurance is discussed. Our main fractional reduction (FR) theorem states that two FI future lifetime random variables with identical distributed curtate future lifetime are stochastically ordered (stop-loss ordered) if, and only if, their fractional parts are stochastically ordered (stop-loss ordered).

The well-known properties of these stochastic orders allow to find lower and upper bounds for different types of actuarial present values, for example when the random payoff functions of the considered continuous life insurances are convex (concave), or decreasing (increasing), or convex not decreasing (concave not increasing) in the future lifetime as argument. These bounds are obtained under the assumption that some information concerning the moments of the fractional part is given. A distinction is made according to whether the fractional remaining lifetime has a fixed mean or a fixed mean and variance. In the former case, simple unique optimal bounds are obtained in case of a convex (concave) present value function.

The obtained results are illustrated at the most important life insurance quantities in a continuous random environment, which include bounds for net single premiums, net level annual premiums and prospective net reserves.     

Copula的投资组合选择模型的应用研究

结合Copula技术和GARCH模型,建立了投资组合的Copula-GARCH模型。由于该模型可以捕捉金融市场间的非线性相关性,因而可用于投资组合的风险分析中。利用这个模型,并结合Markowitz的投资组合选择模型,对我国的一支开放式基金——中信红利精选股票型证券投资基金投资组合的选择进行了优化,本文应用lingo 8.0,在收益率一定的情况下,得到了风险(VaR)最小的投资组合。     

Stochastic Dominance and Conditional Expectation—An Insurance Theoretical Approach

We show that the relation of second order stochastic dominance, which has found widespread use in models of economic behavior under uncertainty, may be described in terms of conditional expectation. If a distribution G second order stochastically dominates another distribution F, then there are random variables g and f with distributions G and F, respectively, such that g can be obtained from f by iterated conditional expectation. In terms of insurance, this shows that the less risky distribution can be obtained by a sequence of insurance contracts each one insuring against the residual risk left over from the previous contracts.     

A Model for Analyzing the Impact of Selective Lapsation on Mortality

Abstract

This paper presents a model for examining the effect of various relationships between mortality rates and lapse rates on the mortality experience of a cohort of insured lives. The approach is individual rather than the aggregate traditionally used in analyzing selective lapsation. The model assumes that insured lives are healthy at policy issue, but later may move to an impaired state from which the lapse rate is zero. Associated with each insured is an unobservable “risk level” random variable, which reflects the heterogeneity of the insured group. Individual mortality and lapse rates are functions of the risk level. A numerical illustration provides some interesting results obtained by using this model.     

Bayesian and Bühlmann credibility for phase-type distributions with a univariate risk parameter

Credibility theory is a statistical tool to calculate the premium for the next period based on past claims experience and the manual rate. Each contract is characterized by a risk parameter. A phase-type (or PH) random variable, which is defined as the time until absorption in a continuous-time Markov chain, is fully characterized by two sets of parameters from that Markov chain: the initial probability vector and transition intensity matrix. In this article, we identify an interpretable univariate risk parameter from amongst the many candidate parameters, by means of uniformization. The resulting density form is then expressed as an infinite mixture of Erlang distributions. These results are used to obtain a tractable likelihood function by a recursive formula. Then the best estimator for the next premium, i.e. the Bayesian premium, as well as its approximation by the Bühlmann credibility premium are calculated. Finally, actuarial calculations for the Bühlmann and Bayesian premiums are investigated in the context of a gamma prior, and illustrated by simulated data in a series of examples.     

Capital Allocation for a Sum of Dependent Compound Mixed Poisson Variables: A Recursive Algorithm Approach

The sum of independent compound Poisson random variables is a widely used stochastic model in many economic applications, including non-life insurance, credit and operational risk management, and environmental sciences. In this article we generalize this model by introducing dependence among Poisson frequency variables through a latent random variable in a linear fashion, which can be translated as a common underlying risk factors affecting the frequencies of individual compound Poisson variables. Despite its natural interpretation, this generalization leads to a highly complicated model with no closed-form distribution function. For this dependent compound mixed Poisson sum with an arbitrary severity distribution, we obtain the Laplace transform and further develop a new recursive algorithm to efficiently compute the probability mass function, extending the well-known Panjer recursion. Furthermore, based on this recursion, we derive another recursive scheme to determine the capital allocation associated with the Conditional Tail Expectation, a popular risk management exercise. A numerical example is presented for the illustration of our findings.     

