The rationale for the superior performance of contrarian investment strategies remains a matter of lively debate. The orthodox view maintains that such strategies generate higher returns because they are fundamentally riskier, whereas the behaviorists suggest that the superior performance is a result of systematic errors in investors' expectations about the future. If the behavioral view is accepted, then the debate centers on what the underlying source(s) of such errors areānaive extrapolation of past performance or biased analysts' earnings forecasts. Using stocks listed on the London Stock Exchange, we found evidence consistent with the view that errors in expectations are more likely to be a result of biases in analysts' earnings forecasts than naive extrapolation of the past. We also found that positive and negative earnings surprises have an asymmetrical effect on the returns of low- and high-rated stocks. Positive earnings surprises have a disproportionately large positive impact on stocks that are priced low relative to four measures of operating performance; negative surprises have a relatively benign effect on such stocks.