Numerical algorithms for Panjer recursion by applying Bernstein approximation

Abstract

In actuarial science, Panjer recursion (1981) is used in insurance to compute the loss distribution of the compound risk models. When the severity distribution is continuous with density function, numerical calculation for the compound distribution by applying Panjer recursion will involve an approxi- mation of the integration. In order to simplify the numerical algorithms, we apply Bernstein approximation for the continuous severity distribution function and obtain approximated recursive equations, which are used for computing the approximated values of the compound distribution. The theoretical error bound for the approximation is also obtained. Numerical results show that our algorithm provides reliable results.

Key words： Compound risk model; Panjer recursion; Bernstein approximation; excess-of-loss reinsurance

References

1 Atkinson K. Elementary Numerical Analysis. New York: Wiley, 2003

2 Bernstein S N. Démonstration du théorème de weierstrass, fondeé sur le calcul des probabilités. Commun Soc Math Kharkow, 1912, 13: 1-2

3 Brunner H, Houwen P J van der. The Numerical Solution of Volterra Equations. Amsterdam and New York: North-Holland, 1986

4 Bühlmann H.Numerical evaluation of the compound Poisson distribution: recursion or Fast Fourier Transform? Scand Actuarial J, 1984, (2): 116-126

5 Dickson D C M. A review of Panjers recursion formula and its applications. British Actuarial J, 1995, 1(1): 107-124

6 Frachot A, Georges P, T. Roncalli T. Loss distribution approach for operational risk. Working Paper, Groupe de Recherche Operationelle, France, 2001

7 Gerhold S, Warnung R. Finding efficient recursions for risk aggregation by computer algebra. J Comput Appl Math, 2009, 223(1): 499-507

8 Grubel R, Hermesmeier R. Computation of compound distributions I: Aliasing errors and exponential tilting. ASTIN Bulletin, 1999, 29(2): 197-214

9 Grubel R, Hermesmeier R. Computation of compound distributions II: Discretization errors and Richardson extrapolation. ASTIN Bulletin, 2000, 30(2): 309-331

10 Hairer E, Nørsett S P, Wanner G. Solving Ordinary Differential Equations II. Berlin: Springer, 1993

11 Hipp C. Speedy convolution algorithms and Panjer recursions for phase-type distributions. Insurance: Mathematics and Economics, 2006, 38(1): 176-188

12 den Iseger P W, Smith M A J, Dekker R. Computing compound distributions faster! Insurance: Math Econom, 1996, 20(1): 23-34

13 Lorentz G G. Bernstein Approximation. Toronto: University of Toronto Press, 1953

14 Moscadelli M.The modelling of operational risk: experience with the analysis of the data collected by the Basel Committee. Temi di discussione del Servizio Studi, 517.2004

15 Panjer H. Recursive evaluation of a family of compound distributions. ASTIN Bulletin, 1981, 12(1): 22-26

16 Panjer H, Wang S. On the stability of recursive formulas. ASTIN Bulletin, 1993, 23(2): 227-258

17 Panjer H, Willmot G E. Insurance Loss Models. Schaumburg: Society of Actuaries, 1993

18 Peano G. Démonstration de l’intégrabilité des équations différentielles ordinaires. Math Ann, 1890, 37(2): 182-228

19 Rajagpoal L, Roy S D. Design of maximally-flat FIR filters using the Bernstein polynomial. Circuits and Systems, IEEE, 1987, 34(12): 1587-1590

20 Sancetta A, Satchell S E. The Bernstein copula and its applications to modelling and approximations of multivariate distributions. Econometric Theory, 2004, 20(3): 535-562

21 Sundt B, Jewell W S. Further results on recursive evaluation of compound distributions. ASTIN Bulletin, 1981, 12(1): 27-29

22 Yang J P, Cheng S H, Wang X Q. Bivariate recursive equation on excess-of-loss reinsurance. Acta Math Sin (Engl Ser), 2007, 23(3): 467-478

23 Yang J P, Cheng S H, Wu Q. Recursive equations for the compound distributions with severity distributions of the mixed type. Sci China, Ser A: Math, 2005, 48(5): 594-609