When Genius Failed: Revisiting Roger Lowenstein’s Classic Twenty Years Later

Roger Lowenstein’s When Genius Failed: The Rise and Fall of Long-Term Capital Management was first published in 2000 (it was republished with a new afterword in 2011) and is now considered a classic book of business history. It tells how a firm consisting of PhDs and financial wizards—including two Nobel prizewinners, Robert Merton and Myron Scholes—conquered Wall Street in the space of four years, making unprecedented gains, only to see their fortunes completely collapse. In the end, the fall of Long-Term Capital Management almost brought Wall Street to its knees, and the firm had to be bailed out through government intervention. The lesson—that speculating in complex derivative securities using a lot of margin is extremely dangerous, and that risk management models don’t adequately take into account the fat tails of the probability distribution (the likelihood of extreme events)—was very clear to everyone who read the book, which was a major bestseller. Of course, nobody making policy in Washington paid attention to it, and the result was that the same mistake was made ten years later, leading to the Great Financial Crisis of 2008. Is the same mistake being made today? One can only guess; what is certain is that derivative securities remain mind-bogglingly complex, enormous leverage is still common, and billions of dollars are still at stake.

The fall of Long Term Capital Management. From SmithsDaily.com

The book is superbly researched; the tale is superbly told. There are plenty of edge-of-your-seat moments, especially when the giants of Wall Street are trying to decide whether to rescue LTCM—and how. But the book is a cautionary tale, and cautionary tales are morality lessons. You know more or less what’s going to happen from page one, and every few pages you are reminded that pride goeth before a fall. It’s an age-old tale of hubris and its comeuppance, of getting too big for your breeches, of success going to one’s head, and a thousand other clichés.

And the missteps that the partners in Long-Term Capital Management made were made not because the formula that they used was wrong, but because they stopped using the formula. The partners were arbitrageurs who used enormous leverage, and when their positions became too big to efficiently arbitrage, they started to place one-sided bets, or failed to sufficiently hedge, or took liquidity for granted. Managing a small risk is very different from managing a large one, and much easier too. Did genius actually fail, or did the partners stop acting like geniuses? Lowenstein’s descriptions of their missteps make me think that the latter was the case.

Same Approach, Different Rigor: Renaissance Technologies

Interestingly, the big bestseller in the financial world of 2020 is the exact opposite of this book, The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution. During the tenure of LTCM, Simons’ Medallion Fund was doing very well; Lowenstein’s book never mentions the fund, Simons, or his firm, Renaissance Technologies. Renaissance and LTCM took the same approach: hire the best and brightest, staff your firm with PhDs, run strictly mathematical models, keep a tight ship, use a huge amount of leverage, and don’t let your secrets get out. Both firms are exemplars of the “quant” model. Perhaps The Man Who Solved the Market should be called When Genius Succeeded.

Why did Renaissance succeed where LTCM failed? Two reasons: hubris (of which LTCM had a lot more than Renaissance) and liquidity (Renaissance traded primarily in liquid instruments while LTCM greatly overestimated the liquidity of their positions).

As a quantitative investor myself, and as someone intimately involved with the quant community at Portfolio123, I take these lessons to heart. I use leverage, but carefully. I display hubris from time to time, but try to keep it in check. I trade relatively illiquid stocks, but manage my trades and my portfolio carefully. I’ve done extremely well by taking unusual positions, doing a lot of research, and thinking mathematically. Does that make me a stock-market genius? No, it doesn’t. I’m an ignorant amateur compared to the folks in these books. I want to continue winning, and thinking myself a genius won’t help me with that. After all, I’ve made plenty of mistakes, and I’m certain I’ll continue to make plenty of mistakes. I want to learn from each and every one.

But one of the merits of Lowenstein’s book is the reminder that there are always bigger mistakes that can be made. The cataclysmic scenario that brought LTCM to its knees was a huge change in the bond market that could not have possibly been anticipated. Suddenly everything that was happening was completely without precedent.

