Saturday, October 29, 2005

The best predictor of mutual fund performance

[10/28/05] According to an S&P study, the most consistently top-performing funds had three common characteristics: longer manager tenure, lower expense ratios, protected downside.

[8/8/05] the expense ratio isn't only the best predictor of performance, it is the only statistically reliable predictor, according to a study by Boston-based Financial Research Corp.

FRC tested 11 popular criteria that investors use in picking mutual funds: Morningstar ratings, past performance, turnover ratios, asset size, expense ratios, fund-manager tenure, net sales as well as four measures of risk/volatility.

The results showed that out of the 11 criteria, the expense ratio was the only one that had sufficient statistical relevance in predicting fund performance. Funds with low operating costs "deliver above-average future performance across nearly all time periods," the study says.

2 comments:

Chris said...

I don't dispute the fact, but I do wonder what causes the expense ratio to be higher.

Is it reasonable to assume that the large fund houses - T. Rowe Price, Fidelity, Vanguard, etc. - attract and retain some of the best talent? The big players all can afford to have low expense ratios across the board, and so perhaps this then helps to explain the performance. Conversely a small operation, or new operation, may be feel beholden to a much higher fee in order to help pay the bills, or line the pockets..., because the total assets under management are likely small.

Just a thought, but I bet that high expense ratios are themselves a result of something that probably relates to lower performance. Something such as greed, a lack of experience, or less resources. Resources could be a crack research staff, so a manager doesn't have the talent, and then has to spread himself/herself even thinner trying to perform at several roles - researcher, salesman, investment advisor, marketer, etc.

Interesting food for thought, good post.

Mike said...

Chris (my loyal reader) pointed that Barrons recently had an article covering this topic

Barron's - Monday, August 15th, 2005

Fund Fees: Up or Down?

By JIM MCTAGUE

BACK IN 1989, A COMPAQ DESKTOP COMPUTER with an Intel 486 processor
retailed for about $15,000, and companies snapped them up because the then-cutting-edge machines promised to provide big gains in productivity.

That same year, the annual cost to investors of owning a mutual fund -- what prospectuses refer to as the "expense ratio" -- averaged 0.97% of assets, or $9.70 per $1,000 invested per year. These annual fees pay for such things as administration, management, operations, advertising and distribution.

Today, you can spend less than $500 for a Compaq that is so powerful it makes the old 486 look like a Model T. Yet the cost of owning a mutual fund has risen to about 0.99%, or $9.90 per $1,000, despite
technological advances that should have lowered funds' overhead.

Where did all the savings from the productivity gains go? That's a matter of contentious debate.

Investor advocates like Russel Kinnel, director of research at Morningstar, the Chicago-based fund-rating service, say that fund managements have spent the money to promote more sales, helping their own bottom lines at the expense of their customers.

Mutual funds, on the other hand, say the industry average distorts the true picture, which sees most investors enjoying lower fees. The overall industry number is higher than in 1987, they argue because it reflects the start-up costs of thousands of new funds launched since that date.

At the end of 1989, there were 2,935 funds with $981 million in assets. At the end of 2004, the number of funds had swelled to 8,044 and they were overseeing $8.1 trillion in assets. This year, through June 30, the number of funds had dropped by 92, to 7,952, according to the Investment Company Institute, the industry's largest trade group.

Expense ratios become very important over time because they can significantly reduce an investor's total return. In addition, says one Morningstar study, expenses are a better predictor of fund performance than historical returns.

In a provocative piece last April for a Morningstar publication "FundInvestor," Kinnel asserted that an investor who randomly selects a fund with expenses in the lowest 25% and five-year returns in the worst 25% probably will have better future returns than the investor who buys a fund with five-year returns in the top 25% and expenses in that same quartile. "Higher expenses don't get you better management," he wrote.

If this is the case, investors should ignore those lists of last year's best-performing mutual funds, Kinnel argued recently. If anything, he said, the previous year's top funds probably are near their apex in both share price and performance.

Kinnel contends that the industry's rapid growth in the number of funds and assets under management since 1989 should have resulted in lower expenses, given the possible economies of scale. His expense ratios are slightly lower than the ones we cited above. Our source was the ICI. According to Kinnel's calculations, the average individual investor is paying 0.96% today versus 0.94% in 1989.

