Abstract:
A number of alternative models of risk such as: variance, mean absolute deviation (symmetric), expected downside risk, value at risk, conditional value at risk (asymmetric) have been developed to quantify and measure risk. Practitioners often use pricing models and simulation tools for describing the behavior of random parameters in financial applications. Within the simulation paradigm for a given decision such risk measures are easily computed. We consider a more challenging problem of computing 'optimum risk decisions'. We propose a framework for optimization which extends the well established stochastic programming paradigm of computing hedged decisions to that of computing the best decision in respect of a given risk measure.
Regards
Franke