©Mike Thompson, 2003 Copley News Service. All rights reserved

That depends on the direction of the movement, doesn’t it? A move up from here certainly would be nice.
Please keep in mind that I am focusing here on the truly long run—and not expressing any views about what might happen in the short run. After all, rallies of 30% and even more are common in secular bear markets—Japan has had something like 9 rallies greater than 25% since 1990, and 3 that were greater than 40%, while the U.S. experienced 13 bull markets of greater than 30% during its secular bears of 1902-21, 1929-49, and 1966-82. So ironically, stocks may well be the place to be in the short run. As I said, there are lots of periods where you get big rallies, even in major bear markets or in stationary markets. But for now, equities aren’t the best place to be in the long run.

What has changed so much today that we can’t rely on historical experience?
Consider just a few powerful facts: This first isn’t economic, but I believe it is a dominant consideration in all decisions, not just in investing. Specifically, the mainland U.S. is no longer an impregnable island fortress in the world. We are now—and are likely to be in the indefinite future—vulnerable to physical attack. Which means that the risks in decisions of all sorts have been increased by a large order of magnitude—especially risks framed by expectations over the long run. 

No argument the world’s a mess. But as you said, there’s always uncertainty. The last century was no paragon of peace.
I hate the euphemism, “the geopolitical situation.” But it seems to be such that the threats are going to be constant. And these risks are entirely different than what would exist in a peaceful world. You can have an economic mess but still feel things will get back to rights. This, on the other hand, scares me plenty. But in economic and market terms, what’s different today is that the expected equity-bond risk premium is historically quite small. 

Again, for the jargon-challenged, when you say the “risk premium” is small, what you mean is that at current prices, the long-run expected return advantage of stocks over government bonds is tiny. And you’re suggesting that is so different from historical experience that stocks aren’t the asset class of choice “for the long term” right now?
Well, Jeremy Siegel is right that the long run (in 20-year periods) real annual rate of return on equities in U.S. history has been 7% with remarkable consistency. Ever since 1800. It’s a very powerful story. There’s no way that you can match that anywhere. So he says what you should do here is just shut your eyes and hang in. 

I hear a “but” coming—
But the average dividend yield during all those 20-year periods that Jeremy looked at was over 4%. Real price appreciation contributed only 2.1% to that long-run 7% real annual return. All the rest was dividends, received and reinvested. By contrast, today’s dividend yield is in the neighborhood of 2%. Which means that in order to achieve 7% real growth over the next 20 years, we’d need to add 5% real growth in earnings to that 2% dividend yield-and that’s not exactly a reasonable expectation over the long run. Impossible, in fact.

But 5% doesn’t sound outrageous. After all, we had years at the peak of the mania when the indices turned in fancy double-digit gains.
There, you’re getting into the difference between what investors expect and what’s rational to expect. Remember, it’s a very important but little-known fact that real growth in earnings and dividends consistently lags long-run growth rates in real GDP—and in many cases even in per capita GDP growth—not just in the U.S., but in all other developed economies. Between 1900 and 2001, for instance, U.S. GDP growth averaged 3.3% in real terms, vs. 1.9% growth in GDP per capita, 1.5% earnings growth and just 1.1% dividend growth. And the U.S. economy was the most successful on the planet! 

But couldn’t valuations simply head higher?
You’re welcome to bet on further declines in dividend yields or on higher P/Es from current levels over the long haul—but it’s not a risk I’d want to take under current circumstances. For one thing, bonds do not need high starting yields to outperform stocks if inflation is lower than expected. For another, the historical average returns that so many rely on as guides to the future are misleading. In a nutshell, the double-digit returns stocks were able to generate over the last century were due to equities starting cheap and getting richer over time. Many investors extrapolate this past performance and expect (at least) as high future returns. But these investors are missing, first, the fact that a part of those realized returns was unexpected windfalls from rising equity valuation multiples and, second, that when starting from high valuation levels it is not reasonable to expect returns as high as in the past. I won’t cite you chapter and verse, but there have been four excellent articles published recently talking about how small the expected risk premium is in today’s market—and I strongly recommend every money manager study them all: Arnott-Bernstein in the March/April ’02 Financial Analysts Journal [www.aimrpubs.org], the pieces by Illmanen and by Hunt and Hoisington in the latest Journal of Portfolio Management [www.iijpm.com] and a paper by Kennedy and Fraser published last year by Cambridge Associates [www.cambridgeassociates.com]. The underlying message in all of them is that until these valuation issues adjust themselves, equities are not necessarily going to be the best place in the long run. 