Repeat Sales Index for Thin Markets

The repeat sales model is commonly used to construct reliable house price indices in absence of individual characteristics of the real estate. Several adaptations of the original model by Bailey et al. (J Am Stat Assoc 58:933–942, 1963) are proposed in literature. They all have in common using a dummy variable approach for measuring price indices. In order to reduce the impact of transaction price noise on the estimates of price indices, Goetzmann (J Real Estate Finance Econ 5:5–53, 1992) used a random walk with drift process for the log price levels instead of the dummy variable approach. The model that is proposed in this article can be interpreted as a generalization of the Goetzmann methodology. We replace the random walk with drift model by a structural time series model, in particular by a local linear trend model in which both the level and the drift parameter can vary over time. An additional variable—the reciprocal of the time between sales—is included in the repeat sales model to deal with the effect of the time between sales on the estimated returns. This approach is robust can be applied in thin markets where relatively few selling prices are available. Contrary to the dummy variable approach, the structural time series model enables prediction of the price level based on preceding and subsequent information, implying that even for particular time periods where no observations are available an estimate of the price level can be provided. Conditional on the variance parameters, an estimate of the price level can be obtained by applying regression in the general linear model with a prior for the price level, generated by the local linear trend model. The variance parameters can be estimated by maximum likelihood. The model is applied to several subsets of selling prices in the Netherlands. Results are compared to standard repeat sales models, including the Goetzmann model.     

Bayesian inference for generalized linear mixed models of portfolio credit risk

The aims of this paper are threefold. First, we highlight the usefulness of generalized linear mixed models (GLMMs) in the modelling of portfolio credit default risk. The GLMM-setting allows for a flexible specification of the systematic portfolio risk in terms of observed fixed effects and unobserved random effects, in order to explain the phenomena of default dependence and time-inhomogeneity in historical default data. Second, we show that computational Bayesian techniques such as the Gibbs sampler can be successfully applied to fit models with serially correlated random effects, which are special instances of state space models. Third, we provide an empirical study using Standard and Poor's data on U.S. firms. A model incorporating rating category and sector effects, and a macroeconomic proxy variable for state-of-the-economy suggests the presence of a residual, cyclical, latent component in the systematic risk.     

The Variance Gamma Process and Option Pricing

A three parameter stochastic process, termed the variance gamma process, that generalizes Brownian motion is developed as a model for the dynamics of log stock prices. The process is obtained by evaluating Brownian motion with drift at a random time given by a gamma process. The two additional parameters are the drift of the Brownian motion and the volatility of the time change. These additional parameters provide control over the skewness and kurtosis of the return distribution. Closed forms are obtained for the return density and the prices of European options. The statistical and risk neutral densities are estimated for data on the S & P500 Index and the prices of options on this Index. It is observed that the statistical density is symmetric with some kurtosis, while the risk neutral density is negatively skewed with a larger kurtosis. The additional parameters also correct for pricing biases of the Black Scholes model that is a parametric special case of the option pricing model developed here.     

The Variance Gamma Process and Option Pricing

A three parameter stochastic process, termed the variance gammaprocess, that generalizes Brownian motion is developed as amodel for the dynamics of log stock prices. Theprocess is obtainedby evaluating Brownian motion with drift at a random time givenby a gamma process. The two additional parameters are the driftof the Brownian motion and the volatility of the time change.These additional parameters provide control over the skewnessand kurtosis of the return distribution. Closed forms are obtainedfor the return density and the prices of European options.Thestatistical and risk neutral densities are estimated for dataon the S&P500 Index and the prices of options on this Index.It is observed that the statistical density is symmetric withsome kurtosis, while the risk neutral density is negativelyskewed with a larger kurtosis. The additional parameters alsocorrect for pricing biases of the Black Scholes model that isa parametric special case of the option pricing model developedhere.     

