[经济学模型] DSGE模型讨论之五——CES functions and Dixit-Stiglitz Formulation [推广有奖]

51楼

rastila

发表于 2012-1-10 23:14:35

wangfaxian 发表于 2012-1-9 12:29
楼主，你好！！线性化DSGE模型在校准后要求求解出稳定的解，并将该解的状态空间形式用于估计。请问：求出 ...

I shall reply u in english, the computer i am using has no chinese input. i'll answer u the first question later on. But the second can be explained here.
There are more than two methods of solving linear rational expectation model, most of them are developed based on understand of linear dynamic system, such as differential equation system. The industrial standard of solving LRE is Blanchard Kahn method, and there are other prominent methods: Binder-Pesaran method, Uhlig's, Klein's, Sims', etc. The Dynare use Anderson-Moore algorithm to solve LRE, which is based on Klein's. And this method do not need to differentiate endogenous variables and jump variables, that's why you do not need to specify jump variables in Dynare. I did not study Juillard's and Uribe's. Thus I can't tell you the difference yet. But I think they won't be very different from each other.

52楼

wangfaxian 发表于 2012-1-12 09:45:01

rastila 发表于 2012-1-10 23:14
I shall reply u in english, the computer i am using has no chinese input. i'll answer u the first ...

谢谢！楼主的解答，我第二问题，不是Blanchard Kahn,klein,sims.Uhlig等人之间的区分，而是Blanchard Kahn
等人的求解方法与Juillard's 及 Uribe's等人的求解方法之间的区分！！因为他们在dynare中都会涉及到。感觉上，前者是在求解析解，而后者却是在求数值解，不知是否？不知楼主见解如何？

宏观经济的辛勤耕耘者