Biggest losses start with the brightest ideas
By Bloomberg
Today is an auspicious anniversary, though it's one I suspect many people may not recall. On Sept. 23, 1998, former Federal Reserve Chairman Alan Greenspan and William McDonough, then president of the Federal Reserve Bank of New York, managed to orchestrate the rescue of the hedge fund Long Term Capital Management.
It was a strange exercise in both herding cats and moral hazard. It wasn't a government bailout, since no taxpayer money was involved. More than a dozen Wall Street banks, many of which had exposure to LTCM, ponied up $3.65 billion to unwind the fund's complex leveraged bets.
Still, the lesson learned was that in the event of troubles, the Fed could be counted on to lend a hand to a) avoid disruption; b) add liquidity and; c) protect the Street against catastrophic losses. In hindsight, it looks like the lessons learned were the wrong ones.
Recall the summer of 1998 when Russia -- a hot investment for bond underwriters -- defaulted on its ruble-denominated debt. This triggered a chain reaction of losses for anyone who either held Russian debt or had assets denominated in rubles.
The biggest of those suffering losses was the wildly overleveraged giant hedge fund. At the time, not many people understood that LTCM was running mathematical models that sought out the tiniest arbitrage opportunities. It was a grand experiment in the power of mean reversion, and the only way some of these trades made any sense was to amplify their size via leverage. And not just a little borrowed money, but an insane 100-to-1 leverage. So certain were the Nobel laureates connected to the fund in the inevitable mean reversion that they engaged in what is now widely accepted as a combination of towering irresponsibility and sheer idiocy.
The miscalculation they made -- as it so often does -- had to do with human behavior and leverage. When you are working with borrowed money, you are subject to the rules and regulations of the lender. Limits on losses with borrowed money are rather bland, standard housekeeping stuff. That is the way risk managers at lenders to hedge funds manage to sleep at night. Sure, they were happy to lend a few billion dollars to make some obscure leveraged bet on a tiny backwater market somewhere in the nether regions of the former USSR -- but they did have some rules.
As the spread between the asset classes that LTCM made its mean-reversion bets on widened rather than narrowed as expected, its losses ballooned to $2 billion in a single month. Once the trade started going against the fund, margin calls began, requiring the fund to either repay the loans or put up more collateral. That was the beginning of the inevitable end.
Roger Lowenstein, in his seminal telling of this story, "When Genius Failed: The Rise and Fall of Long-Term Capital Management," observed that Wall Street took precisely the wrong lesson from the LTCM bailout. The takeaway was that if you are reckless or irresponsible or leveraged up to the gills, well, no worries. The Fed has your back, even if it has to cajole the rest of the Street into ponying up the dough.
Therein lay the moral hazard. In retrospect, the LTCM bailout looks like a rounding error compared with the financial-crisis bailouts in which the Treasury Department provided $700 billion in rescue money to banks and the Fed added more than $3 trillion to its balance sheet to support the economy. If the losses had been imposed on the lenders to LTCM, maybe Wall Street would have been more circumspect about leverage and risk. Who knows, it might even have moderated the huge leveraged bets made on subprime securities, which laid the foundation for the financial crisis.
Here is the kicker: It isn't that the mathematical models designed by LTCM's Nobel laureates had failed. The upshot of the entire episode was that the trades eventually worked, just not on LTCM's timetable. The math was right; what was all wrong was the psychology of the trade. That was, and always will be, the hardest thing to get right.