AN EXAMINATION OF THE SMALL-FIRM EFFECT ON THE BASIS OF SKEWNESS PREFERENCE

This paper tests the hypothesis that the small-firm effect can be explained on the basis of investor preference for positive skewness. Traditional stochastic dominance methodology is extended to consider portfolios including variable weights of investment in a riskless asset. Including a riskless asset provides the result that small-firm portfolios stochastically dominate all other portfolios. This result, which is derived on the basis of 19 years of monthly returns, indicates that the small-firm effect cannot be fully attributed to tax effects, benchmark error, or incorrect assumptions of the CAPM about investor risk aversion.     

Biometric Solvency Risk for Portfolios of General Life Contracts. I. The Single-Life Multiple Decrement Case

Abstract

Solvency II splits life insurance risk into seven risk classes consisting of three biometric risks (mortality risk, longevity risk, and disability/morbidity risk) and four nonbiometric risks (lapse risk, expense risk, revision risk, and catastrophe risk). The best estimate liabilities for the biometric risks are valued with biometric life tables (mortality and disability tables), while those of the nonbiometric risks require alternative valuation methods. The present study is restricted to biometric risks encountered in traditional single-life insurance contracts with multiple causes of decrement. Based on the results of quantitative impact studies, process risk was deemed to be not significant enough to warrant an explicit calculation. It was therefore assumed to be implicitly included in the systematic/parameter risk, resulting in a less complex standard formula. For the purpose of internal models and improved risk management, it appears important to capture separately or simultaneously all risk components of biometric risks. Besides its being of interest for its own sake, this leads to a better understanding of the standard approach and its application extent. Based on a total balance sheet approach we express the liability risk solvency capital of an insurance portfolio as value-at-risk and conditional value-at-risk of the prospective liability risk understood as random present value of future cash flows at a given time. The proposed approach is then applied to determine the biometric solvency capital for a portfolio of general life contracts. Using the conditional mean and variance of a portfolio’s prospective liability risk and a gamma distribution approximation we obtain simple solvency capital formulas as well as corresponding solvency capital ratios. To account for the possibility of systematic/parameter risk, we propose either to shift the biometric life tables or to apply a stochastic biometric model, which allows for random biometric rates. A numerical illustration for a cohort of immediate life annuities in arrears reveals the importance of process risk in the assessment of longevity risk solvency capital.     

Multivariate tests of financial models: A new approach

A variety of financial models are cast as nonlinear parameter restrictions on multivariate regression models, and the framework seems well suited for empirical purposes. Aside from eliminating the errors-in-the-variables problem which has plagued a number of past studies, the suggested methodology increases the precision of estimated risk premiums by as much as 76%. In addition, the approach leads naturally to a likelihood ratio test of the parameter restrictions as a test for a financial model. This testing framework has considerable power over past test statistics. With no additional variable beyond β, the substantive content of the CAPM is rejected for the period 1926–1975 with a significance level less than 0.001.     

Pricing Derivatives using the Asymptotic Expansion Approach: Credit Migration Models with Stochastic Credit Spreads

The purpose of this paper is to evaluate the asymptotic approximation formulas for the price of contingent claims with credit risk, such as credit default swaps and options on defaultable bonds, in a Markovian credit migration model. Often the generator matrix of a credit migration process is assumed to be deterministic; however, a stochastically varying generator matrix is used in this paper. To apply such a model to the valuation of options on defaultable bonds, the small disturbance asymptotic expansion approach of Kunitomo and Takahashi is used in this study.     

ON THE THEORETICAL RELATIONSHIP BETWEEN BUSINESS RISK AND SYSTEMATIC RISK

This article develops a theoretical relationship between systematic risk and business risk. It is an area that has not been substantially developed in the literature. Rubinstein (1973) and Lev (1974) both developed theoretical models of systematic risk allowing for stochastic demand. A model is derived that allows for prices, variable costs and demand to be simultaneously stochastic, utilizing the covariance of the product of random variables. The temporal stationarity of unlevered systematic risk is dependent upon the temporal stationarity of the theoretical structure derived. Insight is gained as to a potential source of the empirically observed temporal instability of levered systematic risk.