Some results on optimal stopping under phase-type distributed implementation delay (Q784790)

The optimal stopping problem of strong Markov processes under random implementation delay is studied. By random implementation delay, the fact that the payoff is not realized immediately when the process is stopped but rather after a random waiting period is meant. The distribution of this period is assumed to be phase-type. The paper is an extension and generalization of the results in [the author, Math. Methods Oper. Res. 75, No. 3, 273--286 (2012; Zbl 1266.60076)], where the implementation delay is a random variable independent of the state process and exponentially distributed. This allows using a resolvent formalism which can be applied in a general Markov setting. In this new paper, it is assumed that the implementation delay is again independent but has now a general phase-type distribution which forms a class of matrix exponential distribution that has a Markovian realization. A new analysis is used as follows. Once the state process is stopped, the exogenous continuous time Markov chain is initiated and the payoff paid out at the time of absorption of this chain depends on the value of the state process at that time. So, the aim is to find a stopping time which maximizes the expected discounted value of the payoff function containing the phase-type distributed random time variable. The main result gives a weak set of conditions under which the optimal stopping problem has a well-defined solution and an optimal stopping time exists. These conditions essentially mean that the optimal stopping region is not empty and that the payoff function cannot grow too fast even though it can be unbounded. In the next sections, the cases of Coxian distribution and the Coxian distribution with scalar diffusion dynamics are studied in more detail. Finally, the results are illustrated with two explicit examples -- a geometric Brownian motion and a square root diffusion known as the Cox-Ingersoll-Ross process.