William J. Bernstein
The Nature of Risk
A simple question: Just what is the risk of a security? Forget balance-sheet ratios, statistical price behavior, or any other abstract measure. Put your feet up on the fireplace, lean back, take a swig of something nice, and contemplate, "What are the bad things that can happen to my corporate securities?" At base, there are four:
- The company will go under—bankruptcy risk.
- You won’t be able to trade the security—liquidity risk.
- The company does not go under, but earnings disappoint—earnings risk.
- Nothing at all befalls the company or its profits, except that its stock is savaged by animal market spirits—speculative risk.
The first two pertain to both stocks and bonds; with the low recovery rates inherent in modern corporate bankruptcies, the only difference is that you are out of pocket sooner with stocks than with bonds. The last two pertain mainly to stocks.
In terms of equity selection, liquidity risk can be dismissed for all but the smallest stocks, so we are left to consider the question: Just how well does the market price the other three risks?
In a rational world, distressed companies should have higher returns to compensate for their obviously greater risk. And the classical three-factor model, which uses price/book ratio (P/B) as a measure of distress, seems to bear this out. However, as even Fama and French will admit, P/B is not a particularly intuitive measure of distress. What happens when we look at more direct measures of distress? In a wonderful working paper, Professor John Campbell and his coauthors at Harvard/NBER used several sophisticated measures of company distress which correlated well with future bankruptcy. Not unexpectedly, companies with high distress had high size and value loadings.
It would be reasonable to assume then, since these companies were highly distressed and had high betas for known risk factors, that a portfolio of their stocks would also have high returns. Alas, no: High distress correlated negatively with return, with the spread between the highest- and lowest-risk portfolios being on the order of about 20% per year. To repeat, the companies at highest risk of bankruptcy had the lowest returns. In a typical bit of academic underspeak, the authors concluded "[the data suggest that] the equity market has not properly priced distress risk."
Outright bankruptcy is bad enough and frequent enough: Over the lifetime of the average investor, most companies will go under. The only hope of profit is to get one’s dividends out before it happens. Worse, however, is that all companies eventually stop growing, and when they do, their stock prices usually get hurt. What is breathtaking is just how quickly this usually occurs. In a famous study by Fuller, Huberts, and Levinson (Journal of Portfolio Management, Winter 1993), stocks were sorted by price-to-earnings ratio (PE). The most expensive quintile, as expected, demonstrated spectacular prior earnings growth. How long did this last after they achieved their lofty valuations? As a rule, above average earnings growth persisted for only six years before it reverted to the earnings growth of the rest of the market—about 5% per year. How much extra growth did these stocks demonstrate during this period? About 20%, total, over the whole period. In other words, if a company was selling at a PE of 60 in year zero, and its price did not change, at the end of six years it would still be selling at a PE of 50. But, of course, by that time, its price would have changed, and not for the better.
A most peculiar situation, this: the most expensive stocks are also the ones most likely to disappoint.
Are some stocks more susceptible to Keynes’ animal spirits? By now, you should be moving your lips. Berry and Dreman, in "Overreaction, Underreaction, and the Low P/E Effect," (Financial Analysts Journal, July/August 1995), demonstrated something that even the most casual market observers are aware of: when glamour stocks have negative earnings surprises, they are taken out and shot, but when value stocks disappoint, the damage is much less. And conversely, when glamour stocks have positive surprises, they do tolerably well, but when a dog surprises, it generally skyrockets. Can you spell Kmart? So once again, the equity markets do not seem to price a very real risk—speculative damage—terribly well.
Disturbing, to say the least. None of the biggest common-sense risks of owning equity are particularly well priced by the market —I’m not talking about artsy-fartsy balance-sheet ratios mind you, but real, definable risks, like the company going kerplunk, decimating its earnings, or simply finding itself at the wrong end of a lynching rope.
This fits all too well with what one sees in the financial media: an overriding obsession with earnings, but little concern about balance-sheet strength until just before rigor mortis sets in. No wonder growth is overpriced and company safety underpriced.
Why, then, aren’t money managers able to take advantage of these obvious inefficiencies? Several reasons:
"The Limits of Arbitrage." (Schleifer and Vishny, Journal of Finance, 1997). When realized returns are the highest, so are fund flows; unfortunately, this is also when expected returns are the lowest.
Tracking error. Even the most successful strategies have rough patches (think about the misery of thoughtful investors in the late 1990s) and, to paraphrase an apocryphal Keynes’ quote, the markets can remain irrational far, far, longer than you can keep your cushy fund manager billet.
Most fund managers cannot transact their way out of a paper bag. A strong balance sheet and/or value strategy entails turnover, and unless it’s done with a light touch, fees, spreads, and market impact will wind up swallowing any excess return and then some.
More attention probably needs to be paid to balance-sheet quality but, as the past decade has shown us, a cheap and simple price-to-book sort provides a pretty good way of capturing most of the above inefficiencies.
We’ve come a long way these seven decades since Ben Graham first emphasized the margin of safety in Security Analysis. Unfortunately, the wrong direction.
Copyright © 2005, 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.