October 19, 2009
THE FIRST STEP IS A GOOD PLACE TO START
Is the stock market overvalued? Wolfgang Münchau says it is in today's FT. He cites some persuasive evidence, based on analysis by smart people: Professor Robert Shiller and Andrew Smithers. The U.S. stock market is overvalued by more than a third, we're told.
Our own work suggests that caution looks increasingly prudent when it comes to risk exposures in asset allocation. But we're not sure. To be precise, we're not sure that a given quantitative profile of the market dispenses timely information about what returns will be in the immediate future. And neither does anyone else.
It's tempting to thinking otherwise, but the future is always unclear. The good news is that the ambiguity oscillates with degrees of vagueness. But that alone isn't enough. The challenge for investing is finding context and structure for managing asset allocation through thick and thin; through times when the future looks reasonably clear as well as during periods when the near term outlook for risk and return is murky.
The good news is that more than half a century of financial economics provides us with the tools and concepts for thinking clearly and productively about designing and managing asset allocation for the long haul, which arrives one day at a time. Different investors will come to different conclusions, but everyone should begin at a common point: the market portfolio.
By market portfolio we're talking of all the major asset classes weighted passively. This boils down to a global mix of stocks, bonds, REITs and commodities. Those four constitute the "major" asset classes in the sense that all are readily available at low cost through ETFs and index mutual funds or, if you're so inclined, actively managed funds. For most investors, these markets (and their various subgroups) constitute the lion's share of the investment choices, and so this is where much of the heavy lifting in the money game takes place.
Weighting all these asset classes by their respective market values gives us a robust benchmark to begin our analysis. Too many investors are convinced that the market decisions in the aggregate tell us nothing; but this is shortsighted. The market portfolio isn't a crystal ball, but neither is it chopped liver. We shouldn't be so naive to think the market portfolio constitutes a short cut to quick success, but neither should we dismiss the collective judgments of all the world's investors.
The first step is recognizing that the market portfolio, flawed though it is, is a valuable resource. Unfortunately, finding a good benchmark that represents all the world's investable assets isn’t easy. Indeed, much of the financial industry pays no attention to this benchmark. No wonder, then, that such a benchmark is generally unavailable, which is why we calculate our own homegrown version. Indeed, indexing a broad-minded definition of the market portfolio is at the core of our asset allocation work in The Beta Investment Report. And for good reason. Everyone needs a neutral benchmark as a starting point to consider the choices. We need to know how this benchmark has performed, and how its risk profile has changed over time. In short, we must study the market portfolio. We're not necessarily going to own it per se, but we need to recognize that in the very long term the market portfolio will deliver average returns and risk relative to the wide variety of money management strategies.
Much of the analysis in my newsletter centers on building a market portfolio index and using this benchmark as a guide for adjusting Mr. Market's asset allocation. As a general treatment, The Beta Investment Report maintains three model portfolio: high risk, medium risk and low risk. All three offer variations on the market portfolio, which is our proprietary Global Market Index. How do we manage these portfolios? It starts with estimating equilibrium returns and risk for the benchmark, a.k.a. the market portfolio.
Projecting the long run return for the market portfolio is essential as a building block for designing and managing asset allocation. The truth is that you'll go blind looking for analytics and data on this critical index. But while much of Wall Street is obsessed with trying to figure out where this or that asset class (or security) is going over the next month, The Beta Investment Report concentrates on forecasting the equilibrium return for the market portfolio. It's hard to overemphasize the power of routinely analyzing this benchmark. At the same time, it's probably the most under-utilized piece of analysis in all of money management. But if we have any hope of gaining strategic insight in the all-important business of asset allocation, we need to have a broad benchmark of the market portfolio and become intimate with its risk and return profile.
It's important to point out that finance theory advises against trying to forecast equilibrium returns directly. A more reliable approach is calculating implied returns by looking at volatility and correlations between the asset classes. In effect, we're reverse engineering the market's prospective return by studying its risk parameters, which offer more reliable insights compared with studying returns directly.
With expected equilibrium returns in hand, the real work begins. At this point, we can start to integrate our views about individual asset classes and whether they're likely to generate higher or lower returns relative to their long run equilibrium performance estimates. If we're fairly confident in our forecast, we'll change the weight of the asset class in our model portfolios up or down, depending on the forecast, relative to the market portfolio weight. But the pesky problem of always have doubts about the future keeps us from straying too far from the market portfolio's asset allocation.
For some investors who are highly confident in their outlook, the resulting portfolio will look radically different from the market portfolio. At the other extreme is an investor who has no particular view, which implies holding the market portfolio as offered. Our model portfolios in The Beta Investment Portfolio generally fall in the middle of these extremes, albeit with three variations of risk.
The real benefit of analyzing portfolio choices in this way provides valuable context and perspective for understanding our particular worldview. Simply going through the process of estimating equilibrium returns, and reflecting on what that implies for the market portfolio and individual asset classes, is an education of immense powerf.
