January 13, 2006
STRATEGIC THINKING--AVAILABLE IN A CONVENIENT 5-YEAR BITE SIZE
Five years is a long time. Long enough to shatter old myths, forge new ones, and remind investors that luck is no trivial factor. A little skill couldn't hurt either. In any case, we submit for your consideration the trailing five-year annualized returns for the major asset classes through yesterday, followed by some observations.
What jumps out at us is the performance dispersion, which is to say the robust variety of returns--ranging from more than 20% a year for REITs down to just over 1% a year for the S&P 500, a.k.a. large cap domestic stocks. The point being that there's been no shortage of opportunity and pitfall in the global capital markets in the last five years. There never is. The winners and losers keep changing, but you can always count on a broad assortment of returns over time for the major asset classes. The not-so-subtle implication: asset allocation still matters, and more than a little.
The mother of all strategic issues in money management all too often becomes subsumed in the chase for the next hot stock. But anyone interested in investment success as a durable notion beyond the end of the month should ignore the widespread obsession with short-term tactical.
The fact that asset allocation is no less relevant in the 21st century than in the preceding generations comes as no shock. Indeed, 2006 is the 20th anniversary of the seminal research from "Determinants of Portfolio Performance," the 1986 paper in the Financial Analysts Journal that first proclaimed the overwhelming influence of asset allocation on portfolios. Market timing and security selection, by comparison, were found to be relatively inconsequential factors regarding diversified portfolios over time.
The eternal question, of course, is what's an enlightened asset allocation going forward? A difficult query to answer, even for an enlightened student of money management. In turn, that suggests investors should spare no effort in coming up with an informed guess, or otherwise hiring someone to do the dirty work.
With that in mind, it's worth remembering that nothing goes up or down forever, at least when it comes to asset classes. Individual securities come and go, but asset classes are forever. That may be small comfort, but it's more than you can say for any particular company or bond. The Acme Computer Co. may hit the skids tomorrow, but equities overall will be here long after we've become fertilizer. Accordingly, paying attention to reversion to the mean (otherwise known as portfolio rebalancing) is one of the few worthwhile pursuits in the study of market history.
Timing, alas, is a crucial part of the equation, at least in the relatively short run. So while we can all congratulate ourselves in recognizing that the REIT group has been one of the globe's hottest asset classes in the last five years, or that the large-cap U.S. equity sector has been a dog, deciding what to do about it for the next five years is another matter.
Inching toward an answer starts with deciding what your "normal" weight in each of these assets should be. "Normal" here is defined as a long-term weighting a la a pension fund's infinite time horizon. For some, that means something approximating the current weight of the asset class as calculated by its relative market cap within the global capital markets. However you calculate it, coming up with a normal weight is fundamental. Why? Because if REITs, for instance, are deserving of a 10% normal weight, then the challenge at any given moment is deciding how much to deviate, if at all, from the normal weight, either with an overweight or underweight.
Coming up with intelligent answers takes time, a bit of number crunching, and in the long run some luck. But for the moment, we can start with step one: knowing where we've been, which hopefully can shed a tiny bit of light on where we're going.
Posted by jp at January 13, 2006 10:58 AM
Speaking as an investor with a medium-sized portfolio and average financial knowledge, this debate begs the question of whether a reasonably intelligent investor has the resources to safely pick individual stocks. My limited experience is that it is much more of a crap shoot, and requires far more time, then the reward justifies. There are too many unknowns that make picking individual stocks problematic for anyone without a staff to cull through all the potential data.
As an example, I bought Johnson & Johnson more than a year ago because it seemed like a well-run, safe health care stock without the dependence on individual drugs of Merck and Pfizer. The stock kind of slumped and then came the botched Guidant acquisition. The lesson to me is that I would either need to spend WAY more time evaluating stocks, or avoid trying. It seems to me that my time is much more profitably used trying to keep abreast of general trends.
There was an article in the WSJ the other day, Reports Help Uncover Red Flags. My feeling is that the people who have the expertise, time and access to data necessary to read these reports and then apply them to each individual company can probably beat an ETF covering the segment. That ain't me, and from my experience, it's not most mutual fund managers either.
Posted by: Individual Investor at January 21, 2006 9:11 AM
Fair enough. The question seems to boil down to this: assuming a globally diversified portfolio, is asset allocation
1) more important,
2) equally important,
3) or less important
than market timing and/or security selection in generating long-term results?
I choose 1, albeit based solely on my own brand of empirical observation.
Posted by: Jim Picerno at January 16, 2006 2:07 PM
Jim: I never said that asset allocation was "of no use" as you seem to suggest. I simply stated that its importance is overrated. You asked me to support my claim and I did.
As for your comment about the papers I cited, "more relevance for academics and modelers than for general investment strategy," we clearly disagree. It's the BHB work that doesn't hold up in practice.
By the way, in my "day job" I acutally manage portfolios.
Posted by: thc at January 16, 2006 11:29 AM
Ah, yes--the smoking guns.
Yes, THC, those are two we're familiar with, having first perused the two texts you cite years ago in our "day job" as a financial journalist. For any serious student of investment strategy, these two papers should be read and digested. But they should also be considered in context, which is to say: they don't refute the relevance of asset allocation. Rather, they somewhat diminish the prestige of the original Brinson, Hood & Beebower asset allocation paper. A subtle distinction, but an important one, at least for investors.
To be sure, both of the papers THC references reveal various "flaws" in the original asset allocation study. But the "flaws" tend to be of a technical nature, with more relevance for academics and modelers than for general investment strategy. That's because even these two critiques don't conclude that asset allocation is of no use. Quite the contrary. Some quotes from the two papers leveling the criticism should suffice as perspective:
From the Ibbotson & Kaplan paper: "We sought to answer the question: What part of fund performance is explained by asset allocation policy? If we think of this issue as a multiple-choice question with '40 percent,' '90 percent,' '100 percent,' and 'all of the above' as the choices, our analysis shows that asset allocation explains about 90 percent of the variability of a fund's returns over time but it explains only about 40 percent of the variation of returns among funds. Furthermore, on average across funds, asset allocation policy explains a little more than 100 percent of the level of returns. So, because the question can be interpreted in any or all of these ways, the answer is 'all of the above.'"
From the Jahnke paper: "There is little doubt that asset allocation is an important determinant of portfolio performance. However, such agreement does not settle the issue of how to do it. What are the appropriate asset classes? Should asset class weights be fixed or dynamic? How should asset allocation be determined? What about the cost of implementation?
Brinson, Hood and Beebower are to be commended for their work, upon which others will build."
Posted by: Jim Picerno at January 16, 2006 8:19 AM
Jahnke and Ibbotson both have done some very interesting work in this area.
Posted by: thc at January 15, 2006 10:34 PM
Flawed? Yes. Irrelevant? No. We could quibble about what percentage is reasonable. But it's a safe bet that for diversified portfolios, the asset allocation decision tops market timing and security selection. Emphasis on diversified portfolio. If you've only got six stocks, asset allocation doesn't matter. Of course, you've got bigger problems, but that's another story. In any case, if there are superior studies, don't taunt us--tell us. What are these studies?
Posted by: Jim Picerno at January 15, 2006 9:28 PM
There have been several credible studies in the last few years that would indicate that Brinson, Hood and Beebower's study was flawed and that asset allocation's role in performance is far less than the 94% they claim. Market timing and security selection really do matter.
Posted by: thc at January 15, 2006 9:05 PM
Posted by: Barry Ritholtz at January 13, 2006 12:29 PM