If I have DSGE model with many unknown parameters like (α, β, γ, δ, …) i.e (unknown values) how to estimate the model and get the values, on other hand, imagine you have the DSGE model equations, how to estimate the parameters (α, β, γ, δ, …), and get their values.
You can either go for calibration, i.e. some form of moment matching or do Bayesian estimation. Both are explained in detail in textbooks like Dave/Dejong.
Thank you very much for you reply
so please provide me with textbooks title for Dave/Dejong , and tell me how to do calibration (I need an example for calibration ) or book explain this step by step
It’ “Structural Macroeconometrics”. See https://www.pitt.edu/~dejong/dbook/docs/Table_of_Contents.pdf
Thank you so much
so this book explain Calibration and Bayesian estimation right , so and do you think this book enough
Yes, Bayesian estimation is explained there. Regarding calibration, I sent you some lecture materials via PM.
Hi prof Pfeifer, may I revive this topic…so it seems there is some trade-off between matching moments via calibration and doing Bayesian estimation, right?. Like if you do Bayesian estimation, you can’t match moments using calibration techniques. And it seems GMM/SMM does both. Thus, matching moments and estimating the parameters, although not using full information as in Bayesian estimation.
Or maybe first calibrate the model to match moments and then use the calibrated parameters as priors in Bayesian estimation. Or that is not necessary?