Monday, September 25, 2017

A critical reappraisal of "Margin of Safety"

A few years ago, I wrote a favorable review of Seth Klarman's Margin of Safety. I recently re-read the book, and today my opinion of it is far less charitable. There are three reasons for that:

1. It's derivative of Ben Graham
When I originally read Margin of Safety, I wasn't familiar with Graham's work. Since then I've read Security Analysis, which made me realize that MoS is essentially a carbon copy of SA. Some of the commentary in SA is specific to the stock market of the 1920s and 1930s, and some of the commentary in MoS is specific to the stock market and junk-bond boom of the 1980s, but beyond that Klarman lifted MoS's writing style, its layout, and even its title from SA.

The main difference between the two books is that SA is longer and more detailed. Anyone who's interested in Klarman's ideas can get a more sophisticated treatment of the same topics from Graham.

2. The writing is moralistic
Klarman is an old-school value investor who buys stocks that trade at low price-to-earnings and price-to-book ratios. There's nothing wrong with that--I prefer to buy statistically-cheap stocks too--but he's dismissive of other investing strategies, and the dismissal has a distinct moral element. He presents his style of value investing as the one true way and portrays speculation as nothing more than a "greater-fool game." There's no acknowledgement that riskier stocks can have positive expected value. Klarman calls art and rare coins "rank speculations" and denigrates short-term traders. The moralistic tone is grating.

When I started investing in 2002, I was a regular reader of Bill Fleckenstein's column on MSN Money. Fleckenstein was a permabear and dogmatically argued that the stock market was overvalued. Despite that, he had great trading instincts. On several occasions, he wrote that the market was due for a bounce right before it did, in fact, bounce. But as far as I know, he never acted on these instincts. He seemed to think that buying high-P/E tech stocks was immoral and that the only legitimate way to trade them was to bet on prices falling back to fair value.

I'm sure that attitude kept his returns far lower than they could've been, and for me, his column was a powerful argument against having ideological biases in investing. MoS evinces the same kind of rigid ideological thinking.

3. Many of the case studies were value traps
Klarman illustrates his investing philosophy with short case studies of stocks that he bought during the 1980s and early 1990s. The case studies are too brief to give a sense of why these investments were successful, but they piqued my interest, and I looked online for more information about the stocks in question.

What I found suggests that many of them had major risks that MoS doesn't mention, while others were classic value traps.

Bucyrus was the recipient of a leveraged-buyout offer in July 1987. After the October '87 crash, investors worried that the offer would be withdrawn, and the stock traded at $10 versus the offer's value of $14 in cash and new securities. At the time, Bucyrus had $9 per share of net cash and an "unprofitable but asset-rich mining-machinery business," and Klarman reasoned that the stock was "a real bargain."

The buyout was consummated, and the stock might have been an intelligent gamble on the deal going through, but the net cash provided less of a margin of safety than one might assume. Sales of Bucyrus's mining equipment had surged in the 1970s, but this gave way to an extended downturn in the 1980s and '90s as demand fell and coal miners reused surplus machinery from the '70s rather than buy new equipment. The company finally filed for a prepackaged bankruptcy in 1994 after years of operating losses. Its 10-K for that year states that "meaningful new machine shipments to domestic coal customers cannot be expected until after the mid 1990's."

City Investing Liquidating Trust
City Investing Company decided to liquidate in 1984, and the following year it put a hodgepodge of assets that it couldn't sell into a liquidating trust. The trust's units traded at $3 compared to tangible book value of $5 and distributions of $9 per unit over the following six years. Sounds great, except that much of the trust's return came from distributing its ownership interest in General Development Corp, a Florida land developer whose stock surged while the trust was liquidating but later became worthless.

GDC was a glorified Ponzi scheme. It sold plots of land on installment, promising that it would improve the land once the buyer had finished paying off the installment loan. GDC underpriced the land and lost money on every sale, but as long as the volume of new installment contracts exceeded the volume of maturing contracts, the company was cash-flow positive. When the U.S. entered a recession in 1990, the number of new installment contracts plummeted and GDC experienced a cash crunch. It filed for bankruptcy in 1991.

An investor who recognized the flawed nature of GDC's business could have hedged the trust's interest in GDC by shorting its stock proportionally, but in doing so, the investor would have tied up more money for a much lower profit.

Esco Electronics
Emerson Electric spun off Esco, its military electronics unit, in 1990 when the outlook for defense spending was bleak. After the spinoff, Esco traded at less than 15% of tangible book value even though it was profitable. Over the next five years, the stock doubled, which would normally be a good return, except that every other defense stock performed better. Northrop Grumman and Alliant Techsystems quadrupled, while General Dynamics was up 550%.

I'm sorry if it seems churlish to criticize Klarman for recommending a stock that doubled, but Esco's underperformance despite the astoundingly low valuation suggests the company had problems he doesn't acknowledge.

Intertan operated RadioShack's international stores. In 1986 its stock traded at $11, which was less than its tangible book value and equal to its net working capital. The stock was cheap because losses in Europe were offsetting healthy profits in Australia and Canada. Klarman mentions that the company turned around its European division and that by 1989 the stock was trading above $60.

Unfortunately that was the high point, and Intertan spent most of the 1990s trading below $10. The company never paid a dividend. Circuit City eventually acquired Intertan for $14 per share in 2004, giving shareholders a total return below the inflation rate.