We are in one of those moments now. As far as I am aware, there has never been a “value inversion” as extreme or as long-lasting as the one we’re in. As far as I am aware, there has never been a stock-market fall and rise as steep and rapid as the one that took place in March, April, and May of 2020. As far as I am aware, there has never been a recession with an onset quite as sudden and drastic as the one we’re in. Interest rates have likely never been as low, at least in the US, and, with the possible exception of the great financial crisis, the Federal Reserve has never tried so hard to prop up the markets.

A good example of the value inversion of the last six years.

How do we invest in unprecedented times and conditions like these? LTCM made the mistake of thinking that the unprecedented conditions they faced were temporary, that things were bound to go back to the way they were, and that this would happen quickly enough to save their hides. As investors today, we cannot make that same mistake. The “value inversion” may last a long time. Interest rates may remain low for ages. This recession is already unlike any other; so were the last two, and the next one surely will be too.

I don’t have any brilliant answers to the above question. My own strategy is based on my belief, which is founded in my research, that it is slightly more likely that what has worked in the last ten years will continue to work in the near future than that something completely different will. And that the riskiest trades are very rarely the most profitable ones.

A Word on Risk Management

What can quantitatively inclined stock investors do to lower the chance that they’ll end up like LTCM and increase the chance that they’ll end up like RenTech?

The essential thing is to manage your risk in a consistent fashion and allow for large deviations from observed conditions. Both LTCM and RenTech used a tremendous amount of leverage, and developed sophisticated risk-managing models. But you don’t have to get terribly sophisticated to understand that frequent small, uncorrelated bets with quick profit-taking is far less risky than enormous, correlated bets that may take many months to be realized. While LTCM started with the small bets, they ended up placing huge ones, and when one went wrong it affected all the others; in addition, all their bets were based on one basic arbitrage principle (spreads will narrow), so when spreads widened in a wide variety of markets, they lost their shirts.

Neither LTCM nor RenTech were stock-pickers like you and me. But in a way, at the end, LTCM resembled the traditional value investor. A traditional value investor places large bets relying on one principle: that prices will revert to the mean. Spreads usually narrow; prices usually revert. But there are times when they don’t. We’re in a period during which expensive stocks have gotten more expensive and cheap stocks have gotten cheaper. That’s not how things are supposed to work. But in hindsight one can see all sorts of good reasons why prices aren’t reverting, and haven’t for a while. I’m not saying one could have been prepared for this back in, say, 2012. I certainly didn’t predict it. But I’ve weathered it, as have some other investors, quite well.

How? By making frequent small, relatively uncorrelated bets on fifteen to thirty stocks at a time and holding for periods ranging from three weeks to six months. By looking at a stock’s value (of course!) but also at everything else about it: its prospects for continued growth, analyst sentiment about it, its share turnover, and most of all a host of stability and quality measures. By investing in stocks that practically nobody has heard of, that nobody on Seeking Alpha is writing about, that the world is ignoring, but that look to me like safe bets.

There may not be any parallel in this technique to what either LTCM or RenTech did, since they weren’t stock-picking at all. But successful stock-picking isn’t just finding stinky stocks that nobody else likes, or brilliant gems that are temporarily undervalued. You can do that all you want, but if you’re not a successful portfolio manager, you’ll find yourself overstretched and on thin ice. What both LTCM and RenTech did right was modeling. Model your portfolio strategy, backtest it, figure out how you can maximize your long-term returns safely, subject your models to the most extreme stress tests you can think of. The way to make money in stocks is to avoid risk, not to court it.


LTCM and Renaissance both developed quantitative strategies that strongly outperformed the market over a period of several years. Renaissance stuck with those strategies, adapting them to some degree, but never radically departing from them. LTCM fundamentally departed from the strategies they used because they had to place larger and larger bets, and the opportunities they had taken advantage of simply didn’t fit with the amounts that they were handling. When genius fails is when it overreaches, when it confidently places bets that no longer fit with its models. When genius succeeds is when it sticks to what works and what has been shown to work.

I take this lesson to heart. I too have developed a quantitative strategy that has strongly outperformed the market in four out of the last five calendar years. Will it fail? I’m certain that at some point it will. But until then, I’ll take my cue from Renaissance rather than from LTCM and stick with my strategy for as long as I possibly can.

Leave a Reply

Your email address will not be published. Required fields are marked *