Why not?
At bottom, what the research shows is that starting price matters. As Antti Illmanen puts it, the objectively feasible long-run equity return is the sum of the dividend yield and the long-run earnings growth rate—and when it’s high, that’s bullish for equities. By contrast, when the subjective investor expectations about returns are excessively high, “that’s bearish for equities, because high hopes make future disappointment more likely.” By extending the timeframe in which they looked at the history of the S&P index back to its origin in 1871, instead of looking merely at more conveniently accessible periods, Hunt and Hoisington demonstrate that stock and bond performance have been significantly affected by three important variables: the inflation rate, the dividend yield on stocks relative to the yield on government bonds and the P/E ratio on stocks. And that the performance of stocks relative to bonds is enhanced when inflation is high and penalized when inflation is low. 

Can you cite some numbers?
According to Hunt and Hoisington, the best 20-year period for stocks was from 1941 to 1961, when stocks outperformed bonds by 14.9 percentage points compounded per year. At that starting point, in 1941, the dividend yield was 7% while the bond yield was only 2.1%—a huge difference. The P/E ratio was 8.8, well below the 131-year average, and GDP deflator rose by an average of 3.6% during the span—an inflation rate that wasn’t the highest for any 20-year stretch, but was above the 131-year average. In other words, for the 20 years that equities did the best, the starting risk premium was the highest, the dividend yield was high, the bond yield was low and inflation was pretty high—a description that in no way describes current conditions.

Well, bond yields are pretty low.
Nominal bond yields, yes. But the research also shows that nominal bond yields, by themselves, tell you very little. Treasury bonds outperformed equities between 1874 and 1894, for instance, even though bond yields started that period at a seemingly low 3.4%. And even though the dividend yield was almost 4 percentage points higher than the Treasury yield. What happened is that the deflation took control, pushing the nominal bond yield even lower. Between 1874 and 1894, prices fell by 1.6% a year. So bonds do not need high starting yields to outperform stocks if inflation comes in lower than expected.

Okay, but those were extreme cases. And, granted, P/E is aren’t low, while dividend yields are. So maybe stocks won’t leave bonds in the dust over the next 100 years like they did in the last. But so what? Aren’t they still likely to be at least a little better?
That’s the crux of my whole argument. The audience I was speaking to in Phoenix, mostly foundation and endowment managers, actually has something in common with most retail investors: they are spenders. They can’t simply lock money up for the long run and forget about it. Spent dividends can’t be reinvested and principal liquidations can be fatal without a long-term bull market working in your favor. Yet when current cash inflow falls below spending requirements, the liquidation of principal is unavoidable. Which means that volatility matters, big time. The theory is that a portfolio can absorb volatility “over the long run” when you hold an asset with a long-run return high enough to compensate for the short-run risks of selling in temporarily depressed markets. It’s a fine theory. But what happens if there is no such asset in the universe? Or none in which you’d dare to risk over 50% of your portfolio?

You’re suggesting you’re reduced to eating your seed corn? 
That’s why the traditional institutional approach, “I will structure my portfolio in this way and make variations on the theme,” won’t work. So what I’m suggesting is, throw it away. You have to be much more unstructured, opportunistic and ad hoc than you have been in the past.