Alas, too much of what passes as investment advice starts at the opposite end of the spectrum. It's tempting to dive into the debate about whether the stock market, or any other market, is overvalued or undervalued. But that courts disaster if the analysis lacks broader context for assessing risk and thinking about asset allocation.
There are no short cuts in designing and managing a portfolio strategy that satisfies in the long run. Fortunately, there's a productive starting point. The bad news is that too few investors are paying attention.
Posted by jp at October 19, 2009 9:53 AM
Yes, indeed, Grantham has done well. Perhaps he'll continue to do well. There are others who've also done well. There'll always be winners. We can also argue that some have won because they're smart--smarter than the rest of us. I assume there'll always be such gnomes--there always have been. And we should study their methods, comments. The opposite side of the coin exists as well. In the long run, the relevant beta is in the middle. The active winners extract their gains exclusively from the hide of the losers.
I'm not saying no one should ever take active risk. Some of us should. The question is how much? And if we agree that we should hold multiple asset classes, the question becomes: should we be active investors in everything? The more active risk we assume, the higher the hurdle.
Remember, too, that even smart investors get whacked at times, perhaps for unusually long periods. It's easy to look back at the late-90s and say: Sticking with value investing was the smart thing to do. But in real time back in those years, it wasn't so easy to rationalize.
Even if we use active managers to flesh out our asset allocation, we still need to manage the asset allocation. Doing so with active managers adds an additional layer of risk. It may pay off; it certainly has in some cases. But it's not a free lunch. People make mistakes, get into trouble, etc. In any case, you're still buying beta with active management, albeit with a twist.
There's no absolute right or wrong answer. Rather, there's just an array of risks to consider and evaluate. Everyone will come to their own decisions, based on their risk tolerance, investment horizon, financial situation and, to your point, confidence (or lack thereof) in alpha.
In short, it's all about risks. Choose wisely. But you still have to design and manage asset allocation. Even if you choose one equity manager, or own one stock, you're making an asset allocation design, albeit an extreme one.
Posted by: JP at October 23, 2009 9:46 AM
But to your point, Jim, you don't need to have many extreme periods in the markets in order to make money. GMO and Jeremy Grantham seem to have made a nice business about just waiting for the extremely fat pitch and making calls on those outliers. And then having the courage to act on it, just as you did to your credit in November through March.
So, perhaps it is "what your gig is" to use a phrase that I experienced firsthand (and you?)but other younger readers might not. I don't like having money invested in irrational, overvalued markets. And this is a doozie. I'd rather have it parked in bonds or in absolute return vehicles.
I don't think we have quite put out this fire though we may have cleared the smoke in the room.
But I think you make a good distinction with your last. There are other asset classes besides the SP500 and other ways to measure achieve returns besides strictly looking at valuations. And a sophisticated investor needs to find them or have his portfolio be down 45% in real terms from the 2000 peak as many are today (SP500).
Posted by: MarkM at October 23, 2009 6:16 AM
Good point. I have high respect for Shiller and Smithers, and their research, although I don't accept it without reservations or to the exclusion of other concepts/analysis. No one approach/theory is perfect.
That said, let's be fair: in recent years (notably just ahead of the 2000 correction and again ahead of 2008's debacle) both were issuing timely warnings, as were others of the value-oriented persuasion.
But there are some caveats as well. One is that relative extreme periods in the stock market lend themselves to accurate forecasts. Unfortunately such periods are rare. It's one thing to forecast a market correction at times when valuations appear historically high, or low. But the market is usually closer to so-called fair value, give or take, and so forecasting is more difficult, in terms of the degree of confidence one has for projecting return. No wonder, then, that the stock market can go for long periods of what appears to be overvalued/undervalued pricing.
For instance, in Smithers current book (Wall Street Revalued) his CAPE valuation analysis shows that the U.S. stock market was overvalued for about 10 years (i.e., the 1990s). And a long stretch of undervaluation dominated for an even longer period (early '70s to late-'80s).
All investing styles can and do suffer long stretches of discouraging results at times. The implication, IMHO, is that we need to look at more than valuation, however enlightened. All the more so if we're managing multiple asset classes.
Posted by: JP at October 22, 2009 10:16 AM
"To be precise, we're not sure that a given quantitative profile of the market dispenses timely information about what returns will be in the immediate future. And neither does anyone else."
Actually Jim, the two gentlemen you cite-Shiller and Smithers- would tell you that the q ratio and CAPE, the measures they use, tell you NOTHING about short term direction of the market. I think you know that.
Short term direction calls and moves are not investing. They are TRADING. But you know this too.
So you can't talk about "finance theory", "volatility and correlations" and the better mousetrap (BIR!)without at least acknowledging the absolutely dead-on 7-10 year projections that have been provided by these two tools. Not in my book.
Posted by: MarkM at October 22, 2009 6:13 AM