A wide range of empirical biases hampers hedge fund databases. In this paper we focus upon survival-related biases and disentangle look-ahead biases due to self-selection of funds and due to fund termination. Self-selection arises because funds voluntarily report their information to data vendors and may decide to stop doing so. By extending existing methodology, we analyze persistence in hedge fund performance over the period 1994–2000, taking into account the above biases. The results show that look-ahead biases due to liquidation and self-selection enforce each other and may lead to overestimating expected returns by as much as 8% per year. Overall, the results are consistent with positive persistence in hedge fund returns at horizons of two and four quarters.

This paper analyzes the properties of a number of estimators that can be used to estimate short-run persistence in mutual fund returns. When data for different funds are pooled, it is advisable to correct for cross-sectional differences in expected returns. However, these adjustments may induce biases in the estimated persistence coefficients and thus lead to spurious persistence. Theoretical derivations, combined with a Monte Carlo study, show that these biases cannot be neglected for the samples that are typically used in applied work. We also estimate the short-run persistence in two samples of U.S. open-end mutual funds using quarterly returns for 1987–1994. An important conclusion is that the results are quite sensitive to the estimation method that is employed.

We analyze the performance persistence in hedge funds taking into account look-ahead bias (multi-period sampling bias). We model liquidation of hedge funds by analyzing how it depends upon historical performance. Next, we use a weighting procedure that eliminates look-ahead bias in measures for performance persistence. In contrast to earlier results for mutual funds, the impact of look-ahead bias is exacerbated for hedge funds due to their greater level of total risk. At the four-quarter horizon, look-ahead bias can be as much as 3.8%, depending upon the decile of the distribution. We find positive persistence in hedge fund quarterly returns after correcting for investment style. The empirical pattern at the annual level is also consistent with positive persistence, but its statistical significance is weak.

Little is known about how different bonus schemes affect traders' propensity to trade and which bonus schemes improve traders' performance. We study the effects of linear versus threshold bonus schemes on traders' behavior. Traders buy and sell shares in an experimental stock market on the basis of fundamental and technical information (past share price evolution, realized earnings, analysts' earnings forecasts, and evolution of the market index). We find that linear and threshold bonus schemes have different effects on trading behavior: traders make more transactions but of a smaller size under the threshold than under the linear bonus scheme. Furthermore, transaction frequency significantly decreases when bonus thresholds are reached but only after building in a safety margin. Under the threshold scheme, the traders' performance is lower (even when there are no transaction costs) than under the linear bonus scheme as a consequence of poorer market timing. This is especially the case when earning money by trading is relatively difficult (i.e., under low profitability conditions). Nevertheless, under low profitability conditions, traders seem to collect more information about the relationships between share price and market returns, earnings, and earnings forecasts, put more effort into understanding those relationships, and thus eventually learn to perform better.