No wonder the institutional types in your audience were shaken. You’re suggesting they behave in most un-institution-like fashion. 
I’ve been to the mat on this with a couple of them over the last couple of months. Yes. The immediate response is questions. “Then how do you deal with your board? How do you deal with your boss?” Very difficult questions like that. But the fact is that institutional investors are already stirring in the direction of realizing that a portfolio more than 50% in equity assets over the long run is much to risky from here. Look at the tremendous growth in hedge funds and other alternative asset classes. And remember, too, that I am not saying the investment landscape has changed forever. The markets are dynamic, and a time will come when dividend yields are back near 4%, or dividend payments will have doubled from current levels, or bond yields will have fallen into the 3% area. Maybe even several of those things will happen. Then we’ll be able to revert to comfortable old habits. But not til then.

You really were trying to win a skunk at a garden party award, weren’t you? It sounds suspiciously like you were urging institutional types to employ—gasp—market timing.
Yes, I am talking about that dirty word, market timing. But why has market timing been considered a dirty word?

That never has made sense to me.
It’s only been because if the stock market always bailed you out over the long run, you couldn’t run the risk of missing out on those long run goodies. 

I’ve always suspected that many investors have never grasped the real implications of “the long run”—despite Lord Keyne’s warning. 
Even more than that, what I am arguing is that if today’s extremely unusual valuation issues—the low risk premium—makes getting those long-term goodies an extremely low probability bet over, say, the next 10 years, then the risk of being out of the market—because it might go up—is much lower. Any upswing that you might miss is far more likely to be a short-term one, than a long-term structural opportunity. Believe me, I am not underestimating how difficult it would be for most institutional investment committees to embrace market timing at this juncture. TIFF’s December 31 commentary included a fictional, but delightful “transcript” of a meeting in which one very typical committee wrestles with the idea. It’s well worth reading. [www.tiff.org] But its key points are: 1) “Investors who rebalance their asset mixes in a disciplined if not mechanistic manner can gain a big edge in markets that display lots of short-term volatility but move neither sharply higher nor lower in a long term basis.” And 2) “Strategies that tend to perform suboptimally under certain market conditions can work surprisingly well under different conditions.”

I can just hear the howls of protest that very few investors have ever been successful at market timing over anything like the long haul. 
I admit that market timing isn’t easy. And I’m not suggesting that it’s the appropriate strategy for the very long term. Nevertheless, the old system wasn’t so easy, either. As I said in the speech, it really wasn’t easy to manage policy portfolios, when market fluctuations drove actual allocations far away from policy allocations. Then the question became, should you rebalance or shouldn’t you? Those were not easy decisions. And how easy was it to decide when to make changes in the long-run policy allocation? Or what changes to make? Should we be 60/40 or 65/35? Then we had to make room for non-traditional investments. The real estate and the hedge funds. That wasn’t easy, either. This job, portfolio management, is always a big challenge, unless you’re Rip Van Winkle. And in a way, I am describing a more interesting and more fun way to make investment decisions than the old way. 

Certainly less restrictive than living in a style box. 
Exactly. I’m not minimizing the challenges, merely suggesting that investors face up to new ones that may be no more difficult really than the old ones—and may be more fun to deal with. Traditionally, you’ve locked yourself into a pattern that you promised yourself you would adhere to, come what may. Being opportunistic gives you much more flexibility. You can wake up every morning saying, “Well, how does the world feel?” It’s a much more interesting way to live. But not at all easy.

No kidding. Because it also implies taking responsibility for your decisions instead of saying, “My mandate made me do it.”
Yes, but in this world you never, ever, ever know the future—and you didn’t know the future three years ago, either. So the question is how would you deal with what you think are the basic forces? I think that now means dealing with a seesaw. The articles I mentioned get to the risk premium question. 

Let’s go back to that burning issue for a lot of portfolio managers today. Just what is the appropriate risk premium? 
What it should be is a whole other question. But if the current risk premium is small, which I think most people agree it is now, then it’s smaller than it has been—and that’s enough to say. And so, given the greater volatility of the stock market versus the bond market, to then say “This [equities] is where I want to have most of my money,” is to make a very risky and hard-to-live-with strategic decision. Because you could go through 4-5 years when equities are earning less than bonds. 

