Analysing the forward premium anomaly using a Logistic Smooth Transition Regression model.

Several researchers have suggested that exchange rates may be characterized by nonlinear behaviour. This paper examines these nonlinearities and asymetries and estimates a Logistic Transition Regression (LSTR) of Fama Regression with the Risk Adjusted Forward Premia as transition variable. Results confirm the existence of nonlinear dynamics in the relationship between spot exchange rate differential and the forward premium for all the currencies of the sample and for all maturities (three and six-month maturities). Results confirm the insight into the presence of speculation barriers and transaction costs in the foreign exchange rate market that would explain, at least partially, the forward premium anomaly.