- Each year the SPIVA results are virtually the same — a large majority of active managers underperform their benchmarks.
- Another demonstration of the failure of active management to generate alpha is to look at the returns of the two leading advisors to pension plans, SEI and Russell.
- Academic studies have found that there is no persistence in pension plan performance beyond the randomly expected — past performance isn’t a reliable predictor of future performance.
Yesterday, we discussed the history and overview of the efficient market hypothesis. Today we'll look at some of the evidence on the efforts of mutual funds and pension plans to generate alpha.
Each year, Standard & Poor's publishes its Indices Versus Active Funds Scorecard, more commonly referred to as SPIVA, presenting the evidence on the performance of actively managed mutual funds relative to their appropriate index benchmark. And as sure as the sun rises in the east, every year the results are virtually the same - with the large majority of active managers underperforming. This is true whether we are talking about stock or bond funds.
A recent study by Vanguard's research department covering the period 1997-2011 found:
- Just 54 percent of the funds managed to even survive the full 15 years. 2,364 funds were either liquidated or merged into another fund in the same fund family, in some cases more than once.
- Investors had a 79 percent chance of picking a fund that underperformed, was liquidated, or had a life cycle that was too convoluted for them to disentangle. And this is based on pre-tax returns. Since the higher turnover rates of actively managed funds tends to make them more tax inefficient, on an after-tax basis the odds of success would have been even worse.
Vanguard's own research led them to conclude that the best predictor of future performance is expense ratios. Complicating matters is that the evidence from many studies shows that there's little to no evidence of persistence of outperformance beyond the randomly expected - past performance isn't prologue. For example, Brad Cornell, in his 2009 study "Luck, Skill, and Investment Performance," concluded: "The analysis also provides support for the view that annual rankings of fund performance provide almost no information regarding management skill."
Another demonstration of the failure of active management to generate alpha is to look at the returns of the two leading advisors to pension plans, SEI and Russell. For the past several years, I've presented the returns of their actively managed funds and compared them to the performance of DFA's funds in the same asset class. (Full disclosure, my firm Buckingham recommends Dimensional funds in constructing client portfolios.) And with very rare exception, each time DFA's funds have outperformed not only overall, but also in every single asset class. Here's a summary of the results for the period 2000-13:
- An equal weighted portfolio of DFA funds would have returned 7.6 percent. A similar SEI portfolio would have returned just 4.2 percent - an underperformance of 3.4 percent a year.
- An equal weighted portfolio of Russell funds would have returned 5.3 percent, underperforming by 1 percent a year an equal weighted portfolio of DFA funds that would have returned 6.3 percent.
Our last bit of evidence looks at Morningstar's performance ranking for DFA's funds. The data is for the 15-year period ending April 10, 2014.
Morningstar Percentile Ranking
DFA U.S. Large Value III (DFUVX)
DFA U.S. Small (DFSTX)
DFA U.S, Micro Cap (DFSCX)
DFA U.S. Small Value (DFSVX)
DFA Real Estate (DFREX)
DFA International Large (DFALX)
DFA International Value III (DFVIX)
DFA International Small (DFISX)
DFA International Small Value (DISVX)
DFA Emerging Markets II (DFEMX)
DFA Emerging Markets Value (DFEVX)
DFA Emerging Markets Small (DEMSX)
Average DFA Ranking
Over the 15-year period, DFA's funds outperformed 80 percent of the actively managed funds that survived the full period - outperforming the majority of actively managed funds in every single asset class. It's also worth noting that DISVX finished in the 1st percentile, and DEMSX finished in the 6th percentile, in supposedly the most inefficient of asset classes. And this outperformance occurred despite the fact that there's a large amount of survivorship bias in the data because Morningstar doesn't account for the funds that either did poorly and were sent to the mutual fund graveyard (where their returns are buried), or were merged (and their performance disappears). Accounting for the "dead" funds would improve the ranking of DFA's funds considerably.
