In all of the recent talk of derivatives and their role in our current problems, there are a number of important lessons to be learned in looking back at Long Term Capital Management.
John Meriwether's Long Term Capital Management was the envy of the financial world from 1994 until 1998, when it lost roughly $4 billion in a couple of weeks. Back then that was a lot of money, and there was fear that the failure would take others down, as LTCM had deals with most of the big investment banks. Its tentacles, much like AIG's, were far reaching, and so a bailout was considered necessary.
What's interesting, and telling, is where their losses came from--and the leverage that was used.
The fund was essentially the arbitrage unit spun out of Salomon Brothers. Its focus was on fixed-income arbitrage, betting that spreads between various types of related instruments would close and return to "normal" historical relationships. They would buy the "underrpriced" instrument and sell the "overpriced" one.
Leverage is used to get a profit from this type of trade. And LTCM used huge leverage, far more than anyone else.
Two members of the team at LTCM were Robert Merton and Myron Scholes, who were behind the Black-Scholes option pricing model and won a Nobel Prize for their work. The formula is credited as the basis for most risk calculations, even today. So LTCM was, some have suggested, an "experiment in mathematical risk management." Sound familiar?
There are a number of good books about the life of LTCM, and the one I just finished is "When Genius Failed: The Rise and Fall of Long-Term Capital Management" by Roger Lowenstein. One thing I find fascinating about the story is the fact that in 1998, having trouble finding opportunities for their massively leveraged capital, the group started selling equity volatility. They concluded that the implied volatility was too high and began selling.
Amazingly, the level of implied volatility that they deemed too high was just 19 percent (think a VIX of 19). They were apparently looking at the historical volatility of 15 percent at the time. They must not have been looking at the fact that the VIX was as high as 35 (48 intraday) in 1997. According to Lowenstein, "they had a staggering $40 million riding on each percentage point change in equity volatility."
When all was said and done, it was largely this highly leveraged volatility bet that did them in. The trade ate up the cash that the company might have used to ride out other positions, and it lost $1.3 billion just on the volatility play.
As far as risk management was concerned, LTCM believed that the type of losses it experienced were "so freakish as to be unlikely to occur even once over the entire life of the universe and even over numerous repetitions of the universe."
"The traders hadn't seen a move like that--ever," Lowenstein wrote. "True, it had happened in 1987 and again in 1992. But Long-Term's models didn't go back that far."
My jaw is hanging open. They didn't run their models two years before their start-up? Or back to the crash of 1987? What kind of "genius" uses risk modeling like that? 
One would think that the banks that had to bail out LTCM learned a hard lesson, one of limiting leverage and the importance of not relying too heavily on models. At the table that day were Bear Stearns, Goldman Sachs, JP Morgan, Lehman Brothers, Merrill Lynch, Morgan Stanley, and Salomon Brothers, just to name a few.
But they either have very short memories or learned something else. Maybe the lesson was actually that you could get away with it.
Meriwether went off and started a new fund with his trading crew in short order. It continued to do well until last year, when he apparently blew up again. Scholes did the same. Those investment firms were using leverage of up to 40 to 1 and relying heavily on essentially the exact same risk models.
So the lessons from LTCM actually were to lever up, get as many counterparties as you can, and collect as much easy cash as you can, despite the huge fat tail risk. If it all blows up you will get bailed out. And you can just start over.
The cycle is made clear in a Reuters article from yesterday. "Many fund managers are opting for so-called nickel strategies, which yield small gains most of the time and build a track record that can help retain client money but bear the risk of huge losses once in a while." The article quotes Hongjun Yan of Yale who states that at least 40 percent of hedge funds invest in nickel strategies.
The term "nickel strategy" comes from Scholes himself, who, as Lowenstein tells it, used the metaphor to explain how LTCM made its money: "Long Term would be earning a tiny spread on each of thousands of trades, as if it were vacuuming up nickels that others couldn't see."
"Black Swan" author Nassim Taleb sees it as picking up nickels in front of a steamroller--and, evidently, that steamroller comes around more often than the models predict.
(This article originally appeared in optionMONSTER's Open Order newsletter of May 29.)
