Random walks in asian foreign exchange markets:evidence from new multiple variance ratio tests

This paper revisits the random walk hypothesis for ten Pacific Basin foreign exchange markets. The results suggest that the null hypothesis of random walk is rejected based on the Lo-MacKinlay variance ratio tests, under conditions of both homoskedasticity and heteroskedasticity for the examined series. The use of a battery of new joint variance ratio tests provide further evidence against the random walk behavior than the conventional variance ratio tests. Therefore, we conclude that these Pacific Basin exchange markets violate the random walk hypothesis and are not in line with the weak-form efficient market hypothesis.