Asymmetric and Symmetric Volatility Models for Exchange Rates in India – The Impact of the Net Purchase of US Dollars by the Central Bank and Net Inflows by Foreign Institutional Investors

Abstract

In this study we model the monthly and the daily US, Euro Zone, UK and Australian exchange rates in India using the symmetric (sGARCH) and the asymmetric (GJR-GARCH and EGARCH) volatility models with the normal, the student t and the skewed student t error distributions. We also investigate the effect of the net US dollars (USD) purchase/ sale by the central bank, the net foreign institutional investor (FII) inflows and the one month forward spot differential on the monthly US exchange rate. Furthermore, we also test the presence of the calendar effect such as the monthly effect and the day-of-week effect on these exchange rates. We find that the models with the normal error distribution tend to fit the monthly log returns of the exchange rates better than those with the non-normal error distribution and the converse is true for the daily log returns. The calendar effects in the mean model are pronounced and Fridays on an average witness an appreciation of the Indian rupee against all the currencies we tested. Month of August has significant impact on the US exchange rate and June on the other three exchange rates. The leverage effect is not pronounced in all the exchange rates. The EGARCH models with the calendar effect dummies in the volatility model are not parsimonious. The net purchase/ sale of USD in a given month by the central bank and the one month forward spot differential do not have any significant impact on the monthly US exchange rate. But the net inflow of USD from foreign institutional investors leads to an appreciation of Indian rupee against USD. Thus the concern that capital inflows, especially the easily repatriable ones, could appreciate the Indian rupee seems to be correct but the net purchase/ sale of USD by the RBI does not seem to be abating the impact.