Reflections on Money Management:
Of Risk Premium Collection, Patience, and Spreading Your Bets

It is all too easy to confuse market efficiency and rationality. More than one skeptic has pointed out the inconsistency between occasionally "barking mad" markets (page xi of The Four Pillars of Investing) and the tenets of the efficient market hypothesis.

I indeed believe that markets are both highly efficient and, at times, highly irrational. The former refers to the extreme difficulty of selecting individual securities that will produce excess risk-adjusted returns, while the latter refers to the fact that risk premia occasionally get completely out of whack, as they did during the late nineties and the entire middle third of the twentieth century.

While I’m not going to reveal precisely how I’ve allocated assets in the past, it’s in the ballpark of the DFA Model Strategies—heavily weighted towards small cap, value, and foreign equities, keeping the bond portion of the portfolio relatively short. And, of course, as passive as possible.

I can tell you two other things:

1. Over the past few decades, the DFA Model portfolios have blown the doors off a conventional market-weighted approach—so much so that during the twenty-year period ending September 2003, only a 58/42 stock/bond mix (using the Lehman Brothers 1-5 Year Aggregate) was needed to equal the 11.52% return of the Wilshire 5000 in that same time period.
2. Any manager using such an approach, let alone one more highly torqued, will not long survive in the world of pension or private wealth management; even the most efficient strategies will underperform the S&P 500 for prolonged periods of time, as occurred in the late 1990s. Very few managers can outlive three or more years of double-digit tracking errors.

To demonstrate just how bad this tracking error can be, I’ve calculated below the returns of the following five all-equity portfolios over the past 20 years (ending September 2003, rebalanced every 12 months):

• Wilshire 5000

• Conventional: 70% Wilshire 5000, 30% EAFE
• DFA Model: Heavily weighted towards value and small
• More torqued: A "four-corners" approach of 25% each domestic and foreign large and small value
• Most heavily torqued of all: a portfolio consisting only of small value stocks—half foreign and half domestic

Their returns were 11.70%, 11.52%, 14.41%, 14.22%, and 14.05%, respectively. Clearly, exposure to the global value factor was of some benefit during the past 20 years, while exposure to the global small factor was not. (For the record, the 20-year Fama-French value factors were about 2% for domestic equity and 3% for international; both size factors were near zero, consistent with the larger magnitude and greater reliability of the value factor.)

While the more highly torqued series (the last two) did not outperform the more balanced DFA Model strategy, this is likely an artifact of the lack of an international small-value series before 1995. In any case, if one accepts the three-factor model, it is irrelevant; more highly torqued portfolios have higher expected returns, and their behavior can be used to demonstrate the unbearable pressure they bring to bear upon investors of all stripes.

Now for the "money shot." Below, I’ve plotted the one-year trailing tracking error, in increasing order of "torqued-ness" towards the value and small risk factors for the four all-equity portfolios (Conventional 70/30, DFA Model, Four Corners, and Two Corners) versus the pure Wilshire 5000.

What you’re looking at is a plot of just how badly you’re going to feel during periods when the S&P 500 is the toast of cocktail party conversation, as it was during the 1920s, early 1970s, and late 1990s.

If you’re an institutional investor, the damage will be significantly worse. Dipping much below the –5% level for more than a year or two is usually fatal. While you might have skated through the 1990s with the Conventional 70/30 portfolio, the DFA Model strategy would have made you a deadster by 1998, and the Four- and Two-Corners approaches would have sent you packing on at least three occasions. For example, in 1997 the return of the Wilshire 5000 was 31.28% versus 25.89% for the Conventional, 13.32% for the DFA Model, 8.26% for the Four-Corners, and 4.02% for the Two-Corners portfolios.

Even Berkshire shareholders would have tossed Warren and Charlie out on their ears with the relative underperformance of the last three portfolios.

If you desire a fuller appreciation of the constraints that hobble institutional investors, I cannot recommend highly enough Fortune and Folly: The Wealth and Power of Institutional Investing, a dead-eye dissection of the pension fund industry by two anthropologists, William O’Barr and John Conley. In the late 1980s, they donned the requisite field garb from Brooks Brothers and went forth to find out what makes this exotic breed tick.

The professors discovered they were dealing with not one, but two separate tribal cultures:

• Corporate pension fund managers. This group was usually "recruited" from within their companies. The quotation marks signify that a transfer to the company’s pension fund division was, and still is, generally not viewed as a promotion. Nevertheless, these organization men and women viewed themselves as "can-do" types whose bias was towards active managers. Further, since they were funding a corporate obligation, their investment results eventually fell to the company’s bottom line, as many owners of common stocks have found out to their recent chagrin. While losses consonant with those in the broad market are tolerable, underperforming the market was not.

• Public pension managers, that is, those managing the defined-benefit plans of public employees. These plans tended to be relatively small, understaffed operations, which is fortunate, since this mandated much more passive management than seen in private plans. The Prime Directive of the public plan manager was "Thou Shalt Not Get Thy Name in the Newspaper." This translated into not lagging the risk-adjusted broad market by too terribly much. Two to three percent behind? No problem. Five percent? Better tell your secretary to hold all calls from the Daily News. While the private fund manager may be rewarded for superior performance, no public manager in the history of the world has ever found herself celebrated on the 11 o’clock news for beating her benchmark.

What does this mean? While you won’t find many pension managers who haven’t heard of Fama and French, my passive-fund sources tell me that they nibble only ever-so-slightly at small stocks and do not tilt at all towards value. I’m not quite sure of the reason for this, but suspect it has to do with the small factor’s positive correlation with the market, whereas the value factor has a negative correlation, making the former a somewhat safer bet than the latter. The Frank Russell Company, one of the nation’s largest pension fund consultants, considers neither small nor value a risk factor, that is, neither deserving of a returns premium.

While the small investor who tilted towards value and small in the late 90s grew mightily discouraged, he or she could not be fired. In 2000, the worm turned with a vengeance—I suspect there will be a very high positive correlation between retirement age within the boomer cohort and portfolio-weighted price-to-book. This investor felt the sting of relative underperformance in the 1990s and considers the longer-term reward of staying his own personal course fair, but not overly generous, compensation for the pain of the Bubble Years.

Obviously, not everyone can, or even should, tilt. If you work for a value company or a small company, it’s probably not a good idea. Even if you "qualify" by virtue of temperament and employment, these risk factors aren’t a sure thing. That’s why they are called "risk premia."

But as an investor, you must, by definition, bear some degree of uncertainty and lay your money down. To value-load or not? To small-load or not? You must choose.

Copyright © 2004, William J. Bernstein. All rights reserved.

The right to download, store and/or output any material on this Web site is granted for viewing use only. Material may not be reproduced in any form without the express written permission of William J. Bernstein. Reproduction or editing by any means, mechanical or electronic, in whole or in part, without the express written permission of William J. Bernstein is strictly prohibited. Please read the disclaimer.