Finite-Time Ruin for the Compound Markov Binomial Risk Model

TL;DR

This paper derives explicit formulas for finite-time ruin probabilities in a discrete-time risk model modulated by a Markov chain, extending classical results to dependent claim processes.

Contribution

It extends Takács' Ballot Theorem to the Markov-modulated setting and develops a multivariate Lagrangian inversion approach for arbitrary initial surplus.

Findings

Derived Takács-type formula for ruin probability with zero initial surplus.

Developed a Lagrangian inversion method for general initial surplus.

Provided distributional results for hitting times in the modulated risk process.

Abstract

In this paper, we study finite-time ruin probabilities for the compound Markov binomial risk model - a discrete-time model where claim sizes are modulated by a finite-state ergodic Markov chain. In the classic (non-modulated) case, the risk process has interchangeable increments and consequently, its finite-time ruin probability can be obtained in terms of Tak\'acs' famous Ballot Theorem results. Unfortunately, due to the dependency of our process on the state(s) of the modulating chain these do not necessarily extend to the modulated setting. We show that a general form of the Ballot Theorem remains valid under the stationary distribution of the modulating chain, yielding a Tak\'acs-type expression for the finite-time ruin probability which holds only when the initial surplus is equal to zero. For the case of arbitrary initial surplus, we develop an approach based on multivariate Lagrangian inversion, from which we derive distributional results for various hitting times of the risk process, including a Seal-type formula for the finite-time ruin probability.