Page 1 of 1

### Capital evolution with five-year step

Posted: Tue Jul 04, 2017 2:51 pm
Dear all,

I am learning to build a global CGE model, but came across some questions.
The question is about the dynamics of capital stock. According to the standard perpetual inventory assumption,
Capital Stock(t+1) = NewInvestment(t)+ (1-depreciation rate)*Capital Stock(t).
However, when considering a five-year time step, shall I modify the depreciation part from “(1-depreciation rate)” to “(1-depreciation rate)^5”?

Moreover, how shall I deal with the new investment in period t? Shall I simply multiply it with 5? That is, shall I use the following equation instead to model the dynamics of capital stock when using a five-year time?
Capital Stock(t+1) = 5*NewInvestmentt+ (1-depreciation rate)^5*Capital Stock(t).

I wonder if anyone have experiences with the recursive dynamic CGE model could spare some time to help me solve them out.
Thanks for your help!

### Re: Capital evolution with five-year step

Posted: Sat Jul 22, 2017 7:27 pm
Hi
Take a look at "Calibration of Models with Multi-Year Periods" on http://www.mpsge.org/mainpage/mpsge.htm by Tom Rutherford.
Cheers
Renger