Why have expense ratios moved up? Kinnel says that middlemen have captured the money saved from running more money more efficiently. Simply put, fund companies are encouraging brokerage houses to sell their products by giving them a small piece of the action. Many funds that are sold through brokers have B, C and D shares, each of which carries a different fee structure.

A sales charge or "load" that, in former times, was paid upfront is now embedded in the annual expense charges. "If this hadn't happened, the dollar-weighted expense figures would have fallen since 1989," Kinnel claims in his article.

He also points an accusing finger at mutual-fund supermarkets, such as Schwab's One Source.

These offer investors the convenience of being able to buy funds from a number of companies in one place and having one statement each month, instead of one per fund. But such convenience comes at a cost. Funds are charged 20 to 40 basis points (each equal to 1/100th of a percentage point) a year for their listing, and the charge is passed through to investors, including those who aren't investing in them through a supermarket.

Sean Collins, an economist with the Investment Company Institute, argues that Kinnel's conclusions are wrong. He says that if you look at individual funds, expense ratios typically have gone down since 1989, as fund assets have grown. Most individual investors have seen expenses fall, he contends, even though the average expense ratio for the mutual-fund universe is up.

Collins resolves this paradox by claiming that many of funds created since 1989 have attributes that skew the overall number. For example, many of the newcomers are small and thus lack economies of scale. "In a period in which you create a lot of new funds, average expense ratios will look like they are rising," he adds.

In addition, he says, the new funds are unfamiliar to investors and so must spend more on advertising and distribution to attract assets. And, he points out, some incur higher expenses by virtue of their focus. International funds, for example, are more expensive to run because it costs more to buy and sell non-U.S. stocks. High-yield bond funds often have higher costs as well.

In 1989, there were 128 international funds, according to the ICI. As of June, there were 824. There's no breakdown for high-yield bond funds in the data; but during the same period, all bond funds increased from 1,004 to 2,023.

Collins also cites the fund supermarkets as contributing to higher costs, but adds that people want the extra services they afford, which weren't readily available in the 1970s and the 1980s. "Kinnel seems to want to roll back the world," he says.

Kinnel doesn't deny that some individual funds are reducing their fees. In fact, he says, investors are flocking to the funds that are. He also finds that advisers and retirement-plan sponsors are providing cheaper funds in their lineups more frequently than in the past.

Vanguard, the only major fund family actually owned by its shareholders, pioneered the low-fee fund and direct selling. Its funds' average expense ratio was 0.23% in 2004, versus 0.39% in 1989.

Mike Miller, a managing director of the company, says it has become more difficult to trim expenses because of new Securities and Exchange Commission disclosure regulations, which he predicts will squeeze smaller players out of the industry. His company's big advantage is that it pays no one for distribution, choosing instead to sell directly to the investor. Vanguard recently reduced expenses even more for investors whose assets exceed $100,000, some to as low as 0.07%.

Kinnel points out that fund groups like Fidelity and T. Rowe Price had expense ratios below average in 1989 and that they've continued to lower them.

"Conversely, Aim and Franklin Templeton are now charging more than 40 basis points above what they charged 15 years ago," he adds.

An AIM spokeswoman said that Kinnel didn't include more than $36 million in fee reductions by AIM so far in 2005. She also said that he failed to note that, since 1992, AIM has introduced seven international funds, seven sector funds and six global funds, all of which tend to have higher expense ratios than U.S. domestic stock funds.

Similarly a spokeswoman for Franklin Templeton, says: "Since 1989, we've seen an evolution in mutual-fund pricing, including changes to front-end sales charges, the addition of 12b-1 fees, and the introduction of alternative share classes. With all of these considerations, simply comparing expense ratios at different points in time does not present a complete picture of shareholder costs." (12b-1 fees usually are imposed to provide promotional funding to get brokers to sell a fund.)

Vanguard's Miller says he believes that the larger funds have tried to reduce costs, though not all succeed in reducing them across all their offerings.

Some of the reduction has been engendered by price competition and the introduction of exchange-traded funds, which have very low management fees. Some of the reductions came about because of actions by New York State Attorney General Eliot Spitzer.

Miller predicts that regulations requiring greater fee disclosure could encourage further expense reductions. Ultimately, the heaviest inflows of cash will be recorded by funds with the lowest costs, he says. And if Kinnel's theory about low costs being the best indicator of future performance is correct, investors in these funds will have a lot to smile about down the line.

http://online.barrons.com/article_print/SB112388888118012232.html