It's also important to note that Morningstar's rankings are also based on pre-tax returns. Their higher turnover increases the hurdle active managers face in order to generate after-tax alpha. The hurdle is so high that Ted Aronson, of institutional investment manager AJO partners, stated:
"None of my clients are taxable. Because, once you introduce taxes… active management probably has an insurmountable hurdle. We have been asked to run taxable money - and declined. The costs of our active strategies are high enough without paying Uncle Sam."
Given that active strategies are generally less tax efficient, if the rankings were based on after-tax returns, DFA's ranking would likely have been quite a bit higher.
It's hard to conclude that the markets are inefficient in the only sense that matters (active funds can persistently outperform) when DFA's funds, which don't engage in stock selection or market timing activities, outperformed the vast majority of actively managed funds who were supposedly exploiting market inefficiencies (mispricings) - without even taking into account survivorship bias and taxes.
We now turn to the evidence on pension plan performance.
It seems logical to believe that if anyone could beat the market, it would be the pension plans of U.S. companies. Why is this a good assumption?
First, pension plans control large sums of money, giving them access to the best and brightest portfolio managers, each clamoring to manage the billions of dollars in these plans (and earn large fees). Pension plans can also invest with managers that most individuals don't have access to because they don't have sufficient assets to meet the minimums of these superstars.
Second, it isn't even remotely possible that they would hire a manager who didn't have a track record of outperforming their benchmarks.
Third, many of these pension plans, if not the majority, hire professional consultants such as Frank Russell, SEI, and Goldman Sachs to help them perform due diligence in interviewing, screening and ultimately selecting the very best.
Fourth, as individuals, it's rare that we have the luxury of being able to personally interview money managers and perform as thorough a due diligence as do these consultants.
Fifth, the fees they pay for active management are much lower than the fees individual investors pay.
With those thoughts in mind, we'll examine the results of two large studies on the performance of pension plans. The 2008 study "The Performance of U.S. Pension Plans" and the 2005 study "The Selection and Termination of Investment Management Firms by Plan Sponsors," both found that despite plan sponsors hiring investment managers with large positive excess returns up to three years prior to hiring, realized returns relative to benchmarks were indistinguishable from zero. They also found there was no persistence in pension plan performance beyond the randomly expected - past performance isn't a reliable predictor of future performance. The authors concluded: "The striking similarities in performance patterns over time makes skill differences highly unlikely."
For those interested in learning more about the evidence on efforts to outperform risk-adjusted benchmarks, my 2011 book, "The Quest for Alpha" presents the evidence on efforts to generate alpha by individual investors, mutual funds, pension plans, hedge funds, and venture capitalists. You'll find that in each case the evidence demonstrates that the vast majority of investors would be better served to act as if the markets are efficient (invest in passively managed funds such as index funds).
Before concluding this section, there's one last point to cover.
It's a Stock Picker's Market
A common excuse heard since 2008 is that the increased correlation (a measure of the strength of the linear relationship between two variables) of stocks has made it difficult for active managers to outperform. Does this hold up to scrutiny?
Correlation shows the directional movement of stocks, not the magnitude of their movement. Magnitude is shown by the dispersion of returns - the size of differences in the returns of individual stocks/asset classes. The greater the dispersion, the greater the opportunity for active management to add value by overweighting the winners and avoiding the losers. Thus, it's the dispersion of returns that we should look at, not the correlations, to see how high a hurdle there is for active management.
In their May 2012 paper, "Dispersion! Not Correlation!" Vanguard Research showed that while correlations of stocks had increased, the dispersion of returns had actually remained stable.
In each of the five calendar years (2007-2011), which included bull, bear, and flat markets, the degree of dispersion was such that at least two-thirds of all stocks either led or trailed the index by more than 10 percentage points - with the range being a low of 67 percent and a high of 79 percent. Clearly, there was plenty of opportunity for active managers to add value. They just don't do it with any persistence, as the annual SPIVA demonstrates. As further evidence, a December 2013 study by S&P, "Dispersion: Measuring Market Opportunity," looked at this issue and found that while increased dispersion increases the opportunity for active managers to distinguish themselves, especially in relative terms, the evidence shows that higher dispersion doesn't increase the likelihood of outperformance by active managers.
Tomorrow we'll continue the series with a look at how markets become efficient and then summarize what we've discussed.