Optimal linear estimators of location and scale parameters using order statistics and related empirical Bayes estimation

Summary

An estimator which is a linear function of the observations and which minimises the expected square error within the class of linear estimators is called an “optimal linear” estimator. Such an estimator may also be regarded as a “linear Bayes” estimator in the spirit of Hartigan (1969). Optimal linear estimators of the unknown mean of a given data distribution have been described by various authors; corresponding “linear empirical Bayes” estimators have also been developed.

The present paper exploits the results of Lloyd (1952) to obtain optimal linear estimators based on order statistics of location or/and scale parameter (s) of a continuous univariate data distribution. Related “linear empirical Bayes” estimators which can be applied in the absence of the exact knowledge of the optimal estimators are also developed. This approach allows one to extend the results to the case of censored samples.     

Empirical Bayes credibility

Abstract

A credibility estimator is Bayes in the restricted class of linear estimators and may be viewed as a linear approximation to the (unrestricted) Bayes estimator. When the structural parameters occurring in a credibility formula are replaced by consistent estimators based on data from a collective of similar risks,we obtain an empirical credibility estimator, which is a credibility counterpart of empirical Bayes estimators. Empirical credibility estimators are proposed under various model assumptions, and sufficient conditions for asymptotic optimality are established.     

Experience Rating in Group Life Insurance

Abstract

Methods for experience rating of group life contracts are obtained as empirical Bayes or linear Bayes solutions in heterogeneity models. Each master contract is assigned a latent random quantity representing unobservable risk characteristics, which comprise mortality and possibly also age distribution and distribution of the sums insured, depending on the information available about the group. Hierarchical extensions of the set-up are discussed. An application of the theory to data from an authentic portfolio of groups revealed substantial between-group risk variations, hence experience rating could be statistically justified.     

Inference about parameters in empirical linear Bayes estimation problems

Abstract

It is shown how one may construct tests and confidence regions for the unknown structural parameters in empirical linear Bayes estimation problems. The case of the collateral units having varying “designs” (i.e. regressor and covariance matrices) may be treated under the assumption that the design variables independently follow a common statistical law. The results are of an asymptotic nature.     

Multivariate Shrinkage for Optimal Portfolio Weights

Abstract

This paper proposes a multivariate shrinkage estimator for the optimal portfolio weights. The estimated classical Markowitz weights are shrunk to the deterministic target portfolio weights. Assuming log asset returns to be i.i.d. Gaussian, explicit solutions are derived for the optimal shrinkage factors. The properties of the estimated shrinkage weights are investigated both analytically and using Monte Carlo simulations. The empirical study compares the competing portfolio selection approaches. Both simulation and empirical studies show that the proposed shrinkage estimator is robust and provides significant gains to the investor compared to benchmark procedures.     

Bayesian comparison of several continuous time models of the Australian short rate

This paper provides an empirical analysis of a range of alternative single‐factor continuous time models for the Australian short‐term interest rate. The models are nested in a general single‐factor diffusion process for the short rate, with each alternative model indexed by the level effect parameter for the volatility. The inferential approach adopted is Bayesian, with estimation of the models proceeding through a Markov chain Monte Carlo simulation scheme. Discrimination between the alternative models is based on Bayes factors. A data augmentation approach is used to improve the accuracy of the discrete time approximation of the continuous time models. An empirical investigation is conducted using weekly observations on the Australian 90 day interest rate from January 1990 to July 2000. The Bayes factors indicate that the square root diffusion model has the highest posterior probability of all models considered.     

Covariance complexity and rates of return on assets

This paper considers the estimation of the expected rate of return on a set of risky assets. The approach to estimation focuses on the covariance matrix for the returns. The structure in the covariance matrix determines shared information which is useful in estimating the mean return for each asset. An empirical Bayes estimator is developed using the covariance structure of the returns distribution. The estimator is an improvement on the maximum likelihood and Bayes–Stein estimators in terms of mean squared error. The effect of reduced estimation error on accumulated wealth is analyzed for the portfolio choice model with constant relative risk aversion utility.     

