A dynamic model of capital investment with uncertain demand and Bayesian learning
本文构建了一个动态模型，研究在位企业在不确定未来需求且通过贝叶斯学习更新对潜在进入者产品质量信念时的最优产能选择与时机，为产能投资过度或不足提供了新解释。
This paper employs a sequential learning framework to extend the static industrial organisation literature regarding optimal capacity choice and timing for incumbents and potential entrants to a dynamic environment with heterogeneous products in which uncertainty over future demand changes over time. The incumbent is uncertain about the quality of the entrant’s product and receives imperfect information signals at random points in time. Based on these signals, the incumbent updates his belief about the entrant’s product in a Bayesian fashion. Once adequately convinced about its true quality, he stops learning and makes a capacity choice according to his conviction. Existing models on capacity choice and timing in an incumbent-entrant context have not accounted for incomplete information and sequential learning in their analyses. As such, there are several novel implications that arise from the modeling framework, as well as new rationales for incumbent inertia and what this suggests in terms of over- and under-investment in capacity relative to the welfare-maximising socially optimal level.