Obviously I don't recommend the book, but if you're still interested in reading it, here's a PDF copy.


  1. MoS seems to be an interesting artifact of the value universe memeplex. It's name is invoked by many as a holy text when few of them have read it and fewer still, as you have found, actually derived value from it.

    The ideological confusion in investing is an interesting observation. I used to work for some people who saw their ideological bent as one of the value-adds of their investment process. It was in fact why I wanted to work for them. Later I realized that their ideological edge was the thing getting in the way of them taking advantage of the field of diamonds that existed when we were together. Today they're still bemoaning the ideological sandpit they have found themselves mired in for almost ten years while other investors have hauled away most of the diamonds. Sadly, I was not one of them because I only figured this out recently.

    Today I watch other friends and otherwise smart, talented investors waste their talents with similar ideological biases. Of course, in a raging, inflationary bull market like this one one of the biases that will be punished most severely is the "dogmatic permabear", but I think you're correct in observing generally that a true investor doesn't bring morality into his profession. Those who do are also speculators, but of a different sort.

    Great work following up on the case studies in a "Where are they now?" fashion. No wonder we hear so little of this "secretive" multi-billion dollar value fund.

    1. Thanks. I've also been in the position of having to un-learn a lot of ideological baggage. Unfortunately it's probably something that every investor has to go through to some extent or another.

  2. I agree with your criticism but not with your review. Almost all books on fundamental investing steal from Graham in fact you could argue that Eugene Fama is a plagiarist for "discovering" the value premium. The way I look at it is at least Klarman is cites Graham and Security Analysis. Number 2: Moralistic comments I have no disagreement with.

    In response to number 3 I would like to use your own criticism against yourself and say that you should have less moralistic view of Klarman's investing style. Also holding someone's feet to the fire for some investment they bought and sold 10-20 years ago is kind of preposterous.

    1. Sure, but there's a difference between owing an intellectual debt to Graham and blatantly rehashing him. Most people who borrow from Graham add their own twist to his style, e.g. Buffett starting out as his disciple and eventually developing more of an emphasis on business quality. Klarman doesn't add anything new IMO.

      I don't think it's preposterous to critique his case studies. These aren't random stocks; they're hand-picked examples of his investing philosophy. If a significant # of them turn out to be duds, that tells us something about the viability of his strategy.

    2. Seth Klarman is more risk arbitrage-y in writing than Graham was in his writing. Graham dedicates like 15 pages total to risk arbitrage.

      My meaning is that he likely sold them several years before the downfall that you mention. If I bought a stock stock at $10 and sold it at $20 six months later you would apparently call that a terrible investment if in 10 years the company went bankrupt.

    3. If you flipped a stock for a quick profit and it later went bankrupt, I would think it fair to question whether the profit was the result of a repeatable process or simply a lucky fluke. Which is what I did with his case studies. In the case of e.g. Intertan, the fifteen years of crummy performance after 1989 suggests it was a fluke.

      Re: holding periods, Klarman owned Esco for at least five years. It's listed under "largest losses" in his 1995 shareholder letter.

    4. "If you flipped a stock for a quick profit and it later went bankrupt, I would think it fair to question whether the profit was the result of a repeatable process or simply a lucky fluke." If you're wondering if his philosophy and strategy were a fluke or repeatable, I'd just point out Klarman has a multi-decade track record you can also use to evaluate this. And when you do that, it definitely appears repeatable.

      The two questions should be how much more efficient have markets become since the book was originally published and how do its lessons apply to today's markets?

  3. Salutations! You are piercing the cult of personality surrounding these false idols, and the world owes you a debt of gratitude. If you meet the Buddha on the road, kill him! But what do you think lies on the other side? Investment nirvana?

  4. It's worth mentioning, though it doesn't obviate your point in anyway, that Graham isn't really responsible for a lot of the stuff he gets credited for. Value, MoS, etc were all around before him; you just had to read widely.

    Graham is rightly credited for synthesising these ideas into an, ostensibly at least, coherent framework.

    I think your points about Klarman get at something deeper. The way in which investment philosophy itself herds an investor.

    1. If you haven't read it already, you might enjoy "The White Sharks of Wall Street" by Diana Henriques. It profiles Thomas Mellon Evans, who started his career buying net-nets in the early '30s. As you say, the ideas were floating around, even if Graham was the first to compile them and provide a theoretical framework.

      Fascinating point about investment philosophies and herding. I'll have to chew it over, but I think you're onto something.

  5. I really appreciated the contrarian post. I have to admit I am one of those that hold this book in high regard. My appreciation stems more from the context provided surrounding the prevailing investment environment during the period. As a younger reader, I found the content invaluable and somewhat timeless. He rails on things that we are railing against today including the rise of passive investing and the reach for yield investors driving all sorts of distortions (debatable, I know). Plus I found his 3 category (if I recall) investment process rather interesting.

    Overall, one of my more favorite qualitative investment books though I may be biased as I have also read all his mid to late 90's LP letters. These must strike to the core of any current active value manager to unless you are just batting 1000 these days. Really appreciate his contributions to the community.

    1. Thanks. That's a fair point to make in the book's defense.

      I had a similar experience with Peter Lynch's books--I really enjoyed them when I'd just started investing, then gradually became more critical of some of his ideas, but I still appreciate the books for giving me a few evergreen investing principles early on.


Note: Only a member of this blog may post a comment.