Why Is It so Difficult to Forecast Exchange Rate Movements?

1034 WordsJun 18, 20185 Pages
Introduction The stability of currency values plays a significant role for economic and financial stability. It is not difficult to see the exchange rate fluctuations are widely regarded as damaging. As the movements of the exchange rate have significant and large effects on the trade balance, resource allocation, domestic prices, interest rate, national income and other key economic variables. Then can exchange rate movements be predicted by these fundamental economic variables? Economists have long taken the view that economic fundamentals determine exchange rates. Nevertheless, in the early 1970s, after the collapse of fixed exchange rate regimes of the Bretton Woods system, excess volatility, nonlinear and disorderly movements in…show more content…
The third assumption is static exchange rate expectations. Both the Monetary Model and the M-F model have ignored the effect of expectations, which is the heart of current exploration of "the disconnect puzzle" in exchange rates. It seems obvious that the expectations from the market are hardly to keep unchanged, especially when policies change or shock emerge. The last assumption is imperfect capital mobility. The Mundell-Fleming Model puts high concentration on the capital flows and ignores the stock level of capital. Moreover, the ad hoc setting of money demand function simply ignores the microeconomic point of view. Literature evidence As the famous study of Meese and Rogoff summarized, none of these models, which are based on economic fundamentals can provide a better prediction than a random walk model, a simple predictor. There are reasons why economic fundamentals are not very helpful in predicting exchange rates, even if currency values are likely determined by these fundamentals. Primarily, the inherent limitations of the fundamentals models. These structural models usually rely on coefficients in the equations to forecast which are based on historical data that may vary over time (Meese and Rogoff 1983). In the research of Meese and Rogoff, they provided the model with
Open Document