RDP 2010-07: Monetary Policy and the Exchange Rate: Evaluation of VAR Models 6.Conclusion

This paper investigates the ability of vector autoregressive (VAR) models to properly identify monetary policy shocks with data simulated from a small open economy DSGE model estimated using Australian data. Overall, it finds that sign restriction models do reasonably well at estimating the responses of macroeconomic variables to monetary policy shocks, particularly compared to VAR models based on a recursive identification structure.

Using an identification procedure that is agnostic regarding the direction of the exchange rate response, the paper examines the ability of sign-restricted VAR models to overcome puzzles related to the real exchange rate. It finds that the sign restriction approach recovers the impulse responses (free of the exchange rate puzzles) reasonably well, provided that a sufficient number of shocks are uniquely identified; if only the monetary policy shocks are identified, the exchange rate puzzle remains. This suggests that identification schemes that are too parsimonious may fail to recover the ‘true’ impulse responses. The paper also finds that measures of central tendency can be misleading and that the true impulses hardly ever coincide with the median. This casts doubt on the common notion that the median impulses are ‘most probable’.

Reality is likely to be even more complex than the ideal laboratory setting created in this paper. As is pointed out, even in this simple setting, identifying too few shocks can make it difficult to recognise the qualitative and quantitative features of the true data generating process. Analogously, in real world applications, a much richer set of sign restrictions and identified shocks than our simple model uses might be required to capture the transmission mechanism. It is likely that there are many more shocks ‘contaminating’ the identification of the transmission mechanism than allowed for in our simple model. It might be worth analysing more complex small open economy models with a richer shock structure and a larger number of possible sign restrictions that would allow the movements in the terms of trade and investment to be captured, for example. Both variables have been shown to be important building blocks to overcome some of the puzzles addressed in this paper and are known to be important drivers of the business cycle for a small open economy such as Australia.