Are Mutual Funds Hurt by Their Shareholders' Decisions?

Forced Decisions Mutual fund critics often argue that mutual funds are undermined by their shareholders. Cash flows from sales, or redemption requests, force managers to make transactions that they otherwise would not. Even if those trades prove to be sound, they generate costs.

The concern about transaction costs is overstated. It does apply to funds that invest solely in illiquid bonds, which have large gaps between their bid and ask prices. With other funds, that rationale dissolves. Brokerage fees for stock trades have effectively gone to zero, and spreads aren't far behind. Trading investment-grade bonds is cheap, too.

True, if a fund owns enough of an equity, it can move that stock's price through its trade requests. But such wounds are avoidable. Even huge funds can avoid being placed in such a situation by holding a combination of cash and blue chips. (No active equity mutual fund can move Amazon's AMZN stock. If the largest such fund, American Funds Growth Fund of America AGTHX, held 2% of its assets in Amazon, that amount would represent less than half of Amazon's average daily trading volume.)

Another, more plausible claim is that portfolio managers' undesired actions are not, in fact, sound. Their forced trades are weaker than their unforced choices. That is certainly possible. Presumably, actions that are thrust upon managers are less thoroughly considered than are their voluntary trades. The former are acts of desperation; the latter are decisions of intent.

Red Herring Such is the apparent finding of an unpublished academic paper, "Do Mutual Funds Trade on Earnings News? The Information Content of Large Active Trades." (Linda Chen, University of Idaho; Wei Huang, St. John's University of Minnesota; and George Jiang, Washington State University.) The authors find that U.S. stock funds' large trades, as they define the term, "have significant predictive power" under certain circumstances.

Veteran financial journalist Brett Arends of Barron's regards this paper as demonstrating that "mutual fund managers would do a better job if investors would just leave them alone." That is a step too far.

In the paper, the authors determined that when calculated over very short time horizons (one, two, and three months) and when evaluating certain types of trades (the authors' definition of large, under specified cash flow conditions), some funds appear to be superior to others at anticipating which stocks will report quarterly earnings surprises, and on average, those superior funds record higher total returns during the next 12 months.

That is an intricate inquiry, only indirectly related to the topic of forced versus unforced trades. The authors questioned the extent to which portfolio managers anticipate quarterly earnings announcements, not whether they make better investment decisions when they are unpressured. Indeed, the authors did not even bother to measure the results of forced trades; they simply tossed those cases aside, as being irrelevant to their study.

My point? The notion that mutual fund managers are bossed around by their shareholders has not truly been tested. Distinguishing between trades that are caused by shareholder actions and those that are not is guesswork when looking from the outside, and not necessarily a straightforward task even from the inside. Understandably, Arends seized upon research that, at long last, appeared to address this common question. But it did not.

The Closed-End Evidence An alternate approach is to compare mutual fund performances to those of closed-end funds. By definition, closed-end shareholders can neither redeem their holdings nor request that new shares be issued. They would therefore seem to be the ideal control group, for measuring the extent to which (if any) mutual funds suffer from the shareholders' actions.

Unfortunately, closed-end funds are an odd breed. They rarely compete directly with mutual funds, partially because many closed-end funds use leverage (which mutual funds generally do not), and partially because they tend to be offerings. For example, exactly six large-blend U.S. stock closed-end funds are currently unleveraged.

Happily, closed-end funds being a veteran breed, all six funds possess 10-year records. Five have performed adequately to well during that time period; one has been notably poor. The table below shows their average results over the past decade, computed with and without the outlier, against those of the average large-blend mutual fund.

Call the contest a draw. The full group (such as it is) of closed-end funds matched its open-end rivals. The closed-end funds have slightly higher total returns but also slightly higher standard deviations, leading to identical Sharpe ratios (which measure risk-adjusted return) among the closed-end and open-end funds. That outcome does not support the hypothesis that forced sales hurt mutual funds.

Removing the weakest of the closed-end funds, Foxby Corp FXBY, leads to a different conclusion. I must confess: Discarding outliers to craft a new narrative is a wicked practice. However, I also must point out that doing so while extending the analysis to include the trailing five- and 15-year periods leads to a consistent result: Without Foxby, the closed-end funds have been moderately but reliably stronger.

Why Not? A sketchier study you will rarely encounter. It does, however, beg a final question. Why aren't there more conventional closed-end funds? Funds that attempt not to differ from open-end funds by being leveraged or following a niche strategy, but instead simply by being better? It could well be that fund managers benefit from not facing cash flows. Many observers over the years have made that suggestion. Yet, in real life, the proposition has barely been tested.

One of those mysteries, I suspect, that will be resolved by considering how funds are marketed. Not every query about how funds are structured has an investment-related answer.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.