Academic journal article Financial Management

The Telling Trades of Mutual Funds

Academic journal article Financial Management

The Telling Trades of Mutual Funds

Article excerpt

Under the assumption that mutual funds trade at quarter commencement, some funds exhibit and exploit persistent stock selection talent; that is, the stocks purchased consistently outperform the stocks sold, and the higher turnover of these funds indicates that managers are capitalizing on their forecasting abilities. However, any evidence of sustained stock selection skill disappears when alternate trade-timing assumptions are considered, suggesting that some skilled managers are electing to trade earlier than previously assumed. Overall the results question the appropriateness of the quarter-end trading assumption and the validity of existing studies that employ it.


The question as to whether or not fund managers have stock selection ability is an intriguing one examined in the literature and approached from a variety of angles. If funds are able to forecast stock returns, then some researchers have argued that skilled managers' portfolio holdings should persistently outperform unskilled managers' portfolio holdings. Other researchers have posited that aggregate fund purchases should subsequently outperform aggregate fund sales. Believing that true talent should be exhibited repeatedly, I feel that a better test of mutual fund stock selection ability lies in a combination of these two approaches. Therefore, in this study, I ask whether some funds are able to consistently purchase stocks that subsequently outperform the stocks they sell. Furthermore, I test for this recurrent forecasting ability under a variety of trade-timing scenarios, questioning the appropriateness of the most widely accepted trade assumption.

Historically, researchers questioning whether mutual fund managers possess stock selection ability have examined whether the returns generated by fund holdings are inter temporally related. That is, the assets of funds headed by skilled (unskilled) managers should generate superior (inferior) returns year after year. The results of these performance-persistence analyses are mixed. While Jensen (1968) and Malkiel (1995) argue against fund managers possessing private information, Goetzmann and Ibbotson (1994), Elton, Gruber, and Blake (1996), and Wermers (2003) discover that previous performance is related to future performance. However, this positive persistence may be driven largely by repeat losers (Brown and Goetzmann, 1995), generate benefits that are entirely subsumed by expenses (Wermers, 2000), or be attributed not to managerial ability but instead to common factors driving stock returns, as well as to expense ratios, load fees, and turnover (Carhart, 1997).

Recently, researchers have approached the question of mutual fund forecasting ability from a different perspective. The use of quarterly ownership reports has enabled the focus to shift from fund holdings to fund trades, which are perhaps more likely to be triggered by private information (as compared to the passive decision to maintain holdings). Nofsinger and Sias (1999), Wermers (1999), Cai and Zheng (2004), and Chen, Jegadeesh, and Wermers (2000) discover that stocks widely purchased by mutual funds earn higher subsequent returns than those widely sold. While I believe that this recent approach is better geared toward uncovering any managerial stock selection ability, most of these analyses study mutual fund trading patterns as a group (i.e., cumulating all funds' trades in a given stock).1 Conclusions drawn from this methodology (i.e., testing the percentage change in a stock's institutional ownership) are questionable, as the results could have been driven by a few large trades rather than produced by widespread fund behavior.

The methodology employed in this paper to study funds' trading activities draws from both strains of the literature. That is, after inferring an individual fund's trades (instead of aggregate fund trades), I then follow in the footsteps of the more traditional fund performance studies by asking whether some funds consistently make better (ex post) purchasing and selling decisions than others. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.