Home Bias in Currency Forecasts

The "home bias" phenomenon states that empirically, economic agents often under-utilize opportunities beyond their country borders, and it is well-documented in various international pricing and purchase patterns. This bias manifests in the forms of fewer exchanges of goods and net equity-holdings, as well as less arbitrage of price differences across borders than theoretically predicted to be optimal. Our paper documents another form of home bias, where market participants appear to under-weigh information beyond their borders when making currency forecasts. Using monthly data from 1995 to 2010 for seven major exchange rates relative to the US dollar, we show that excess currency returns and the errors in investors' consensus forecasts not only depend on the interest differentials between the pair of countries, but they depend more strongly on interest rates in a broader set of countries. A global short interest differential and a global long interest differential are driving the results.