Recent research has examined the performance of stocks held by U.S. mutual funds and found they realize abnormal returns. The result is significant as it stands in contrast to the general consensus from traditional performance studies that mutual funds do not possess superior information. Employing a unique dataset, I examine the performance of the monthly stock holdings and trades of a sample of Australian fund managers. When stock holdings are observable, performance measures can be constructed that are more precise than traditional fund manager performance measures. I find the stocks held by fund managers realize abnormal returns consistent with some stock selection ability across fund managers. Examining the performance of their individual trades, I find that the stocks they buy realize abnormal returns whereas for sell trades I find no evidence of abnormal returns. Overall, the results suggest fund managers have the ability to select stocks that realize positive abnormal returns thus providing out-of-sample support for similar recent findings for U.S. mutual funds.

The binary classification of firms into profits or losses represents a powerful heuristic. The literature that has examined the impact on the firm of this earnings heuristic has focused on the earnings management actions of small profit firms. The impact of this earnings heuristic on the actions of firms reporting accounting losses and the decision-making effects the heuristic may have other than earnings management have not been examined. In this study we hypothesize that reporting an accounting loss acts as a heuristic trigger for firms to exercise the abandonment option and discard unproductive investments. The results are consistent with the hypothesis. We find that there is a sharp and economically significant discontinuity around zero in the level of investment in labor between small profit and small loss firms. The discontinuity is due to loss firms having a lower-than-expected level of investment in labor, given their economic fundamentals. Further tests show that this discontinuity is due to the exercise of the abandonment option. We find that firms switching from a profit to a loss cut labor to a greater extent than other firms with similar changes in earnings that do not pass the loss threshold. Taken together the results are consistent with the accounting loss heuristic acting as a major disciplinary or incentive altering event that resolves agency problems.