“…When applied on US data, this methodology generally leads to a rejection of the hypothesis (see for instance Evans and Lewis, 1994, Campbell and Shiller, 1991, Shea, 1991. Somewhat better results have been obtained with data on other countries, although the evidence in favour of REH is far from overwhelming (see among many others Prats Albentosa and , Cuthbertson et al, 2000, Ghazali and Low, 2002, Cooray, 2003. Using another approach, based on a linear regression of the expectation error on the lags of interest rates, Johnson (1997) Kalev and Inder (2006) express serious doubts about the econometric validity of these results; with more sophisticated -although linear-methods, they reach opposite conclusions.…”