A RISK-SENSITIVE TWO-STAGE STOCHASTIC MILP FOR BUILDING MATERIALS CAPACITY AND PORTFOLIO PLANNING

Abstract

A risk-sensitive two-stage stochastic mixed-integer linear programming (MILP) model is developed for capacity and product-portfolio planning in the building materials industry. The model integrates first-stage investment and capacity decisions with scenario-dependent production, inventory, trade, and material-flow decisions. Expected costs are combined with Conditional Value-at-Risk (CVaR), robust optimization, operational constraints, and stress testing to account for uncertainty and risk. The results demonstrate how changes in capacity, import limits, and energy availability affect investment decisions, production planning, and import requirements. The proposed approach provides a systematic and auditable framework for supporting strategic decision-making under uncertainty.