Index Funds Aren't All Equal

High costs and managers who hold cash too long -- or pay inflated commissions -- can shave returns. So it pays to compare

Does it really matter which Standard & Poor's 500-stock index fund you own? After all, these funds have identical portfolios -- composed of the 500 stocks in the S&P 500, held in proportion to their weight in the index. You bet it does. A BusinessWeek analysis of these funds reveals some startling differences in their ability to keep pace with the benchmark.

Some funds -- including the $96.9 billion Vanguard 500 Fund -- come very close to matching the index' return year after year. But others miss by surprisingly wide margins, thanks to factors that can include high expense ratios and managers who sit on cash, churn holdings, and -- according to a recent Securities & Exchange Commission probe -- pay inflated trading commissions to cover investment research. That has drawn criticism since S&P 500 funds buy what's in the index, so they don't need Wall Street's buy and sell recommendations. The lesson? "All index funds aren't equal," says Jerry Chan, a vice-president and index researcher at J.P. Morgan Chase (JPM ).

Deviation from the index -- plus or minus -- is called "tracking error," and in 2003, shortfalls were especially pronounced, thanks largely to managers who held cash in a rising market. Indeed, over 80% of the S&P 500 index funds that Morningstar follows fell short of the performance ideal of a plain-vanilla index fund: to deliver the benchmark's return, minus the fund's expense ratio. According to Morningstar data, last year the average S&P 500 fund missed by the amount of their expense ratios plus an additional 0.38%, or 38 basis points (one basis point is 1/100th of a percent).

That margin of error is startling to industry experts. When it comes to the S&P, "you should be within five basis points" of the index's performance, adjusted for fund expenses, says Gus Sauter, Vanguard's chief investment officer. "Ten basis points would be stretching the limit."

Because shortfalls compound over time, it's important to scrutinize the performance of index funds just as you would an actively managed fund. Indeed, someone who had invested $10,000 in the S&P 500 in 1995 would have had $28,179 on Dec. 31. That's $865 more than you'd have in an identically priced investment that returned a mere 0.38% less each year, according to Morningstar.

In evaluating index funds, start with what's generally the biggest driver of tracking error -- the expense ratio. Since these costs are deducted from a fund's earnings, keeping them low can mean the difference between getting a near-market return and underperforming. The average S&P 500 fund charges about 0.48% of assets, notes Jeffrey Ptak, a Morningstar mutual-fund analyst. Expenses range from a low of 0.15% at SSgA's S&P 500 Index to 1.5% at several funds. "Costly index funds are about as self-defeating a proposition as you can find," says Ptak.

When it comes to tracking error, the expense ratio is only part of the story. Consider the Invesco S&P 500 Index fund. The gap between its 2003 return of 27.62% and the index' 28.67% gain is 1.05 percentage points. But the fund's expense ratio of 0.65% accounts for only part of the shortfall. A spokesman says Invesco switched the fund from an outside manager to one in-house last August because of "unacceptable" tracking error.

Invesco declined to comment on the reason for the fund's underperformance. But in general, when index funds lag by more than their expense ratios, cash and transaction costs such as commissions are primarily to blame. In a bull market, it's imperative to invest incoming cash immediately since stock prices are more likely to rise than fall. Some S&P 500 funds, including those run by Vanguard, Fidelity, and T. Rowe Price (TROW ), can put their cash to work quickly by buying S&P 500 futures contracts and then replacing them with stocks. This helped all three funds finish close to the index in 2003.


Still, some index funds are prohibited from buying futures; check the "Statement of Additional Information" filed with the SEC for restrictions. Others, such as Columbia Large Company Index fund, don't generally use them: "Futures don't track the index perfectly -- they can be over- or undervalued," says portfolio manager Eric Remole. "Using them can introduce tracking error." The fund's "Z" shares trailed the S&P last year by 70 basis points -- that's 19 more than its expense ratio. All but 3 basis points of that gap can be explained by the fact that fund expenses are taken out of assets throughout the year, leaving less to invest each day, Remole says.

Commissions also contribute to tracking error. These fees aren't factored into a fund's expense ratio but are deducted when stocks are bought and sold. If a manager is paying inflated commissions to generate "soft dollars" -- credits that can be used to pay for investment research -- "tracking error will get pretty large," says John Chalmers, associate finance professor at University of Oregon.

How much is a fund paying in commissions? Check its Statement of Additional Information. The Morgan Stanley S&P 500 Index Fund, for example, paid $271,513 in commissions in fiscal 2002. The Vanguard 500 Fund, in contrast, paid $4.386 million. But as a percentage of assets, the Morgan Stanley fund's rate was higher -- 0.0169%, vs. 0.0057%.

The Vanguard 500 fund is generally able to negotiate lower commissions than its smaller rivals. Still, in a lawsuit filed in U.S. District Court in New York in August, a shareholder alleges that the Morgan Stanley S&P 500 Fund paid inflated commissions in 2002 to earn soft-dollar credits to buy stock research. Morgan Stanley responds: "Three appellate courts and various district courts have rejected similar claims made by the same attorney against other firms. We believe this case is equally lacking in merit." In 2003, the Morgan Stanley Fund's "B" shares -- the largest among its share classes -- missed the S&P by 1.83%. That gap consisted of a 1.5% expense ratio plus 0.33% of additional underperformance. In 2002, however, the fund trailed the index by less than its expense ratio.

It also helps to check a fund's annual report for turnover. When Standard & Poor's (which, like BusinessWeek, is a unit of The McGraw-Hill Companies) adds or drops a stock in the index, S&P 500 funds must follow suit. Typically, the index turns over about 5% of its market cap a year, says S&P index strategist Srikant Dash. But in part because funds also buy and sell as shareholders come and go, the average S&P 500 fund turned over 13% of its portfolio in 2003.

At some funds, turnover was far greater. The $105 million MTB Equity Index Fund had an 87% rate. Why? Shareholders came and went quickly as "institutional portfolio managers used our fund" to get quick exposure to the market and then bailed out once they "figured out what to do with the money," says co-manager Peter Hastings. The fund levies expenses of just 0.25%. Yet in 2003, it trailed the S&P 500 by 0.77%.

High index-fund expense ratios are likely to decline thanks to SEC scrutiny. But tracking error may remain a problem. The upshot: While many investors think of these funds as commodities, their performance may continue to be far from uniform.

By Anne Tergesen & Lauren Young

    Before it's here, it's on the Bloomberg Terminal.