Bill Miller is a well-known stockpicker. He is chairman of Legg Mason Capital Management and oversees some $35 billion. Although not as famous as Warren Buffet, his reputation is on par with Mr. Buffet when it comes to picking stocks. No less an authority than CNN Money proclaims him “the greatest money manager of our time.”
Bill Miller is most famous for beating the S&P500 index 15 years in a row with his mutual fund Legg Mason Value Trust (LMVTX).
During the streak there were several years that saw his fund underperforming the S&P500 leading into the final months or even the final weeks of the year. But somehow he always ended the year with a miraculous finish to beat the S&P500 and grasp victory from the jaws of defeat.
Nearing the end of 2006 he was in a familiar position: losing to the S&P500 with the year coming to a close. There was a lot of media attention on whether this was the end of the streak.
If memory serves me, he performed well at the end of the year. But the deficit was too much to overcome and the streak ended in 2006.
What Does Bill Miller Prove About Active vs. Passive Management?
(Reader warning: In the interest of space I’m about to make some generalizations.)
With the rise of index investing over the last 30 years there has been an ongoing debate about whether investors are better off choosing actively managed mutual funds or passively managed index funds.
Supporters of passive management have compiled an impressive body of evidence indicating that passive management is likely to outperform active management over short periods of time, and even more importantly it is highly likely to outperform active management over long periods of time.
Supporters of active management cite Bill Miller as proof that actively managed funds can outperform passive funds over long periods of time. After all, Bill Miller did just that. He beat the (mostly) passive S&P500 index for 15 years.
So who’s right? Does Bill Miller prove that active management can beat passive management?
Evidence is Different than Proof
First, it’s important to remember the difference between proof and evidence. Bill Miller might be evidence of some proposition, but not necessarily proof. We have evidence of a Big Bang, but we haven’t proved it. We have evidence of the Loch Ness Monster, but we haven’t proved it.
If someone tells you that Bill Miller is proof that active management works, the proper response is that he might be evidence, but he’s not proof.
Now that we all know the difference between evidence and proof, we can discuss whether Bill Miller is evidence that active management works.
(For ease of use I’m going to use a shortcut. When I say “does active management work” I’m really saying “can an active manager consistently outperform a passive approach over long periods of time.” Similarly, when I say, “is active management better than passive management” I’m really saying, “can an active manager consistently outperform a passive approach over long periods of time.”)
Point
Supporters of active management see Bill Miller as evidence that active management is better than passive management.
Counterpoint
However, supporters of passive management would rightfully argue that the existence of a single active manager that beats the S&P500 index for 15 straight years is a strong piece of evidence that active management doesn’t work.
Passive management supporters recognize that there may be an occasional “lucky” fund manager that beats the index over a long period of time. They would argue that with hundreds or thousands of fund managers in the world there ought to be more Bill Millers through simple random chance. The index doesn’t have to beat all managers over a given time period, it simply has to beat the vast majority. They would say the dearth of Bill Millers supports their argument that active management doesn’t work.
Point
Supporters of active management would argue that 15 years is simply too long for someone to beat the index by chance. He must have special insight that allows him to do it year after year.
Counterpoint
Supporters of passive management would respond that 15 years may not be sufficient time for the index to prove its superiority over all other investing approaches. (For some reason when I write the last sentence I picture Chris Farley saying, “All other tropical storms must bow before El Nino“).
There may be a lucky active manager that beats the index for 15 years. But continue for another few years and the lucky manager will eventually revert to the mean.
I Got to Thinking…
There’s no question that Bill Miller has handily beaten the S&P500 index for 15 years. But what has he been up to since the streak ended? Unfortunately not much. He has been beaten up the last couple years. If this were a prize fight he’d be down for the count.
As a supporter of passive management I thought to myself, “If the last couple years have pulled down Bill Miller’s average return to that of the index, then passive management has once again proven its superiority, Bill Miller gets his comeuppance, and order reigns in the universe.” After all, why should 15 years be the right testing period? Why not 17-and-a-half? Or twenty? Or more?
If the passive supporters are right, then the longer the testing period the more likely Mr. Miller should revert to the mean.
The Numbers
I gathered the numbers for SPY and for Bill Miller’s fund Legg Mason Value Trust. I chose SPY because it’s the most well-known index fund that tracks the S&P500 and because it has data stretching back through most of Mr. Miller’s streak.
SPY started trading in January 1993 so I started my comparison there. By that point Bill Miller had already beaten the index for a couple years.
I’m not going to describe my methodology in detail since most of my readers want the bottom line without the details. If you’re one of the chosen few that likes the details, post a comment and I can describe more in the comments.
From January 29, 1993 through the end of the streak in December 2005, Mr. Miller outperformed SPY quite spectacularly:
LMVTX: 13.7% annual return
SPY: 9.5% annual return
Miller’s numbers would look even better if I could include the first 2 years of the streak (SPY didn’t start until January 1993).
Slightly more of Miller’s returns came through dividends. Including the effect of taxes and taking into account the reinvestment risk would help close the gap a bit, but Miller still comes out far ahead.