Not a pleasant prospect.
Of course, you could also go through 4-5 years when they’re earning more. But the long run here is not necessarily going to bail you out. Or, even if it does, the margin by which equities will outperform could be too small to compensate for the volatility. The whole argument turns on the risk premium question. If the risk premium were 6 percentage points or 7 or something like that over bond yields, if you were looking for double-digit equity returns from here over the long run, then to hell with the volatility. Fine. But the margin is too small. I mean, I prepared that speech for an audience of endowment funds and charitable foundations. They’re all spenders. So the volatility really matters.

Only if they want to continue funding their operations!
Exactly. Since current income is less then they spend, they’re continuously liquidating principal—making volatility really matter. We’ve learned even in the pension fund world today that both bond and stock volatility matters because the funding margins are bad. So it looks to me like we’re in investment circumstances where you have to play it by ear and you have to play it without fixed positions.

Institutions? Play it by ear?
I don’t deny that it’s a difficult case to sell to a board, particularly in the foundation and endowment area, where the boards include a lot of people who aren’t professional investors. Or that it’s a difficult strategy to execute. But nothing in investing is easy to execute. It really isn’t. 

No, but the other thing people forgot during the 20-year bull market is that investing is all about taking risks to get rewards. You’ve got a whole generation of portfolio managers and investment boards who’ve convinced themselves that if they diversify, stick to a style, and hold on for the long term, they’re home free.
Don’t forget, “ buy on the dips!”

Well, that’s a given. Except that it hasn’t worked lately.
This is a different world. But my message is not necessarily gloom and doom. I mean, investing, well done, could earn you as much or more, maybe, than before. A different world doesn’t have to mean down the tubes. 

For supremely skilled money managers. And what makes this now a “ a different world” for investors—different from all the other rough patches the market has encountered in the last 100 years—is basically that dividends are so puny? 
Yes. But don’t forget too that the Ibbotsen data, which is what everybody hangs their hat on in terms of the long-term returns, includes not just dividends, but dividends reinvested. It’s the reinvestment, and compounding of those dividends that really makes the huge difference. If you’re compounding 4.5%, instead of 2%, over the long run, the difference is immense. 

At least dividends aren’t in quite as bad repute now as they were a year or two ago—
Yes, recent events suggest that maybe the dividend yield problem could be solved. The President’s proposal to eliminate the double taxation of dividends, with all its problems, maybe won’t ever become a reality, but it has changed the focus of attention. Barbara and I were at a client meeting two days ago and heard a Sanford Bernstein presentation that was much more hopeful than anything I am saying. They feel that dividend payouts are going to rise rapidly and restore the equity risk premium. Fine. Great. If they do, I’ll drop this argument and go back to the conventional approach. But I’m much less sure than they are. 

Leaving aside the political questions, which aren’t insignificant, there doesn’t seem to be a lot of enthusiasm in the corporate world for raising dividends—or, more to the point, much ability to do so. 
Well, earnings are very depressed now.

And corporate balance sheets are in terrible shape.
True, the payout ratio on reported earnings ratio is very high. But there are little straws in the wind, and I’ll mention two. First, Microsoft. I mean, it was only pennies. But that they felt that they had to pay a dividend; that they couldn’t just sit on the cash; that they had to begin to share it with the shareholders. That is a total change in viewpoint.

No argument.
The other thing happened a few weeks ago. Goodyear passed their dividend. In recent years, if a company passed its dividend, the stock went up, it all but got an award. Yet when Goodyear said it was passing its dividend, the stock took a 10%-15% hit. 

Just like in the old days.
Exactly. So the market is beginning to think about dividends in a more positive light. Whether President Bush’s proposal goes through or not, the focus of attention has begun to change. Here’s another important aspect of dividends: In the old days, no matter what the earnings were, we knew what the dividend was. It had a signaling quality that was important. Today, with all the skepticism about earnings—dividends are a tangible statement by management. “This is what we know we can earn in the long run, because we don’t expect to have to cut this dividend.” Dividends tell a story again. They’re not just about courtesy, but something more palpable to the shareholder. Bush, in his odd way, threw this thing into the marketplace and made people stop and think. I don’t know how long it takes to play itself out, but the trend is right. By contrast, my case for throwing out policy portfolios very much hangs on yields staying stubbornly low. 

