Currently, I am running a very basic dsge model with a tfp shock in Dynare. However, I find the the correlation between capital and output in the “APPROXIMATED MATRIX OF CORRELATIONS” is around 0.70, which is very high in my point of view. I am wondering is this because k(-1) vs k issue? or it is very normal ?
In unfiltered data, a correlation above 0.7 or 0.8 is normal for standard RBC models. Using hp_filter=1600 brings it down to about 0.3. But you should note that this correlation is between output and capital k in Dynare’s end of period stock notation, i.e. the capital decided by investment today and used in tomorrow’s production function.