Two-Stage Mixed Integer Stochastic Programming and Its Application to Bond Portfolio Optimization
Alreshidi, Nasser Aedh
MetadataShow full item record
We consider a two-stage stochastic bond portfolio optimization problem, where an investor aims to optimize the cost of bond portfolio under different scenarios while ensuring predefined liabilities during a given planning horizon. The investor needs to optimally decide whether to buy, hold, or sell bonds based upon present market conditions under different scenarios and varying assumptions, where the scenarios are determined based on interest rates and buying prices of the bonds. Three stochastic integer programming models are proposed and applied to real-data from Saudi Sukuk (Bond) Market. The case-study results demonstrate the varying optimal decisions made to manage bond portfolio over the two stages. In addition, the three stochastic programming models for bond portfolio optimization, are tested on a large set of randomly generated instances similar to the Saudi Sukuk (Bond) Market. The results of computational experiments attest the efficiency of the proposed models.