Even the straws in the wind you’ve identified scarcely mean corporate dividend policies change overnight. 
No, that’s right. Though if the market senses the trend, it may matter. But the other thing that makes today’s market environment such a different world is something Antti Illmanen spends a lot of time exploring in that Journal of Portfolio Management piece I mentioned: The big difference today, between what are rationally expected returns and what are subjectively expected returns. If you ask most people on Wall Street about long-term earnings growth they’ll say it’s 9% or something like that—

Which is patently silly.
Really crazy. Since in the long run, per share earnings do not keep up with GDP growth, but lag. But that means that there’s this whole problem of deflating expectations. During the 1990s, certainly, and even going further back, a lot of what the last century was about in the stock market was a succession of positive surprises. Certainly since 1982. So if the 20th Century was the century of equities, maybe now we’re going to have a succession of negative surprises. There are a lot of other places we could have them, but earnings growth in particular looks vulnerable to negative surprises; there has to be a deflation of these expectations to get back to realistic valuations. But it takes time to unwind. Illmanen talks about how, except for the 1929 Crash and its aftermath, the last century was characterized by the growing recognition that everything was wonderful in the U.S. The thing is, we have looked at this history here—which has been wonderful—without any sense of how lucky we were. He points out that there were many other places in the world where it wasn’t a good idea to be in equities in the 20th Century. You can think about the future of this country in many different ways, but you can’t feel quite as secure about it as we did in the 1940s or the 1960s or even during the 1980s, when we were experiencing that wonderful turnaround from how badly we were doing in the 1970s. 

Aren’t you getting a little nostalgic? I’ve never gotten the idea that Americans felt terribly secure as the nation entered WWII, or in the Depression. We certainly didn’t during Vietnam or Watergate—
I really meant coming out of WWII. Not during the war. I was very much there and remember 1942 only too well. In fact, the way I felt then is the way I feel today. Are we in something that maybe we’re not going to win? A lot of people felt that way in 1942. If you look at the stock market, it made a major long run bottom, just about the time we took Guadalcanal that summer. That we could really take something, even just a little island, was electrifying. Then in November, we landed in North Africa. And the long-term bull market was on its way. But the first 6-8 months after Pearl Harbor were grim. The Japanese were taking everything and the Germans were pouring across Russia. It was dismal. I sense a lot of that same mood now. I mean, How in the hell are we ever going to lead normal lives again? Breathe free, with the war and terrorism alerts. 

You mentioned that bringing people’s return expectations down to realistic levels could take a long time. And dividend yields likely will take a while to climb back to even average levels. Realistically, the way most institutions operate, even if they ultimately take your advice to abandon policy straightjackets, it’ll take them forever to implement that decision. Doesn’t that make it likely they’d do it just about the time they probably should be coming back the other way? 
That’s very possible. Amazing things go on in the institutional world. The great asset to be in, over the last 10 and 20 years, has been bonds, yet you see only 15%-18% allocations to bonds. Now, they’re playing around with the hedge funds and all those things as bond substitutes, but at much less risk, the bond play over the last 10 years was enormous—and kept going even after the stock market went down. This is part of the reason that I suggest that they just blot out whatever allocations they had in the past and try to start fresh from today. And remember, I’m not suggesting they say, “I’ll never touch the stock market again.” That would be wrong. I am suggesting that they just can’t put stocks away and forget them. That they have to be much more ad hoc and flexible in their asset allocation decisions. 

