The economics of capacity expansion in the mexican cement industry

An investment planning model based upon mathematical programming techniques is developed and implemented to analyze and evaluate recent capacity expansion and distribution decisions in the Mexican cement industry. The technological features of cement production and investment reveal that significant economies of scale at the plant level are present. In addition, the physical characteristics of cement imply that transport costs will have a substantial influence on delivered prices. In consequence, markets for cement will be geographically segmented, limiting demand and in turn plant size. Given plant-specific scale economies, the issue of minimum optimal scale becomes important when deciding to build a new plant or capacity addition. The combined effect of transport costs and scale economies will underlie the capacity expansion decision and, more concretely, the choice of size, location and timing of expansion. The Mexican cement industry is characterized by a small number of firms facing inelastic demand, high barriers to entry, increasing vertical downstream integration, geographic price discrimination and imperfect efforts to coordinate investment and output. A spatial competition model is used to illustrate why entry barriers and excess capacity could occur because of the preemptive motives of existing producers. The lack of competitiveness in the Mexican cement industry has been associated with market failure, symptoms of which include the presence of price regulation, suboptimal plant sizes, and inefficient capacity expansion and distribution practices. By use of the investment planning model (which is based upon simplifying assumptions), it is shown that a significant number of resources could be saved by investment planning according to our guidelines. This model, which assumes that a central planner makes all the decisions for the industry, solves for cost minimizing levels of production, investments, shipments, imports and exports, given that cement demand grows over time, and is geographically segmented. Finally, an evaluation of distribution practices for one year and a multiperiod sensitivity analysis of the model's solution to changes in key parameters is presented.