Forward rate unbiased hypothesis, risk premium and exchange rate expectations: estimates on Pakistan Rupee-US Dollar

In this paper, forward market unbiasedness hypothesis (FRUH) is tested and its underline assumptions of rationally formed expectations and non existence of time varying risk premium is examined empirically in case of Rupee/US$. Taking in to account the non-stationarity of the spot and forward rates series, we tested this hypothesis by two approaches. First approach relates the changes in spot rates to the forward premium. The results overwhelmingly reject the hypothesis of forward rate unbiasedness hypothesis. In fact, the estimate is significantly negative and away from 1. This confirms the existence of forward discount anomaly or forward discount puzzle for 1-month forward market. Following the recent literature, we also tested forward rate unbiasedness hypothesis using the tests of cointegration. The result suggests the presence of the cointegration relation between spot and forward rates. However, it fails the restriction that this relation is (1, -1). This therefore implies that cointegration test too fail to accept the forward rate unbiasedness hypothesis. We also found that forward rate unbiased hypothesis fails due to fact that market’s expectation regarding exchange rate movements were not rational and also due to presence of time varying risk premium.