Stochastic programming methods in asset-liability management

This paper reviews some of the stochastic programming (SP) frameworks that are useful in applications to asset-liability management (ALM). Two such frameworks include recourse models and SP with decision rules. Recent advances also provide a representation for the Conditional Value-at-Risk risk measure that can be easily optimized in SP. Uncertainty in ALM stochastic programs is represented through discrete scenarios that are often generated through time-series methods. Sophisticated methods, such as those incorporating stable distributions, are needed to capture typical characteristics of financial data.