An empirical comparison of GARCH option pricing models

Recent empirical studies have shown that GARCH models can be successfully used to describe option prices. Pricing such contracts requires knowledge of the risk neutral cumulative return distribution. Since the analytical forms of these distributions are generally unknown, computationally intensive numerical schemes are required for pricing to proceed. Heston and Nandi (2000) consider a particular GARCH structure that permits analytical solutions for pricing European options and they provide empirical support for their model. The analytical tractability comes at a potential cost of realism in the underlying GARCH dynamics. In particular, their model falls in the affine family, whereas most GARCH models that have been examined fall in the non-affine family. This article takes a closer look at this model with the objective of establishing whether there is a cost to restricting focus to models in the affine family. We confirm Heston and Nandi's findings, namely that their model can explain a significant portion of the volatility smile. However, we show that a simple non affine NGARCH option model is superior in removing biases from pricing residuals for all moneyness and maturity categories especially for out-the-money contracts. The implications of this finding are examined. JEL Classification G13     

A mean/variance approach to long-term fixed-income portfolio allocation

Abstract

Long-term investments in bonds offer known returns, but with risks corresponding to defaults of the underwriters. The excess return for a risky bond is measured by the spread between the expected yield and the risk-free rate. Similarly, the risk can be expressed in the form of a default spread, measuring the difference between the yield when no default occurs and the expected yield. For zero-coupon bonds and for actual market data, the default spread is proportional to the probability of default per year. The analysis of market data shows that the yield spread scales as the square root of the default spread. This relation expresses the risk premium over the risk-free rate that the bond market offers, similarly to the risk premium for equities. With these measures for risk and return, an optimal bond allocation scheme can be built following a mean/variance utility function. Straightforward computations allow us to obtain the optimal portfolio, depending on a pre-set risk-aversion level. As for equities, the optimal portfolio is a linear combination of one risk-free bond and a risky portfolio. Using the scaling law for the default spread allows us to obtain simple expressions for the value, yield and risk of the optimal portfolio.     

Heavy tails and currency crises

In affine models of foreign exchange rate returns, the nature of cross sectional interdependence in crisis periods hinges on the tail properties of the fundamentals' distribution. If the fundamentals exhibit thin tails like the normal distribution, the dependence vanishes asymptotically; while the dependence remains in the case of heavy tailed fundamentals as in case of the Student-t distribution. The linearity of the monetary model and heavy tail distributed fundamentals are sufficient conditions for fundamentals-based repeated joint currency crises. An estimator for the extreme exchange rate interdependencies is obtained and applied to Western, Asian and Latin American currency block data.     

Credibility in Favor of Unlucky Insureds

Abstract

The classical Bühlmann credibility formula estimates the hypothetical mean of a particular insured, or risk, by a weighted average of the grand mean of the collection of risks with the sample mean of the given insured. If the insured is unfortunate enough to have had large claims in the previous policy period(s), then the estimate of future claims for that risk will also be large. In this paper we provide actuaries with a method for not overly penalizing an unlucky insured while still targeting the goal of accuracy in the estimate. We propose a credibility estimator that minimizes the expectation of a linear combination of a squared-error term and a first-derivative term. The squared-error term measures the accuracy of the estimator, while the first-derivative term constrains the estimator to be close to constant.     

A global consistency result for the two-dimensional Pareto distribution in the presence of misspecified inflation

A global consistency result for the ML estimator of a misspecified two-parameter Pareto distribution is proved. The misspecification is due to the assumption of a wrong inflation rate, which violates the i.i.d. assumption in the model. We also investigate how far away from the true parameters one finds the ML estimator of the misspecified model (asymptotically for a small misspecification r). Finally, for the case where the misspecification depends on the number of observations n, i.e., r=r _n , and where $r_{n}\stackrel{n\to \infty}{\longrightarrow}0A global consistency result for the ML estimator of a misspecified two-parameter Pareto distribution is proved. The misspecification is due to the assumption of a wrong inflation rate, which violates the i.i.d. assumption in the model. We also investigate how far away from the true parameters one finds the ML estimator of the misspecified model (asymptotically for a small misspecification r). Finally, for the case where the misspecification depends on the number of observations n, i.e., r=r _n , and where r_n? ^{n? ￥}0r_{n}\stackrel{n\to \infty}{\longrightarrow}0, we prove a central limit theorem for the ML estimator.     