Certainly you must recognize that notion makes professionals accustomed to running their portfolios—not to mention their lives—according to quantitative benchmarks extremely nervous—
But you can do the market timing with a lot of quantitative stuff, tactical asset allocation. Wells Fargo, First Quadrant, they’re all doing tactical asset allocation on a quantitative basis. If I were going into that, that’s how I’d do it. But there are methodologies, too; you don’t have to fly by the seat of your pants. 

Aren’t you really suggesting investors rediscover the sort of wide-ranging money management style you employed way back when at Bernstein-Macauley? 
Yes. I’m glad you reminded me. When I came into the business, we allocated between stocks and bonds for clients depending on their individual circumstances. This was 1951, the shadow of the Crash was still palpable. I had two associates who were a lot older than I who had been through the whole thing. So we managed conservative people with small stock exposures and aggressive people with larger stock exposures. But we had a lot of leeway. We didn’t have a policy portfolio framework in mind at all. Or the notion that in the long run, stocks are the place to be. Not at all. Because everyone’s experience was of the market fluctuating, without a trend. It was only well into the 1950s—really, after 1958, when the world didn’t come to an end when stock yields fell below bond yields and growth entered our vocabulary—that “stocks for the long run” began to acquire some—what’s the word?

Currency?
Yes. Then, of course, the whole notion got knocked into a cocked hat in the 1970s. After the bubble of 1968 blew apart. It was the victory over inflation in the early 1980’s that finally really changed things—If you look at the data, we had constant positive surprises on inflation for nearly 20 years. All during the 1980s and well into the 1990s, the inflation rate kept coming in lower than expected. Even in 1999-2000, with 4% unemployment, the inflation rate was still low. Now, what inflation does is reduce visibility, so it kills off the appetite for risk. When inflation is low, you feel that you know more about the future, and are much more willing to take risks. So this was a wonderful sequence of events to raise P/Es and make people increasingly hopeful—and justifiably so. Likewise, I think that many of the productivity gains in the 1990s that we hear so much about, particularly in areas like retailing, came in response to these disinflationary pressures. The constant pressure on the price level was downward, not upward—and that meant that you had to cut costs to stay alive.

But now we’re at the point where we might have too much a good thing—and find out, contrary to Mae West’s wisdom—that it’s not wonderful. 
Yes, if it reaches the point where companies are shedding employees who can’t find jobs. Previously, they could find jobs. And now prices are very different. Meanwhile, we’ve also had to change our perception of the world outside. Maybe Europe really is history in an economic sense. They can’t get their act together. Japan—maybe they are going to pull themselves out of their morass, eventually. But meanwhile the dynamic is really in the rest of Asia—and in China. We’re just beginning to sense what that can mean, both good and bad.

It certainly has deflationary implications in these parts. 
Yes, it’s not linked to high P/Es.

And multiple expansion was really the signal event in the market in the 1980s and 1990s. What earnings growth there was, was dwarfed by the inflation in P/Es. 
Well said. And that’s also critical to the case I’m making. If there is room for continued multiple growth, then the low dividend yields don’t matter. Then you’re going to get your risk premium. It could happen. But I don’t think you place a bet on that. 

You can’t forecast animal spirits? 
At no point can multiple change be extrapolated, either way. Because you don’t know what kind of mood change is going to come. On August 15, 1982, the market took off like a shot—and never looked back. Why, at that moment? A case could be made two years earlier or three years earlier for the same thing to happen, but it didn’t. You can’t forecast multiples. You can say, “This multiple looks low, it’s low historically, or it looks low to me, or maybe relative to other companies. But in a macro sense—particularly after you’ve had almost 20 years (or, if you go back to 1942, 60 years) of multiple expansion, you just can’t forecast it. I just keep coming back to the notion that in a world in which the longer-run expected returns on bonds and stocks appear to be very closely aligned, much of the institutional money management rulebook that’s been compiled over the last 30-50 years ought to go out the window. And things like tactical asset allocation, which once was much more popular than it is today, deserve a fresh look. 

