Countless money managers set out every day to beat the stock market. Most are very smart, most are very diligent, and most are doomed to fail.
Why? Because smart as these money managers are, the market is smarter. In fact, the market knows everything that can be known about every stock an investor might buy or sell.
This, at least, is the way the world theoretically would work if every stock market were an “efficient” stock market, instantly reacting through price changes to the latest data about a company's prospects. If, for example, Pfizer were to make an important announcement, scores of traders would hit their keyboards, and dozens of investment analysts would adjust their financial models, so the stock's price would quickly reach a new equilibrium reflecting the latest news.
So how smart, meaning how efficient, are the world's stock markets? Or, to put the question a different way, what are the chances that most of those money managers really can beat the market?
This is an excellent time to ask. The 10-year period ending in 2004 saw one of the great bull markets of all time, followed by the worst stock market downturn in a generation, followed by a pretty decent recovery in 2003 to the closing weeks of 2004.
A manager who knew just when to step in or out of the markets, or who made timely decisions on when to hold certain types of stocks such as technology, telecommunications and energy, or who avoided (or shorted) companies before they imploded over accounting scandals, could have posted fabulous returns. All that manager really had to do was buy stocks before they went up, and sell stocks before they went down. How hard is that?
To find out, I used Morningstar's Principia software (updated through October 2004) to compare various types of actively managed mutual funds against index funds that seek only to match, not beat, certain stock market benchmarks. All else being equal, in an efficient market, we would expect an index fund to consistently rank in the upper half of its category, because index funds tend to have lower trading and operating costs than actively managed funds. On the other hand, if active managers usually add value, we would expect the index fund to land consistently in the lower half of the group.
My research is summarized in the accompanying table. Before we look at the results, however, we should consider “survivorship bias.” Mutual funds that perform poorly tend to go out of business. Morningstar's October 2004 data only reflects the performance of funds that still were around in October 2004. The bad track records of funds that folded before that date are removed from the database, and thus the overall performance of actively managed funds is made to look better than it actually was.
Keeping survivorship bias in mind, let's look at the data for stocks in large U.S. companies. The Vanguard 500 Index fund, which closely tracks the performance of the Standard & Poor's 500 Index, outperformed about three out of four of its actively managed peers when we measure returns over 10- and 15-year periods. More recently, it outperformed 59 percent of the actively managed funds over three years, but only 45 percent of the actively managed funds over five years. (If you are looking at the table, note that the data is presented in percentile form, where the best-performing funds would be in the 1 st percentile and the worst would be in the 99 th .)
I suspect the reason that Vanguard 500 Index fund finds itself in the 55 th percentile over a 5-year period is because it stayed fully invested in large-company stocks during the tough years of 2000-2002, while many actively-managed mutual fund managers held larger cash positions and ventured into other asset classes. At December 31, 2002, when the market was near its post-bubble lows, the average large-cap mutual fund had 88% in U.S. stocks, 5% in cash, and 7% in other investments, while Vanguard 500 Index had 99% in U.S. stocks and 1% in cash.
It is a little trickier to perform this analysis for U.S. small-company stocks, because there are no index funds that have a sufficient history. So for this analysis, I estimated an index fund return by subtracting a hypothetical 20 basis point (0.20%) annual expense burden from the return of the S&P 600 Index. Investors today pay that 20 basis points to invest in the index through the iShares S&P 600 Smallcap Index Fund.
In every period — looking back over three, five, 10 and 15 years — our hypothetical index fund landed in the upper half of all the funds in the category. Fewer than half the active managers were able to generate sufficient additional return to compensate for their costs, even after we boost the active managers' statistics because of survivorship bias. Investors who may have assumed the small-cap stock market is naturally inefficient should take note of the consistently superior returns for the index.
Percentile Rank Of Annualized Returns
|Vanguard 500 Index||41||55||22||27|
|S&P 600 Smallcap Index*||29||41||37||44|
|MSCI EAFE Index*||41||56||74||95|
|Vanguard European Stock Index||31||42||40||--|
|MSCI Japan Index*||63||69||--||--|
|*Less hypothetical expenses|
The MSCI EAFE (Europe, Australasia, and Far East) index is a good proxy for non-U.S. stocks, but again, there are no index funds that have a sufficient history for comparison purposes. Therefore, I used the returns of the index itself, reduced by a hypothetical 35 basis point (0.35%) estimated expense charge, which is equal to the current expense ratio of the iShares MSCI EAFE Index Fund. The mutual funds analyzed were all diversified, distinct funds categorized by Morningstar as foreign large cap. The index's performance, less the hypothetical expenses, is as follows:
- 3 Year Return — 41st percentile
- 5 Year Return — 56th percentile
- 10 Year Return — 74th percentile
- 15 Year Return — 95th percentile
The performance of the EAFE against actively managed funds is awful. The statistics reveal that in this asset class, an active manager can add value.
