Efficient Frontier
Efficient Frontiers Logo
William J. Bernstein

Margin of Safety in the Asset-Class Era

Earlier in the year, editorial duties led me to reread Benjamin Graham’s The Intelligent Investor. This lighter, more colloquial version of Graham’s massive classic, Security Analysis, was aimed at the small investor and had the added advantage of a more current perspective—the heretofore most recent (fourth) edition was written in 1970-1971 and published just before the old master died in 1976.

No concept is more associated with Graham than the "margin of safety": that extra cushion of value which minimizes the probability of underperforming the default option of the prudent, risk-averse investor—the high-grade corporate bond.

For starters, one needs to realize that Ben Graham’s definition of risk is couched only in the very long term. For him, what we today would consider risk—the day-to-day volatility of the market—connoted opportunity. Short-term turbulence was none other than the wise investor’s manic-depressive old friend Mr. Market, one day willing to take inflated shares off the investor’s hands, the next, offering to sell them back cheaply.

Graham’s analytic strength lay in the humanities, not mathematics. His intuitive grasp of history told him that from time to time, the economy went south in a big way, causing large numbers of companies to fail. "Fair-weather investor" was one of the most grievous insults "The Dean" could hurl. The true test of any strategy was how well it weathered storms:

. . . the margin of safety is counted on to protect the investor against loss or discomfiture in the event of some future decline in net income . . . most fair-weather investments, acquired at fair-weather prices, are destined to suffer when the horizon clouds over—and often sooner than that . . . The danger of growth-stock investing lies precisely here. For such favored issues the market has a tendency to set prices that will not be adequately protected by a conservative projection of future earnings.

Stock purchases were best made in the foulest of weather. For Graham, nirvana was a company with no debt selling at a market cap less than the amount of bonds it could safely issue, should it so desire. Nineteen seventy-one was definitely a fair-weather year, but even then, Graham was able to find one such gem—National Presto—selling for three times before-tax earnings.

Fortunately for Graham, he had never heard of Kahneman and Tversky or Thaler and Benzarti. He did not care (although he surely knew) that human beings intuitively experienced risk in the short term and that his description of risk was as feasible for most people as successful dieting or perfect communism.

Here’s how Graham’s margin of safety worked. First, he would not even look at a stock unless it possessed a clean balance sheet, with current assets more than twice current liabilities and earnings more than three or four times annual bond and preferred obligations. Next, earnings had to be reasonably stable and preferably growing. (Earnings growth was merely a nice freebie for Graham; realizing that it usually didn’t persist, he never paid up for it.) If such a stock could be found selling at eleven times earnings, then it offered the shareholder an earnings yield of 9%. If high-grade bonds yielded 4%, then the stock offered a 5% annual advantage over the bond. Over ten years, this annual cushion grew to 50%.

That 50% advantage provided as fool-proof a margin over the safe bond as any reasonable investor could demand in the world of stocks; to come out behind the bond, the already undervalued security would have to fall a great deal more. Graham allowed that this did happen often enough, but surely, if one could own a list of, say, twenty such issues, the odds of aggregate failure should approach zero.

The Intelligent Investor contains many small pleasures for the informed modern reader, and none is as great as Graham’s foreshadowing much of the modern finance literature, especially the efficient market hypothesis. Even the first editions of the book, written decades before Eugene Fama shook the world with his demonstration of the random nature of security prices, display an intuitive sense that it was nearly impossible for the average analyst to successfully pick stocks. The reason was simply that he was buying and selling mainly from other analysts. Graham was the first to realize that as analytic techniques improved, the competition got that much keener and the job only harder.

The first index funds, available only at the institutional level, were brought out just as Graham was scouring the galleys. If a list of twenty stocks lowered the chance of failure to acceptable levels, then surely the entire market would increase the margin of safety that much more. As Graham surveyed the opportunities facing the buyer of stocks in 1971, he was discouraged; it was impossible to put together an adequate list of stocks that met his criteria. He fell back on his usual advice to hold perhaps 50% of a portfolio in high-quality equity issues, sheepishly admitting that while such a list came nowhere near providing a margin of safety, perhaps times had changed and that the margin could no longer be obtained.

In the years since the publication of The Intelligent Investor, stocks became historically cheap two more times—in 1974 and 1982—each time selling at around eight times earnings, or at an earnings yield of about 12%. By Graham’s definition, on neither occasion did this provide an adequate margin of safety, i.e. at least 5% more than the yield of long-term corporate bonds. (Corporates yielded about 8% in 1974 and an astonishing 14% in 1982.) Yet, these were both excellent times to buy stocks, and in 1982, bonds as well.

Which brings us to the current date. By Graham’s reckoning, we’re well up the proverbial creek, with stock earnings yields more or less equal to the long-corporate rate of about 5%— no margin of safety at all. One can finesse Graham’s formulation by using expected stock returns, which at present are about 7% nominal (5% long-term earnings growth plus a generous 2% dividend assumption), which offers a 2% annual margin of safety.

Is 2% per year an adequate margin when the risk of Chapter 11 is spread over the entire market? Certainly a question that reasonable people might bruit about. I would answer in the affirmative, because I think that in the very long term, long bonds are more risky than stocks due to their exposure to future inflation. To eliminate the inflation risk to bonds, one has to shorten maturities down to a yield of about 2%-3%, by which point even Graham might admit that a proper margin is now present.

Clearly, though, it’s not 1971, let alone 1974, any more. We’ve traveled only part way back from the Wizard-of-Oz market of the late 90s.The real question facing today’s investor is whether we’ll ever fully return to Kansas.

To Efficient Frontier Homepage

Copyright © 2003, 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.