Stochastic Dynamic Duality: Theory and Empirical Applicability
探讨随机动态生产问题中的对偶关系，发现当价格预期具有马尔可夫结构时，供给方程等无法通过间接利润函数偏导直接获得，因此对偶方法在实证中可能比原始方法更复杂。
Abstract This paper explores duality relationships for a broad class of stochastic dynamic production problems. Assuming that the decision maker maximizes the expected present value of profit, it is shown that product supply, negative factor demand, and negative quasifixed factor acquisition equations cannot be directly obtained by partial differentiation of the indirect profit function if price expectations have a Markovian structure. Consequently, empirical application of duality to many stochastic dynamic problems is quite complex and may be more difficult than a primal approach to the problem.