Credibility Estimation of Distribution Functions with Applications to Experience Rating in General Insurance

This article presents a new credibility estimation of the probability distributions of risks under Bayes settings in a completely nonparametric framework. In contrast to the Ferguson's Bayesian nonparametric method, it does not need to specify a mathematical form of the prior distribution (such as a Dirichlet process). We then show the applications of the method in general insurance premium pricing, a procedure commonly known as experience rating, which utilizes the insured's claim experience to calculate a proper premium under a given premium principle (referred to as a risk measure). As this method estimates the probability distributions of losses, not just the means and variances, it provides a unified nonparametric framework to experience rating for arbitrary premium principles. This encompasses the advantages of the well-known Bühlmann's and Ferguson's approaches, while it overcomes their drawbacks. We first establish a linear Bayes method and prove its strong consistency in nonparametric settings that require only knowledge of the first two moments of the loss distributions considered as a stochastic process. Then an empirical Bayes method is developed for the more general situation where a portfolio of risks is observed but no knowledge is available or assumed on their loss and prior distributions, including their moments. It is shown to be asymptotically optimal. The performance of our estimates in comparison with traditional methods is also evaluated through theoretical analysis and numerical studies, which show that our approach produces premium estimates close to the optima.     

The linear growth credibility model

Abstract

The investigation of evolutionary models, i.e. models allowing the risk parameter to change in time, has been one of the main topics of research in credibility theory in the last few years. In the present paper a very special (but rather practicable) evolutionary model is defined and recursions for the credibility estimator are stated.     

On the asymptotic distribution of weighted least squares estimators

Abstract

In the ELB (Empirical Linear Bayes)-approach to credibility, the unknown structural parameters are substituted by a set of parameter estimates. The weighted least squares estimators are known to be asymptotically normally distributed when the design variables are independent and identically distributed random variables. It is demonstrated that, with probability one, the conditional asymptotic distribution, given the design, is the same as the unconditional distribution. Estimation of the asymptotic covariance matrix will also be considered.     

Value at risk linear exponent (VARLINEX) forecasts

Abstract

The VARLINEX (value at risk linear exponent) forecasting procedure is presented in this paper, which explicitly adjusts the forecasts when the loss functions of the forecaster are asymmetric. The theory of order statistics is applied to derive the VARLINEX forecasts and their corresponding confidence intervals, which are distribution-free. An empirical study based on our method is carried out for the S&P 500 returns and compared with the RiskMetrics? and the EVT method. It is found that our method can perform very well in relation to EVT and always performs much better than RiskMetrics?.     

Interval Estimation of Actuarial Risk Measures

Abstract

This article investigates performance of interval estimators of various actuarial risk measures. We consider the following risk measures: proportional hazards transform (PHT), Wang transform (WT), value-at-risk (VaR), and conditional tail expectation (CTE). Confidence intervals for these measures are constructed by applying nonparametric approaches (empirical and bootstrap), the strict parametric approach (based on the maximum likelihood estimators), and robust parametric procedures (based on trimmed means).

Using Monte Carlo simulations, we compare the average lengths and proportions of coverage (of the true measure) of the intervals under two data-generating scenarios: “clean” data and “contaminated” data. In the “clean” case, data sets are generated by the following (similar shape) parametric families: exponential, Pareto, and lognormal. Parameters of these distributions are selected so that all three families are equally risky with respect to a fixed risk measure. In the “contaminated” case, the “clean” data sets from these distributions are mixed with a small fraction of unusual observations (outliers). It is found that approximate knowledge of the underlying distribution combined with a sufficiently robust estimator (designed for that distribution) yields intervals with satisfactory performance under both scenarios.     

Sampling distribution of the relative risk aversion estimator: Theory and applications

Brown and Gibbons (1985) developed a theory of relative risk aversion estimation in terms of average market rates of return and the variance of market rates of return. However, the exact sampling distributions of the relative risk aversion estimators have not been derived. The main purpose of this paper is to derive the exact sampling distribution of an appropriate relative risk aversion estimator. First, we have derived theoretically the density of Brown and Gibbons' maximum likelihood estimator. It is shown that the centralt is not appropriate for testing the significance of estimated relative risk aversion distribution. Then we derived the minimum variance unbiased estimator by a linear transformation of the Brown and Gibbons' maximum likelihood estimator. The density function is neither a central nor a noncentralt distribution. The density function of this new distribution has been tabulated. There is an empirical example to illustrate the application of this new sampling distribution.     

Cash Flow Matching

Abstract

We propose a scenario-based optimization framework for solving the cash flow matching problem where the time horizon of the liabilities is longer than the maturities of available bonds and the interest rates are uncertain. Standard interest rate models can be used for scenario generation within this framework. The optimal portfolio is found by minimizing the cost at a specific level of shortfall risk measured by the conditional tail expectation (CTE), also known as conditional valueat-risk (CVaR) or Tail-VaR. The resulting optimization problem is still a linear program (LP) as in the classical cash flow matching approach. This framework can be employed in situations when the classical cash flow matching technique is not applicable.