Sure, if opportunities are smaller in scale, and more fleeting, they’ll go more than ever to the flexible and nimble. But that’s a profoundly disturbing message for portfolio managers who think the be-all and end-all of job security is hewing to a benchmark. 
I got to that point toward the end of my speech. I said that in this looser, more opportunistic environment I foresee the abandonment of the dreadful, depressing, defaulting process of putting managers into cubbyholes—large-cap growth, small-cap value and such foolishness —along with this stifling, stupid obsession with tracking error instead of absolute results and risks incurred. 

Did any of the PMs in the audience stand up and cheer?
There was some response. But good managers should be given as much latitude as possible. I think that if you have a nose and a sense of how to do research in the investment arena, if it works in one area, it should work in another. With a good manager, breadth is very important. Give them as much opportunity as possible. If they have skill, don’t compartmentalize it. 

But how do you know you have a skilled manager without benchmarks? How do you know whether he’s adding performance, or just being pulled along in the market’s slipstream?
To me, the bogey is not what somebody else is doing. It is what do I need for what I hope to achieve, or do in my retirement, or to fund my grants or to keep the faculty happy. I need a return of such and such and I can stand just about this much volatility. How much is this manager contributing toward that goal? The volatility is tricky because it can be very volatile, if it’s got a low covariance. But when I look at my portfolio, how much is this manager helping me to get where I’m going? My question isn’t whether they are beating the S&P or the Russell, but how am I doing? I happen to be rereading some Benjamin Graham from the 1950s, and one of the things he stressed is that the market is only interesting if it’s going up a lot and you want to sell. Or going down and you want to buy. But what other people are doing in the market is not relevant to what you’re doing. He emphasizes that over and over again. Once you get in the benchmark business, you’re asking the question differently.

Very differently.
“How am I doing, relative to how other people are doing, blah, blah. It’s not interesting. Selecting managers is difficult enough, so that if somebody is delivering for you and helping you get where you need to go and the relationship works, I don’t care where he ranks, because the ranks change. I can hear all the consultants hissing and sputtering. Maybe I’m wrong, but it’s going back to first principles, and that’s always a good idea. 

But without policies and benchmarks, how do you ride herd on rational investment decisions? 
My point is that has to be a constructive question to ask. And, in a good organization, should lead to interesting and useful answers. Look, the whole process around “I want to be in the stock market for the long run,” grew out of some original work, but then it turned out to be true—and eventually just became encrusted. The whole benchmark thing grew out of a piece Bill Sharpe did about indexing, using the S&P 500. It was useful to identify styles, but I still feel that somebody who’s good at one style should be good at another. At least, I would tell a good manager, “Don’t feel constrained.” If you have skill, let’s get as much out of it as we can.” Because styles do go out of fashion. As it is, the manager who is constrained huffs and puffs and says, “Well, I outperformed my bogey, even though we had very poor returns.” But so what? Both the bogey and that manager lost money! 

Ah, but they’ve avoided dreaded tracking error!
Which points to another reason the changes I’m suggesting won’t happen overnight. There is an encrusted—I’ll probably lose all my friends for saying this—people structure in this industry that keeps it going. Investment officers have to rethink their roles, which in many ways could be more interesting, but very different. The consultants also. The ratings agencies’ boxes—whole bureaucracies have grown up around them. They’d all be uprooted, if people followed my advice to think about the investment process in new ways. 

There you go again, telling institutional types to go against the grain. Take new kinds of risks because the old formulas won’t produce the sort of gains they need from here. What exactly do you have in mind?
It’s impossible to be specific at this point, but—besides opportunistic market timing—the options have to fall into the general category of nontraditional assets: hedge funds, venture capital, real estate, private equity, etc. The good news is that investors are becoming accustomed to portfolios more complex than just stocks, bonds and cash. And that adding nontraditionals diminishes covariances within portfolios, making the overall package less risky than it used to be.

