Research Discussion Paper – RDP 8603 Risk Premia, Market Efficiency and the Exchange Rate: Some Evidence Since the Float
May 1986
Abstract
The Australian dollar was floated in December 1983. Since that time the exchange rate has become more volatile and has depreciated significantly. The aim of this paper is to examine the behaviour of the foreign exchange market since the float. In particular, the paper considers whether the joint hypotheses underlying the notion of speculative efficiency – namely market efficiency and risk neutrality – hold in the post-float forward market. Under the speculative efficiency hypothesis, the forward exchange rate is a rational expectation of the future spot exchange rate. The paper examines this speculative efficiency hypothesis for forward rates of different maturities, by examining whether the forward rate provides the best available forecast of the future spot rate.
The paper endeavours to make full use of the data available by sampling more finely than the contract interval. This procedure, however, involves some econometric difficulties. In particular, the residuals from OLS estimation will be serially correlated, following a low order moving average process. Consequently, the estimated standard errors will be inconsistent. To overcome this, the relevant equations are estimated by first obtaining consistent parameter estimates by OLS and then estimating a consistent asymptotic covariance matrix.
The findings of the paper can be summarised as follows. For the post-float period as a whole, the speculative efficiency hypothesis can be rejected for the 30-day forward market but not for the 15-day and 90-day forward markets. However, some evidence of parameter instability is found. In particular, there is evidence of a structural break in several of the reported equations after February 1985; the time of the first major depreciation. For the period after February 1985, each of the markets was found to be speculatively inefficient, in the sense that other available information improves upon the forecast of the future spot rate that is provided by the forward rate.
It must be stressed that due to the joint nature of the hypothesis it is impossible to state whether the observed deviations from this definition of speculative efficiency were due to market inefficiency (i.e., agents not using available information optimally) or risk aversion (i.e., risk averse speculators may drive a risk premium or wedge between the market's expectation of the future spot rate and the current forward rate). Because of the limited sample period since the float it is difficult to conduct more sophisticated tests for the existence of particular forms of risk premia in the forward market.