Exchange Rate Market Expectations and Central Bank Policy : The Case of the Mexican Peso-US Dollar from 2005-2009

This paper examines two approaches characterized by different tail features to extract market expectations on the Mexican peso-US dollar exchange rate. Expectations are gauged by Risk-Neutral Densities (RNDs). The methods used to estimate these densities are the Volatility Function Technique (VFT) proposed by Malz (1997) and the Generalized Extreme Value (GEV) approach suggested by Figlewski (2009). We compare these methods in the context of monetary policy announcements in Mexico and the US. We find evidence that US surprises, which are measured following Kuttner (2001), have significant effects on exchange rate variations. Around event days in Mexico and the US, the results also indicate that, although both VFT and GEV suggest similar dynamics at the center of the distribution, these two methods show significantly different patterns in the tails. Our empirical evidence shows that the GEV captures better the extreme values of the distribution around monetary policy event days given its unique procedure that allows for longer asymptotically well-behaved tails. This explains the main differences