Year End Review

The year end reviews of Mutual Fund performance have been published and as usual they present information that is likely to be like poison in the hands of most investors. Most dangerous are the lists of "top performing" mutual funds which will undoubtedly wind up on the buy list of many investors despite the fact that past performance of mutual funds has been proven by extensive academic research to be of little or no use in predicting future performance. Same goes for the gushing articles about investing geniuses at mutual fund company X or Y.

The sections aren't useless however, and if you connect the dots a bit you can reach some interesting conclusions. In the midst of a puff piece on the mutual fund company T Rowe Price
the NYT notes that:

... T. Rowe Price has come lumbering back. According to Morningstar, the average return of its funds was 9.85 percent, annualized, over the five years ending Nov. 30, edging out rivals like American Funds, Fidelity Investments and Vanguard.

Now there is a useless statistic !! The average return of a fund company is an absolutely meaningless statistic. No one invests their money in equal amounts across all of the funds of a particular company so no investor is ever going to get the average return of a fund company (and btw is that a simple average of all funds, only stock funds, weighted by size of the fund etc etc ?).

And comparing the average return across fund companies is equally pointless. Does Fidelity have more short term bond funds designed to provide lower returns at lower risk than T Rowe? Does T Rowe have more international funds than its competitors and the statistic simply reflect the great performance of international stocks in the last five years.

The article goes on to state:


With that success has come new customer cash. T. Rowe Price collected nearly $40 billion in net inflows from the beginning of 2002 through October last year, nearly double its amount from 1995 through 1999.

Sounds great right ? Well not exactly, the large inflow of cash makes it exceedingly unlikely that going forward the performance of T Rowe will continue to be as stellar as described. In fact T Rowe funds will increasingly run into one of the many problems with actively managed (as opposed to index) funds. The large amounts of cash they need to invest will be a drag on performance.

Interestingly that same NYT section presents the findings of an academic study on just what the impact of "liquidity trades" trades necessitated by investments or redemptions in a fund as opposed to "valuation trades" trades actually based on a manager's strategy. The study found a hypothetical portfolio of valuation trades to outperform the liquidity trades by 3.5%.

The article notes (my bolds)

The relative performance of the two liquidity portfolios was nearly opposite that of the valuation-motivated versions, with the buys now lagging the market and the sells outperforming it. In fact, the liquidity-motivated sells outperformed the buys by two percentage points a year, on average. In other words, the transactions that managers must make to meet liquidity needs eliminate much of the profit from trades made for valuation reasons. Transaction costs, which were not included in the calculations, coupled with management fees, eliminate the rest of that profit for the average fund

The Professors conclusions:

The professors interpret their findings to mean that mutual fund investors pay a high price for the chance to invest or withdraw money at any time. And the cost is borne by all investors, whether or not they take advantage of the opportunity. The investment implication is clear: don’t pay for liquidity unless you need daily access to your money.

How does this relate to the T Rowe story ? Quite simply a fund company "blessed" with massive cash inflows is going to burdened with the need to make large amounts of liquidity trades that are going to be a drag on performance.

Once again the academic research leads to the conclusion that actively managed mutual funds have a tremendous hurdle to beat simply to match their relevant index. The professors conclude that the liquidity trades, transactions costs and management fees (not to mention the taxes paid by investors) drag down the performance of actively managed mutual funds. Since index funds invest across an entire index fund consistently they do not engage in buys and sells of individual stocks based on liquidity.

What is the best choice for investors according to the study's author:

Professor Gibson says he instead favors open-end mutual funds that significantly restrict short-term trades

While the Professor seems to still hold out hope for active managers (probably because he did not quantify all the costs of an active fund) In my practice I have found an even better solution:

The funds of Dimensional Funds Advisors (DFA) although they are open end funds preempt the problem of excessive movement in and out of their funds. They make the funds available to individuals only through advisors committed to long term investing. In fact if they see a particular advisor is rapidly moving monies in and out of their funds they throw him out. Thus even their index funds targeted at smaller parts of the market (micro cap or emerging markets value for example) have little need to engage in trades simply to create liquidity.