Addressing Price Puzzle- SVAR, FAVAR and SVEC Approach(working-paper)

Price puzzle is an anomaly in empirical VAR analysis that suggest an increase in output and inflation in response to a contractionary monetary policy (rising interest rates). In order to study the effects of monetary policy on output and inflation, I use three different model setups - a SVAR with Cholseki decompsition, Factor-Augmented VAR model (with three factors) and a Vector Error Correction(VEC) model. The SVAR with Choleski decomposition as suggested in the literature shows that there is a significant price puzzle problem. Hence, we see that both output and inflation rises in response to a contractionary monetary policy. Results from FAVAR model suggests that it is a signifiacant improvement over the SVAR model as it suppresses the responses of the output and inflation and produces IRFs which are closer to the theoretical models. This result is similar to the results of Bernanke et al. (2005). Finally, I use a Vector Error Correction (VEC) model approach to model the effects of a contractionary monetary policy. I use GDP
GDP growth, trade openness, inflation, exchange rate and federal funds rate in the VEC model. The results demonstrate that imposing restrictions using a long-run cointegrating relationship can resolve the price puzzle problem. The GDP and inflation drops in response to the rising interest rates. This result is similar to Krusec (2010). However, the difference is that Krusec (2010) uses a three variable in his model while I consider the effects of exchange and trade openness along with other variables. Inclusion of trade openness and exchange is crucial as we are trying to build the response of inflation and output. In doing so, I find two long run-cointegrating relationship among the variable as opposed to Krusec (2010)’s one. And the restrictions in the model changes. However, we come to similar conclusion that the considering a long run cointegrating VAR model solves the anomaly of the price puzzle