×

Ruin probabilities and investment under interest force in the presence of regularly varying tails. (English) Zbl 1091.62102

The model under consideration is a natural extension of the classical Cramér-Lundberg risk model: it is assumed that an insurance company which has the wealth \(Y(t)\) at time \(t\) may invest an amount \(K(t)\) in a stock or market index described by a geometric Brownian motion and invest the remaining the reserve \(Y(t)-K(t)\) in a bond with a positive interest \(r>0\). The authors studied the behaviour of the ruin probability \(\psi(x,K)\), which depends on the initial capital \(x\) and a strategy of unvestment \(K=K(t)\). In particular, they focused on the optimal strategy \(K^*\) which minimizes \(\psi(x,K)\) over all predictable strategies. The paper consists of three parts.
In the first part, the authors derived the Hamilton-Jacobi-Bellman equation for the survival probability for the optimal strategy and obtained the asymptotic behaviour of the optimal ruin probability in the presence of positive interest force and claims with tails of regular variation. These results extend previous authors’ results [see Insur. Math. Econ. 30, No. 2, 211–217 (2002; Zbl 1055.91049)] with zero interest and results of C. Klüppelberg and U. Stadtmüller [Scand. Actuarial J. 1998, No. 1, 49–58 (1998; Zbl 1022.60083)] without investment possibility.
The second part presents an existence theorem for integro-differential equations for the survival probability of an insurer who invests a constant fraction \(k\) of his wealth \(Y(t)\) in a risky stock and the remaining wealth \(Y(t)-kY(t)\) in a bond. This theorem is rather general and extends a result by G. Wang and R. Wu [Stochastic Processes Appl. 95, No. 2, 329–341 (2001; Zbl 1064.91051)]. Finally, in the third part, the authors derived the asymptotic behaviour of the ruin probability of an insurer, introduced in the second part, in the presence of claims with regularly varying tails.

MSC:

62P05 Applications of statistics to actuarial sciences and financial mathematics
91B30 Risk theory, insurance (MSC2010)
90C39 Dynamic programming
49N90 Applications of optimal control and differential games
Full Text: DOI

References:

[1] DOI: 10.1214/aoap/1028903531 · Zbl 0942.60034 · doi:10.1214/aoap/1028903531
[2] Bingham NH, Regular variation, Cambridge University Press (1987)
[3] Burton TA, Volterra integral and differential equations, Academic Press (1983)
[4] Davies B, Integral transforms and their applications 41 (2002)
[5] Embrechts P, Modelling extremal events, Springer (1997) · doi:10.1007/978-3-642-33483-2
[6] Embrechts P, Insurance: Mathematics and Economics pp 55– (1982)
[7] DOI: 10.1007/s007800100057 · Zbl 1002.91037 · doi:10.1007/s007800100057
[8] DOI: 10.1016/S0167-6687(02)00101-4 · Zbl 1055.91049 · doi:10.1016/S0167-6687(02)00101-4
[9] DOI: 10.1016/S0167-6687(00)00049-4 · Zbl 1007.91025 · doi:10.1016/S0167-6687(00)00049-4
[10] DOI: 10.1016/S0304-4149(01)00148-X · Zbl 1058.60095 · doi:10.1016/S0304-4149(01)00148-X
[11] Klüppelberg C, Scand. Act. J. pp 49– (1998)
[12] Korn R, Option pricing and portfolio optimization – Modern methods of financial mathematics, AMS (2001) · Zbl 0965.91020
[13] Lundberg F, Approximerad framställning av sannolikhetsfunktionen, Akad. Afhandling (1903)
[14] DOI: 10.1016/S0304-4149(00)00083-1 · Zbl 1047.60040 · doi:10.1016/S0304-4149(00)00083-1
[15] DOI: 10.1016/S0304-4149(98)00012-X · Zbl 0932.60044 · doi:10.1016/S0304-4149(98)00012-X
[16] DOI: 10.1214/aoap/1037125862 · Zbl 1019.60041 · doi:10.1214/aoap/1037125862
[17] DOI: 10.2307/1427849 · Zbl 0892.90046 · doi:10.2307/1427849
[18] DOI: 10.1214/aoap/1031863173 · Zbl 1021.60061 · doi:10.1214/aoap/1031863173
[19] DOI: 10.1016/S0304-4149(01)00102-8 · Zbl 1064.91051 · doi:10.1016/S0304-4149(01)00102-8
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.