The High Cost of Technical Analysis and Speculation

Abstract

We compare transaction record information and survey data of a sample of U.S. full-fee brokerage clients for the 1964-1970 period to those of a similar sample of Dutch discount brokerage clients for the 2000-2006 period. Two striking findings from this comparison pertain to technical analysis. First, investors in the more recent data report using technical analysis five times more frequently than investors in the older data. Second, using technical analysis is very costly: We estimate the monthly cost of using technical analysis to be 60 basis points. Paralleling these results is our finding that investors report having speculation as a primary investment objective 75 percent more often in the recent data than in the older data. The marginal impact of reporting to have speculation as a primary investment objective is to increase the monthly return standard deviation by 180 basis points, with no statistically significant impact on net returns. In line with studies using the older data, we find that technical analysis and speculation are primary drivers of excessive trading and that excessive trading impairs investor performance. In contrast to the existing behavioral finance literature, however, we find only weak evidence for the claim that overconfidence is a primary driver of turnover and investor performance.