But nontraditional assets also entail new risks—lack of transparency and fee schedules bordering on highway robbery, among them. 
Clearly, there are issues that need to be addressed–and will be—as nontraditional assets assume more prominent roles in investment portfolios, and those are a couple of the most obvious. I can also point to at least three other problems with non-traditional investments in institutional portfolios. But then again, nothing is easy in this business. The first is that hedge funds, private equity and such don’t have a “market return” in the way that stocks and bonds do. So manager choice is a much more critical variable—and the proven best managers often aren’t terribly eager to take on new money. A second is illiquidity. An institution that must regularly liquidate stock, bond and cash assets to finance current spending almost inevitably sees the percentage of illiquid assets in its portfolio constantly rise. This imposes a practical limit on how much the fund can commit to nontraditional assets as long as its current cash flow lags expenses. The third is volatility. Research shows that private equity returns, for instance, are so volatile that—even under highly optimistic assumptions—private equity would underperform a common stock portfolio in about one-third of the years in 5-10-year periods, and in about 25% of the years in a 20-year stretch. Which would be hard to live with, to say the least. 

No kidding. And you want institutions to jump into those arenas without the security blanket of all the benchmarks that have been developed to tell them how they’re doing versus their peers? You’re talking massive reputational risk— 
Yes, reputational risk is huge. But reputations shouldn’t be so hard to maintain. Nontraditional assets shouldn’t really pose a risk if you keep your nose clean. Besides, I really believe that investment committees and managers should welcome the opportunity to move away from the beaten path. Surely, flexibility, creativity and out-of-the-box thinking offers more rewards than just doing the same old boring thing forever. 

Okay, I’m sure all the money managers you know are much better than average. But what happens in your investment world without benchmarks to all the below-average managers? Or, more pointedly, how to I avoid having them run my money—into the ground?
What can I say, other than it wasn’t easy to make sure that didn’t happen under the old rules, too?

Isn’t that where due diligence and constant vigilance enter into the process? 
As I said, this isn’t an easy business. I did a piece a long time ago, called Where, Oh Where Are the .400 Hitters of Yesteryear? [www.aimrpubs.org] complaining that nobody hits .400 in baseball anymore. The defense has gotten too good. Likewise, even the best managers don’t beat the market by as much as they used to. I did a big statistical study of mutual funds and the pattern was very consistent. The interesting thing was that the bottom-performing people stayed on the bottom. And the good ones don’t beat the market by as much as they used to. Now, you’d think that there would have been something learned out of modern portfolio theory and all the clap-trap that we built into that process, but evidently not! 

Okay, but it is also true that the cemetery at the end of Wall Street is filled with the remains of managers who tried to time the market and failed. Very few have succeeded over anything but the relatively short term. 
Yes, and that’s why I would want to use some sort of quantitative asset allocation model, because I wouldn’t trust my own judgment—or at least, I’d like to use a model as an important input. Being disciplined about changing asset mixes is crucial, but the fact is that strategies that perform sub-optimally under certain market conditions can work surprisingly well in others. If you’re in a long run bull market, tactical asset allocation means getting out of stocks too soon all the time because they seem to be out of line. But if you have this kind of market (or even not such a long uptrend), then a tactical asset allocation strategy will work much better. It is essentially a contrary opinion strategy. And well-suited for a rather horizontal market.

It’s probably also like investing for the long term in the sense that it was a very good idea—until everyone in the world started investing that way—at which point it stopped working. 
I don’t know who said this first, but it’s a great truth about the long run: Investing for the long run works only as long as people don’t believe it. The Dow 36,000 crowd or whatever thought everybody ought to believe. And most people did believe it at that point. Which was why it couldn’t work. Likewise, until everybody agrees with me, market timing will be a good way to invest. But when everyone agrees with me, I’m going 100% in stocks!

Thanks, Peter. 

Reprinted with permission, copyright 2003, 
Kathryn M. Welling and Weeden & Co. LP

Illustration: Ann Field

"My point is that we've reached a funny position where the long run doesn't work. Where long-run evidence doesn't fit circumstances as they are today."

"What the research shows is starting price matters...For the 20 years that equities did the best, the starting risk premium was the highest, the dividend yield was high, the bond yield was low and inflation was pretty high—a description that in no way describes current conditions."