Pulse extraction under risk and a numerical forestry application

The problem of optimal pulse harvesting of a resource under risk is discussed. The fundamental importance of stationarity in the stochastic process is investigated and the limiting optimal stopping rule is derived. A numerical example from forestry is wed lo discuss the expected present value and the optimal slopping criterion as functions of time. Finally, the probability distribution of different optimal harvesting ages is calculated. In 1987, a publication will appear closely related to this one. It will contain a purely analytical derivation of the numerical results presented in this paper. Furthermore, it will contain a more general numerical optimization model, where any first order autoregressive price process and any price-age relationship can be used. Hence, this paper is a "popular" version of the final paper. Preliminary results show that the qualitative results discussed in this paper hold in very general cases.