Wednesday, January 26, 2005

Maybe fund managers can beat the market afterall. An Analysis of Mutual Funds by Alexander, Cici,Gibson

SSRN-Does Motivation Matter When Assessing Trade Performance? An Analysis of Mutual Funds by Gordon Alexander, Gjergji Cici, George Gibson

Alexander, Cici, and Gibson provide some interesting evidence that mutual fund managers may be better at picking stocks than previous literature has suggested. The authors show that when mutual fund trades are broken down by the motivation for the trade (for example liquidity driven, tax-loss driven, or for valuation reasons), fund managers do earn abnormally positive returns when trading for valuation reasons.

In the words of the authors:

"We argue that a fund manager who believes that a stock is significantly mispriced will want to trade in its shares. However, heavy investor outflows (inflows) will constrain the manager by forcing him or her to control liquidity by selling (buying) stocks. Accordingly, we condition fund trades on the direction and magnitude of investor flows."

"The hypothesis that fund managers possess the ability to value stocks finds strong support. Valuation-motivated buys (i.e., large buys concurrent with heavy outflows) outperformed their benchmarks by an average 7.95% in the following year whereas
valuation-motivated sales (i.e., large sales concurrent with heavy inflows) underperformed by an average of 1.10%. The 9.05% differential between buys and sales is economically and statistically significant. Results are slightly stronger when October [the month where tax trading would occur given the October tax year end for mutual funds] sales and mandated reporting-month trades are excluded in order to remove tax motivated and window-dressing trades."

Opposed to the valuation trades, fund managers fail to do as well for liquidity driven trades:

"In sharp contrast to valuation-motivated buys, liquidity-motivated buys (i.e., small buys concurrent with heavy inflows) underperformed their benchmarks by an average 1.65% in the following year."

To further buttress the findings, the authors also investigate "buys that add currently unheld stocks to the portfolio and sales that fully terminate existing positions." The findings are again consistent with the managers ability to value stocks better than the overall market.

"We find that initiating buys outperformed their benchmarks by an average of 4.19% in the year after the trade, whereas terminating sales underperformed by an average of 1.07%. The 5.26% differential is both economically and statistically significant, providing confirming evidence that fund managers possess the ability to value stocks."

A few questions that this raises:

  1. How are managers picking these winners and what does this ability say about market efficiency? For instance: are the returns from buys that initiate positions driven by IPOs? and if so are the losing trades rewards to the investment bankers? This would be consistent with Nimalendran, Ritter, and Zhang's recent work.
  2. Why can the managers pick winners better than they can identify losers?
  3. Why don't closed end funds (which are largely free of liquidity driven trading) perform better than open-ended funds?
  4. Unrelated, how large a role does the October tax-year end for mutual funds in returns for October? December? January? How does this affect the January Effect?

Overall a VERY interesting paper! Definitely I^3!!! Although I would suggest a new name. When I clicked on the article I expected a managerial pay paper.

Suggested Citation
Alexander, Gordon J., Cici, Gjergji and Gibson, George Scott, "Does Motivation Matter When Assessing Trade Performance? An Analysis of Mutual Funds" . AFA 2005 Philadelphia Meetings.

No comments: