Description: There are three commonly used methods for compiling exposure and death data for use in obtaining crude mortality rates. They are lives, policies and amounts of insurance. This paper investigates the properties of estimators based on lives, amounts and robust alternatives. It is concluded that lives provide the best estimator, even when there is a moderate dependence of mortality rates on amounts of insurance. Presented at the Fourteenth Actuarial Research Conference, lowa City, lowa, September 6, 1979.

Description: The theory underlying the Moving Weighted Average graduation method is restated in the language of linear algebra which provides for an enrichment of the family of formulas. A comparison of the results of applying six members of the family to 100 simulated data sets is provided.

Description: The purpose of this note is to show how certain formulas of life contingencies can be derived almost painlessly under the uniform distribution of deaths assumption.

Description: The conventional way to analyze grouped observations on a single variable is to construct histograms or, sometimes, fit a normal distribution. We extend this latter approach by allowing for transformations to near normality.

Description: All claim reserve systems endeavor to derive estimates of the unpaid portion of claims which have been incurred over a given period of time, making use of information on numbers of claims and amounts paid on claims as of the valuation date. In this paper, a claim reserve system is described. While the system does produce the traditional type of single-point estimates of claim reserves, it also provides the user with frequency distributions of claim reserves

Description: In Volume LXXV of the Proceedings there appeared a paper entitled Risk Loads for Insurers by Feldblum. Confidence intervals for the betas in TABLE 4 [Profit Margins and Their Standard Deviations by Line of Business] of this paper can be generated using the assumption of underlying normal distributions. However, bootstrapping operational or not for interval estimates does not require the normality assumption. The method of operational bootstrapping does not use random numbers. The method for the 4-dimensional case, used to generate the interval estimates in the current paper, is a natural generalization of the bivariate case. I use operational bootstrapping to generate interval estimates for Betas shown in the TABLE 5.

Description: In this paper,we present a method which first converges a two dimensional data to a univariate one and then uses the interval censoring technique to estimate the probability of failure of a Joint-life status from a heavily censored data. Our study shows that the life annuities of joint-life status calculated assuming independence overestimates the ones evaluated from an actual sample of population.

Description: By the use of resampling and extreme value statistics we will develop a method to reduce the time and costs of testing insurance company insolvency. Most ruin models require assumptions about the surplus distribution and/or assumptions about the claims count and severity distributions. The actuary may not be comfortable in making these ruin assumptions. Instead, he or she may generate such distributions by running extensive computer simulations of corporate models, I introduce a method to reduce the number of simulations necessary' to estimate the 1 and 0.1 percentiles of the surplus distribution. This approach 'cuts off' the tail at approximately the tenth percentile of the sample distribution and generates new tails for the distribution by resampling. Averages of the new order statistics approximate the I and 0.1 percentiles.

Description: In this paper, we study a survival regression model known as Cox's proportional Hazards model. We assume that the data available are already grouped and develop a minimum distance estimation procedure that is also applicable when the data are also left or right censored. Asymptotic properties of the estimators are established and the numerical implementation of the method based on an iterated reweighted least squares algorithm is discussed.

Description: The two-parameter Pareto distribution is a commonly used model in reliability and risk modeling. Minimum variance unbiased estimates of the parameters of Pareto distribution are not known. In this paper, we propose and investigate a method of estimation of the parameters of the Pareto distribution. This method is based on minimization of the Kolmogorov-Smirnov distance between the empirical cumulative distribution function [cdf] and the cdf of the Pareto distribution.