THE ASSET-LIABILITY MANAGEMENT STRATEGY SYSTEM AT FANNIE-MAE

Fannie Mae uses the asset-liability management strategy (ALMS) system to support portfolio management decisions regarding a large, highly le veraged portfolio of home mortgage assets. The ALMS system employs a c ombination of Monte Carlo security cash-flow simulation methods and op tions-based pricing methods to estimate a holding-period return distri bution for each asset and liability considered in the portfolio optimi zation. It uses the return distributions to determine which portfolio composition maximizes the expected utility of return on portfolio equi ty. The return distribution calculations are distributed in parallel a cross a network of Unix workstations. We implemented this distributed computation fairly easily using commercially available client-server d atabase-management software, transforming the feasibility prototype mo del into an interactive-interface, multi-user, distributed-processing application for use by Fannie Mae staff.