Modified R/S and DFA Analyses of Foreign Exchange Market Efficiency under Two Exchange Rate Regimes: A Case Study of Iran

Abstract: Based on efficient market hypothesis, financial markets are impossible to forecast. The purpose of this paper is to examine the weakform efficiency of the Iranian foreign exchange rate (defined by the Rial/Dollar) during time period 1999:25:01 to 2010:17:06 from long memory viewpoint. For this, we have employed three methods of scaling analysis including classical rescaled range (R/S) analysis, modified rescaled range (M-R/S) analysis and detrended fluctuation analysis (DFA). We have divided the time period into two sub-periods, 1999:25:01-2002:21:03 and 2002:21:03-2010:17:06. In the former time period, Iran had a fixed exchange rate regime and in the latter period, the country followed a managed floating exchange rate regime. The obtained results from these methods are not the same. To achieve more explicit conclusions, we’ve used two more widely applied econometric tests namely augmented Dickey-Fuller (ADF) test and Phillips-Perron (PP) test to determine whether or not the time series under consideration behave as random walk consistent with the weak-form efficiency. The findings indicate that the result of DFA is in line with the econometric approach. We conclude that the Iranian foreign currency market at the first sub-period is less efficient relative to the second sub-period. Another important result is that relying on only one method to make a conclusion about market efficiency may be very misleading. Therefore, one should first carefully select more reliable methods and thencompare their results to achieve a reliable conclusion

Keywords:

Forecasting , Scaling , long memory , Rescaled Range Analysis , Detrended Fluctuation Analysis , Unit Root Tests , Random Walk , Efficient Market Hypothesis , Rial-Dollar Exchange Rate series , High-frequency data