DOSPERT's gambling risk-taking scale predicts excessive stock trading

Using a data set that combines trading records in a financial investment simulation with survey responses, this study provides evidence that a domain-specific variant of risk-taking propensity, namely risk taking in gambling (but not in investing) situations, predicts the volume of trades of financial investors. We find that investors' gambling risk-taking propensity, measured by the Weber, Blais, and Betz [2002], Domain-Specific-Risk-Taking (DOSPERT) gambling subscale, increases the number of trades made and hence transaction costs, as well as the extent of their day trading. The short (four-item) gambling risk-taking propensity DOSPERT subscale thus provides a useful diagnostic addition to risk attitude assessment instruments for private investors.