By the way, if you’re thinking, “Big deal, it’s just a few percent difference” then repent now. A few percent is a big deal when compounded over several years.
Now let’s include the numbers for 2006, 2007, and 2008 through May 15 (from Jan 1993 through May 2008):
LMVTX: 10.74% annual return
SPY: 9.1% annual return
Including the last two-and-a-half years causes a slight decrease in SPY’s performance, but a staggering decrease in Miller’s performance. Miller seems to be reverting to the mean.
Finally, I calculated the performance for an investor investing in LMVTX vs. SPY assuming a 25% tax on dividends. SPY partially makes up the gap after-tax because a smaller share of its return comes from dividends. The results (from Jan 1993 through May 2008):
LMVTX: 10.0% annual after-tax return
SPY: 8.8% annual after-tax return
If you’re like me, you appreciate the tax deferral opportunities of a fund with a low dividend yield.
Final Thoughts
Bill Miller earned his keep over the last 15 years as he thoroughly trounced the S&P500. However, SPY is making up lost ground over the last 2.5 years. It hasn’t caught up to Miller yet. Only the future can tell if or when it will.
In my view the existence of a single Bill Miller is strong evidence that active management can’t consistently beat passive indexing over long periods of time. If active management worked (in other words, if active managers could consistently outperform the index) then there would be many Bill Millers that consistently beat the index year after year.
Think about it. The index has vanquished the vast majority of active managers over the years, but it doesn’t get any headline love from the media. But when an active manager beats the index for 15 straight years it makes headline news in the financial section of the New York Times. We’re still talking about it years later.
It ought to make you think twice before you pick an actively managed mutual fund.
Finally, if you believe that it’s just a matter of picking the right manager, you still face difficulty of identifying those managers ahead of time. It’s once thing to marvel at Bill Miller’s outstanding returns over the years, but it’s quite another thing to identify the next Bill Miller without the benefit of hindsight.
Post Script
The New York Times (registration may be required) has this interesting graph showing Miller reverting to the mean after jumping out to an early lead in the late 1990’s and early 2000’s.
Nice entry.
I’m not a fanatical proponent of active management, but I would offer a counter point.
Why is annual outperformance the metric of success? If you’re a value investor, for example, why would we expect an active manager to outperform over every twelve month interval? There are periods where the market is generally overvalued or the market is ignoring the values the manager sees. It doesn’t make sense to me that you’d expect to always beat the market within those arbitrary time intervals.
And indeed, if you look at the late nineties, a lot of value managers actually had to close up shop. They simply couldn’t keep investors during the bubble period when they were underperforming.
In contrast, Miller was actually more of a momentum investor in the late nineties and was killed during the crash. Yes he lost less than the S&P, but still performed horribly vs. many traditional value funds.
So, anyway, I’d like to see a similar analysis of something like Dodge & Cox stock fund or one of the other deep value managers who’ve been around for decades (Dodge and Cox started in 1930 or something, IIRC). I have no idea how many periods they underperformed the indexes, but it would be interesting to see how many rolling 10 or 15 year periods they underperformed.
And to your first point, what would the conditions be to prove either position? I think it’s rather impossible. Both sides can only point to evidence.
total_return:
I agree with you if you’re saying what I think you’re saying.
I use Bill Miller as an example because he is well-known and received a lot of press attention the last few years. Personally I would rather have a fund that beat Bill Miller over the 15 years even if it didn’t beat the index every year. I think that’s what you’re saying.
Outperformance over a long period of time is more important than annual outperformance, assuming you can keep invested during the entire period.
It’s path independent. Let’s suppose I have the choice between two investments. The first is the Bill Miller portfolio. It beats the index every year for 15 years and quadruples my money.
The second investment, on the other hand, loses money for 14 straight years, but in year 15 it produces such an amazing return that I end up with six times my money for the 15 year period.
I choose the second, assuming I can stay invested during the entire 15 years.
On the other point, you correctly said: “And to your first point, what would the conditions be to prove either position? I think it’s rather impossible. Both sides can only point to evidence.”
Exactly. Some people choose to draw conclusions based on a single piece of evidence. I call them “first graders.” After first grade the world gets more complicated and we have to learn to deal with conflicting facts.
It’s like a court of law. Both sides have certain facts that fall in their favor. They emphasize those facts to the jury. But neither side is supported by each and every piece of evidence.
That’s why the jury gets paid the big bucks to weigh the evidence and determine a winner. If it were easy work they would just pull anyone off the street to do it. Oh wait, they do.
Thanks for your well-thought out comment.
P.S. I looked at the Dodge and Cox fund and briefly compared it to the Bill Miller fund. In the short amount of time I had, it looks like Miller jumped out to a lead and led through early 1998 but Dodge and Cox has made up ground since then. Overall Miller is still ahead if you start in Jan. 1991.
Miller vs. Dodge and Cox
Miller’s Countrywide, Bear Stearns, home builders, etc., have been such a disaster that his last 2 1/2 years have wiped out all of the outperformance of his 15-year streak.
Miller is a monkey with a typewriter.
[…] Bill Miller vs. the S&P500 […]