But the picture becomes more complex when we look at certain subsets of foreign stocks. For example, I compared the results of the Vanguard European Stock Index Fund to a group of actively managed funds classified by Morningstar as European equity funds. Because few European equity mutual funds have been around for more than 10 years, I did not compare 15-year returns. In this group, the Vanguard fund, which tracks the MSCI Europe stock index, placed well into the top half. Most active managers have failed to match the index. There is no Japan stock index fund with a long track record. I used the returns of the MSCI Japan index, less a 59 basis point estimated annual expense rate, which is the expense load borne by the iShares MSCI Japan Index Fund. I compared the returns to a group including all distinct, large-cap funds classified by Morningstar as Japan equity funds. Over the past three- and five-year periods, between 60 percent and 70 percent of the actively managed funds outperformed the index.
I did not consider longer terms because only five of today's Japan equity mutual funds have been operating for 10 years or more, which creates too small a sample and too great a risk of survivorship bias. Still, the data for the shorter periods indicates that it is worthwhile to choose an active manager for Japan, which is what we do at Palisades Hudson Asset Management.
To beat the indexes, an active manager must generate enough additional return to offset the extra costs that active management usually entails. Those extra costs are considerable. The average expense ratio for U.S. large-cap mutual funds in Morningstar's database is 1.40%. The subset of that group categorized as index funds sports an average expense ratio of 0.68%, meaning the average actively managed fund starts off with an annual 0.72% disadvantage compared to the average index fund. Large-cap index funds offered by Fidelity, Schwab, and Vanguard all have expense ratios less than 0.20%, making the index advantage for those particular funds greater than one percentage point.
This advantage is even larger for European large-cap index funds. The average expense ratio for European large-cap mutual funds in Morningstar's database is 1.61%. The subset of that group categorized as index funds sports an average expense ratio of 0.79%. In this asset class, the average actively managed fund starts off with an annual 0.82% disadvantage compared to the average index fund. The expense ratio of the Vanguard European Stock Index Fund is a mere 0.32%.
“When an asset class does relatively well, an index fund in that class does even better.” This is Dunn's Law, propounded by California attorney and investment theorist Steven Dunn.
A 1999 study of indexes for the five-year period ending on December 31, 1998 by John Rekenthaler, Morningstar's research director, found that, “...indexes with the strongest total returns tended to post the best relative showings (in their categories), while the losers remained substandard performers (within their categories).” In other words, when an index has a good year, it tends to be a top performer; when the index declines or otherwise struggles, its actively managed competitors tend to beat it.
Going back to 1997, Vanguard 500 Index finished no worse than the 37 th percentile in years in which the stock market overall went up. The fund finished in the 62 nd percentile in 2000 and the 44 th percentile of its Morningstar category during 2001 and 2002, the three years in which the market declined.
As we have noted, managers of non-index fund may hold significant amounts of cash, which helps when the stock market is declining, and they also may explore other market areas that are not necessarily part of their fund's mandate, a practice called style drift. Index funds do not deviate from their mandates, while actively-managed funds tend to have more leeway and will exercise that during certain market conditions.
Morningstar's Rekenthaler concluded: “Since indexes are pure plays on a particular asset class and (active) funds are not, the bottom line must be that, when indexes are compared to generally similar funds, the indexes are rewarded for their purity during good times and punished during bad times.”
But the theory expressed in Dunn's law and supported in Rekenthaler's study may not hold up when we look at results in markets that may be less efficient. Since 1997, the best years for U.S. small company stocks were 1997, 1999, and 2003. The S&P 600 Smallcap Index with the hypothetical 20bps expense ratio would have placed in the 42 nd , 68 th , and 71 st percentiles, respectively, during those years. The worst year for small caps during this time period was 2002. The hypothetical fund would have finished in the 33 rd percentile.
The best years for real estate equities were 1997, 2000, 2003 and 2004. Vanguard REIT Index finished in the following percentiles for the real estate fund category during those years: 88 th , 67 th , 52 nd , and, through October 2004, 49 th. During real estate's negative years, 1998 and 1999, it finished in the 55 th and 58 th percentiles, respectively.
Even if Dunn's law were valid, I'm not sure how useful it would be. It suggests that an investor should use index funds during periods when an asset class is doing well, and should use actively managed funds cycles when the class is doing poorly. So the investor hiring the money managers should do what the managers themselves are trying to do: Figure out whether the market (for a particular type of investment) is going to go up or down, and invest accordingly. It's easy, until you try to do it. More constructively, one can argue that because stocks tend to rise more than they fall, one should just stick with index funds.
In the end, performance tells the story. U.S. and European large-cap index funds and U.S. small cap index funds continue to provide long-term returns in the top half of their categories. Actively-managed funds in asset classes such as Japanese equities and real estate equities seem to provide greater returns. Not every market is equally efficient, and not every benchmark is an equally good index-type investment. So it makes sense to index some asset classes and actively manage others.
That's how we have been doing it for years at Palisades Hudson Asset Management. This review of the latest data reinforces our view that this is the way we should proceed.