Forecasting foreign exchange rates is an important but difficult process; therefore, it is important to use a superior forecasting model. The paper takes up this criterion and proposes to describe and forecast foreign exchange rates by developing an intrinsically nonlinear model with variable and dynamic speeds of adjustment. It is found that the speed of adjusting the random (or expected) to the equilibrium rate is very slow, implying that fiscal policy (statistically insignificat) and monetary policy (statistically significant) may be ineffective to induce changes in the adjustment speed. We also find that the nonlinear dynamic model improves forecasting performance, implying that nonlinearities in the sense of functional forms are exploitable for improved point forecasting of foreign exchange rates.

This article examines the dynamic and stochastic behavior of the beta coefficient (to be referred to as the currency beta) of the unbiasedness hypothesis (UH) in foreign exchange markets. We argue that the dynamics and stochastics of currency betas can be attributed to the dynamic behavior of various macroeconomic variables from different sectors of an economy, in addition to the trend variable considered in previous research. Incorporating four macroeconomic variables from the financial, real, and external sectors into the currency betas of eight currencies (developed and emerging) under a logarithmic change specification used to test the UH, we attempt to simultaneously test the behavior of currency betas in terms of nonstationarity, shifts in the mean and variance, and randomness. The vast quantity of empirical tests and results strongly suggests that the changing characteristics of currency betas are readily apparent and have important implications for the reconciliation of the controversies surrounding the legitimacy of the UH, for government exchange rate policies, and for the forecasting of future spot rates, across the developed and emerging economies under study. We also find different tales from developed and developing countries (JEL F31, F37, F47, G15).