We merge survey data on a sample of individual investors containing test-based measures of financial literacy with administrative records on their assets holding and trades before, during and after the financial crisis of September 2008. This dataset allows us to design three tests of the benefits of financial literacy by comparing the decisions actually taken by individuals with a dominated alternative. We find that high-literacy investors are better at timing the market, since conditional on exiting the stock market they are more likely to exit before rather than after the crash following the collapse of Lehman Brothers. High-literacy investors are also more likely to trade according to the prescriptions of normative models and to detect intermediaries’ potential conflicts of interest. However, though statistically significant these effects are economically small. In fact, far too many investors, even among those with high literacy, tend to choose the dominated alternative along all dimensions of choice examined. This suggests that literacy may be a poor edge against financial mistakes.
