Stochastic Integer Decision Making in Operations and Finance

Two different stochastic decision models are developed for incorporating uncertainty and risk aversion into operational and financial decision making. First, a location decision model is developed for optimizing the quantity and placement of critical transformer spares over a network of industrial sites. Operational and financial sources of uncertainty are incorporated through the framework of a two-stage stochastic integer program. A conditional value-at-risk (CVaR) objective function captures risk aversion. Computational results show the risk averse model results in policies with lower loss as a result of the acquisition of more spares as a hedge against catastrophic scenarios. Second, an extension of the financial index tracking model of Cornuejols and Tutuncu (2007) is developed. To incorporate uncertainty and risk aversion, the distributionally robust stochastic programming (DRSP) framework developed in Delage and Ye (2010) is applied. Computational results show the DRSP index tracking model results in portfolios with improved tracking performance and resilience to market shocks.M.A.S