EHL(2000), "Optimal monetary policy with

[hardmann]

Dear Tom;

When I study this example and its code, I found the paper does not explain how to use data of macroeconomic, and its rats code declared these series but dose not use data instruction to read these from external file. Comparing to it,I read again code of Watson (1993), "Measures of Fit for Calibrated Models", which use follows instructions to read external data.
open data watson_jpe.rat
cal(q) 1948
data(format=rats) 1948:1 1988:4 y c invst h
"
I want to know whether target series of DSGE is true series ,which need read from external file, or simulation without true series like "stoch_simul" instruction of Dynare .

[TomDoan]

Many DSGE's aren't designed to be fit to actual data. This is one of them. Aside from the fact that the disturbance processes aren't rich enough dynamically to fit real-world data (one is an AR(1), the other two are white noise), there are also only three of them when there appear to be four natural observables (r, output gap, wages and prices).