An estimation of the dynamic factor demand model with a stochastic differential equation

本文利用随机微分方程对日本天然气工业的动态要素需求模型进行了检验和估计。动态要素需求模型通常被用来估计航空、汽车、电力等行业的生产率和效率。这一领域的研究大多建立在Translog型成本函数的基础上，以欧拉方程作为企业动态投资行为的表征。本研究试图将一般微分方程型的欧拉方程转化为随机微分型。这一变化使我们发现了企业投资行为的细微波动。日本天然气行业的每家公司每年不仅会进行大的、不稳定的投资，而且还会进行小规模的投资。这项研究揭示了两种投资在动态最优化基础上区别的简便方法。本文受到系统科学、随机分析和实物期权领域的研究者的关注。

This paper has examined and estimated a dynamic factor demand model with a stochastic differential equation on Japanese gas industry. A dynamic factor demand model is usually used to estimate productivity and efficiency on industries such that airlines, automobile, electricity and so on. All most of all researches in this field are built by a translog type cost functions, with Euler equation as a representation of dynamic investment behavior of firms. This research tried to change usual differential equation type's Euler equation to a stochastic differential equation type. This change has made us to find a detail fluctuation on the investment behavior of firms. Each firm in Japanese gas industry makes sometimes not only big-sized lumpy investment but also small-sized investment every year. This research lighted up the easy method of the difference between these two investments on the dynamic optimization root.This paper was paid attention to from the researchers of system science, stochastic analysis and real option's fields.