Forward Sourcing or Spot Trading? Optimal Commodity Procurement Policy with Demand Uncertainty Risk and Forecast Update

Abstract

We consider a commodity procurement problem where a firm satisfies a future customer demand with uncertainty risk via spot trading and forward sourcing. Although the firm can make demand forecast update and hence, remove demand uncertainty when the selling season arrives, it is still susceptible to a high emergency logistics cost at that time spot. Therefore, in this paper, the tradeoff between the mismatching cost of supply and uncertain demand (highest at the beginning of the planning horizon) and the high at-once delivery cost (highest at the ending of the planning horizon) is investigated. We develop a two-stage model and derive the optimal procurement policy for the firm. We also characterize the optimal parameters by assuming demand follows a bivariate normal distribution. Finally, extensive Monte-Carlo simulation is conducted and we quantify the value of forward contracts and the value of information update, using the crude oil data.