Bayesian statistics and the monte carlo method: bayesian statistics and the monte carlo method

ABSTRACT

We discuss the application of the Bayesian statistical paradigm in conjunction with Monte Carlo methods to practical problems. We begin by describing the basic constructs of the Bayesian paradigm. We then discuss two applications. The first entails the simulation of a two-stage model of a property-casualty insurance operation. The second application simulates the operation of an insurance regime for home equity conversion mortgages (also known as reverse mortgages). In this simulation, we built separate models to (1) predict the appreciation of individual home values and (2) predict the annual mortality experience of individual insureds. A feature of this work was the simulation of the parameters of these models in order to explicitly incorporate their variability into the model. We conclude the work by considering (1) model validation issues and (2) alternate forms of scenario testing - i.e., those employing pseudorandom numbers, quasi-random numbers, or even more subjective schemes.

REFERENCES

• 1: Berger, J. 1985. Statistical Decision Theory and Bayesian Analysis, 2nd ed. Springer-Verlag, New York.
• 2: Cramer, H. 1947. Mathematical Methods of Statistics. Princeton University Press, Princeton, NJ.
• 3: DeVinti, T. R. and T. N. Herzog. 1991. Modeling Home Equity Conversion Mortgages. Transactions of the Society of Actuaries 43:261--275.
• 4: Downs, A. 1989. High Housing Prices---First Quarter of 1989. Brookings Institute, Washington, D.C.
• 5: Foster, C. and R. van Order. 1984. An Option-Based Model of Mortgage Default. Housing Finance Review 3 (4):351--372.
• 6: Herzog, T. N. 1990. Discussion of "AIDS: Exponential vs. Polynomial Growth Models" by H. H. Panjer. Insurance: Mathematics and Economics 9:291--293.
• 7: Herzog, T. N. 1999. Introduction to Credibility Theory, 3rd ed. ACTEX Publications, Winsted, CT.
• 8: Herzog, T. N. and G. Lord. 2002 (to appear) Applications of Monte Carlo Methods to Finance and Insurance. ACTEX Publications, Winsted, CT.
• 9: Herzog, T. N. and D. B. Rubin. 1983. Using Multiple Imputations to Handle Non-response in Sample Surveys. In Incomplete Data in Sample Surveys, Vol. 2, ed. W. G. Madow, I. Olkin, and D. B. Rubin, 209--245. New York, NY: Academic Press.
• 10: May, J. and E. Szymanoski. 1989. Reverse Mortgages: Risk Reduction Strategies. Secondary Mortgage Markets. 6(1):16--21.
• 11: National Association of Realtors. January, 1989. Existing and New Home Median Sales Price of Single-Family Homes. Home Sales. Washington, D.C.
• 12: Panjer, H. H. 1989. "AIDS: Exponential vs. Polynomial Growth Models". Insurance: Mathematics and Economics 8:203--209.
• 13: Wade, A. 1989. Unpublished tables of mortality rates for female lives, supplementary material to the 1989 Trustees' Report of the Social Security Administration. U.S. Department of Health and Human Services.
• 14: Waldman, M. and M. Gordon. 1988. Determining the Yield of Mortgage Securities. In The Handbook of Mortgage-Backed Securities, ed. F. Fabozzi, 257--277. Chicago, Illinois: Probus Publishing.