Date of Award

Summer 2011

Document Type


Degree Name

Doctor of Philosophy (PhD)


Business Administration-Finance

Committee Director

John Doukas

Committee Member

Mohammad Najand

Committee Member

Licheng Sun

Committee Member

David Selover


This dissertation examines the performance, skill and trading characteristics of individual currency traders by examining daily returns and transaction data for 428 individual currency traders from 2005 to 2009. Additionally, we examine whether technical trading strategies are profitable for individual currency traders.

The first essay examines the performance and trading characteristics of individual currency traders. Examination of daily returns for 428 accounts from March 2004 to September 2009 shows traders are able to earn positive excess returns, even after accounting for transaction costs. Additionally, the results reveal that day traders not only trade more frequently than non-day traders, but also outperform them based on raw, passive benchmarks and on a risk-adjusted return basis. Furthermore, sorts on trade activity, measured as the mean number of trades per day per account, and account turnover, show a positive association between performance and trade activity. Robustness checks of gross performance and trade activity, proxied by mean number of trades per day, are similar when analyzing a second data set that consists of 74 accounts from July 2010 to August 2011. Consistent with the prediction of the calibration theory the results also show that the more traders trade, the more feedback they receive, which, in turn, decreases their overconfidence and increases performance.

The second essay examines whether individual currency traders are skilled. Unlike previous studies that examine the predictability of R2 for professional investors, who actively manage their portfolios and do not follow benchmarks, and find that R2 can predict future performance, this study reveals just the opposite: R2 does not predict future performance for individual currency traders. Despite the lack of predictive power of R2, we report that individual currency traders are skilled. The R2 measure lacks predictive power because R 2 is not persistent, which is because individual currency traders change their trading styles over time, while earning positive and persistent alphas. Our analysis of trade activity, drawdown, and market timing provides additional support that individual currency traders possess trading skills. Top traders also have the ability to mitigate downside losses, and a sizable percentage of them can time currency market factors. We find that 68.78 percent of trades executed by the top traders are profitable net of transaction costs, and profits do not arise from chance.

The third essay investigates whether technical currency trading is profitable. The results show that the use of technical analysis by individual currency traders is negatively associated with performance. Further, the technical trading model developed here adequately describes the cross-section of returns for individual currency traders. This result arises because individual currency traders use well-known technical indicators to trade currencies. This implies that such currency traders suffer from reduced performance.