I am interested in constructing a specific use of the compound Poisson process. Specifically, in insurance.

Take a standard Poisson process, $N(t)$, with distribution $N(t) \sim \mathrm{Pois}(\lambda t)$.

Take $X(t)$ to be a compound Poisson process with $Y_{i}$ jumps at random times and Poisson distributed with rate $\lambda$ per hour.

Suppose $X(t)$ models the arrival of claims at an insurance office, which arrive at a rate $c$ per hour. Also, suppose we have an initial sum $\mu$.

Find the probability that the company goes bankrupt and find the value of time at which the company goes bankrupt.

We have the survival to time $t$ to be $S(t) = \mu + ct - X(T)$.

$X(t) = \sum_{i=1}^{N(t)} Y_i$.

We have the properties $\mathbb{E}[X(t)]=\mu \lambda t$ and $\mathbb{V}[X(t)] = (\sigma ^2 + \mu ^2) \lambda t $.

Suppose we have $Y_{i} \sim N(2,9)$ and rates arrive at $c=1$ per hour, with initial sum $\mu = 20$ and rate $\lambda = 1$.

Then $\mathbb{E}[X(t)] = 2 \lambda t = 2t$ and $\mathbb{V}[X(t)] = 13t$.

So we have want to find the probability of $S(t) = 0$ and the time at which this occurs.

$P(S(t)=0) = P(20+2t=X(t)) = P(\sum_{i=1}^{N(t)} Y_i = 20 +2t) $

?? At around this point I lose sense of the question. Can someone put some positive direction into it? Perhaps it is formulated incorrectly (or inappropriately!).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.