Was LTCM a lower-tail event?

I have to admit I’m a bit vague on the specifics of what happened in the LTCM blow-up. In my mind, it has always been inextricably linked with Russia’s debt default, although this many years later I’m not even sure that LTCM owned any Russian debt. (Maybe LTCM simply owned spread product, which gapped out in general in the wake of Russia’s default.) In any case, the general consensus, at least as I’ve understood things, has been that LTCM had a strategy which worked until it didn’t — that they were picking up small profits and leveraging those into large profits, while leaving themselves open to a large market reversal associated with some kind of event risk.

But here’s Eliezer Yudkowsky:

While LTCM raked in giant profits over its first three years, in 1998 the inefficiences that LTCM were exploiting had started to vanish – other people knew about the trick, so it stopped working.

LTCM refused to lose hope. Addicted to 40% annual returns, they borrowed more and more leverage to exploit tinier and tinier margins. When everything started to go wrong for LTCM, they had equity of $4.72 billion, leverage of $124.5 billion, and derivative positions of $1.25 trillion.

This actually rings much more true to me. The real killer of LTCM was not Russia, it was the fact that their strategies simply weren’t working any more. And that rather than look for other strategies, they reacted by piling on the leverage. Does anybody know if this is true? Did LTCM’s leverage rise sharply in its final year of operation?

This entry was posted in Uncategorized. Bookmark the permalink.

6 Responses to Was LTCM a lower-tail event?

  1. Archer Martin says:

    I hate to sound like a jerk, but rather than speculate, read Roger Lowenstein’s excellent book, When Genius Failed. He goes into gory detail about LTCM’s strategies, even many specific positions, and the body blows they took for weeks until the Fed forced the bail-out.

    This is based on memory, but I read the book in the last 9 months. I am sure what I say below is accurate, but there may be other important elements that I have omitted.

    Basically, they took an arbitrage strategy that worked well in markets they knew well, and went full bore into markets where they had no experience, and no statistical history, and began doing the same thing (basically going long the risky side of the trade and assuming the “spread” would correct). They also took naked positions, when their former strategy was to effectively trade the volatility, and they were always long volatilty.

    So first, they went from being disciplined traders to essentially being cowboys, although they continued to believe they were disciplined. Second, they went into tons of markets, and by virtue of being in so many markets, were convinced they were diversified.

    Oh, and they forgot completely about liquidity. They took very large positions, in many market so large that they’d be hard to unwind even in normal circumstances. Again, they forgot that not everything trades like bonds.

    But in a crisis, as Lowenstein points out, all correlations move to 1. All markets flee to safety. And everywhere, LTCM was on the risky side of the trade. They got hammered. Their positions were under water. They were getting massive margin calls. But selling would make things worse.

    And very soon, everyone realized what distress LTCM was in and began selling into their trades.

    Get the picture? They weren’t smart at all. Initially, they were brilliant, but they went way beyond the turf where their experience and models applied. That’s hubris, and it took them down.

  2. brad setser says:

    nicholas dunbar’s book is also good. LTCM was long GKOs, tho not to an extent that killed the fund. what killed them was that they were consistently long the illiquid bond v. the liquid one — as martin notes, they were on the riskier or more illiquid side of every trade (tis hard to think of holdings 29 year treasuries rather than treasuries as risky, but they were betting the spread between the two would fall, and in a flight to liquidity, the spread increased). they also had sold — if memory serves — a lot of insurance against big equity market moves. they were betting that equity vol would fall, and it rose — that really hurt (again working from memory).

    but they certainly were long russia. it wasn’t a core position tho. what killed them was the flight to quality and illiquidity when in all markets they were generally long the illiquid and short the liquid.

  3. dWj says:

    My information is also from When Genius Failed, in which Lowenstein makes some egregious errors and makes clear that he doesn’t understand the efficient market hypothesis, which is silly in a different way from what he seems to believe. (Most quixotically, he seems to think it forbids fat tails.)

    Anyway, according to Lowenstein, yes, leverage was increasing toward the end; the fund was actually returning money to investors for the purpose of increasing leverage for remaining investors. Also, as the previous comment noted, when the end was near, other traders sold against LTCM’s positions, exacerbating the problem in the short-term — not so short, Keynes would have noted, as LTCM could remain solvent. Warren Buffett actually offered to buy their positions, which, with his capital, probably would have rebounded somewhat, the prospect of furious unwinding no longer in the picture. That ended up falling through for a reason having to do with timing and paperwork that I don’t remember.

  4. Andy says:

    Donald MacKenzie talks about this in An Engine, Not a Camera. Actually, he mostly speaks about how LTCM’s positions got widely imitated and those positions did worse than other similar positions — they were bad just because LTCM and a host of imitators owned them, and everyone tried to get out the door at once. Pretty interesting, although I’m not sure if he came up with it himself.

Comments are closed.