Stochastic optimization models of actuarial mathematics (Q2215270)

Authors of the paper review the conceptual optimization models of insurance activity related with the stochastic programming. In such a case, operations of an insurance company is described by a random discrete time process. For instance, dynamic model of stochastic programming can be expressed by the following relation \[ X_{t+1}=X_t+c(x)-d(X_t,\gamma)-x\xi_t,\ t\in\{0,1,2,\ldots\},\ X_0=u, \] where \(x\) is the nominal value of the insurance portfolio, \(c(x)\) is a deterministic inflows of premiums per unit time from the insurance portfolio, \(d(X_t,\gamma)\) is a dividend strategy depending on the current capital \(X_t\) and a parameter \(\gamma\), \(\xi_t\), \(t\in\{0,1,2,\ldots\}\), are independent and identically distributed observations of insurance pay-outs. The problem is to find values of parameters \(\{u,x,\gamma\}\) which maximize or minimize a suitable vector function \(\{V_i(u,x,\gamma), i=0,1,\ldots\}\). The methods for finding such optimal solutions is also described in outline.