The Fed and the May 6 'Flash Crash' – By Mark Spitznagel

Thanks to Matt for passing this along. Starting tomorrow, I'll be on vacation without Internet access until the end of next week, so the blog will be quiet until then.

Regulators have been busy searching for the cause of the May 6 "flash crash" when the market dropped by 9.3% and then recovered within minutes. I think it's a good bet no cause will be found; there is still no consensus on what triggered the one-day 20% stock market crash of 1987. But even if there was no trigger, market conditions created by the Federal Reserve's easy money policy definitely made the crash more likely.

The market is a critical system. To illustrate, let's consider another fragile system: the earth's crust. Imagine geologists scouring through the debris of a big earthquake in search of its trigger—as in, "Let's investigate anyone jack hammering in the minutes leading up to the quake." It is intuitively obvious that earthquakes don't have identifiable triggers. We know that big earthquakes, which happen very rarely, are nothing more than many little earthquakes piled on top of each other due to stresses built up within intricate networks of faults. These little fissures cascade into enormous ruptures. The more correlated the fissures, the more delicate the system.

Back to markets. Think of every investor holding a risky position. Then think of all of these investors together in a big herd. Each member of the herd focuses on what the others will do next, since the only reason anyone takes a position is because others are initiating like-minded ones.

When imitative behavior starts happening in markets en masse, expect funny things to happen to liquidity. All you need to know about market dynamics—as I learned as a Chicago pit trader—is that market prices always adjust to the level where market-makers see balanced two-way order flow between buyers and sellers. All market-makers want to do is buy at the bid price, sell at the offer price, and at the end of the day go home unscathed. When there are only buy orders, for instance, expect market-makers to be unwilling to sell to those buyers until the price has adjusted to the point where they see roughly equal buyers and sellers again. To expect them to do anything else is to imagine them as charities.

So when you combine imitative behavior with noncharitable market-makers, there will be seismic waves from time to time. What makes our current system particularly prone to global ruptures is that hair-trigger traders have crowded into exceedingly risky bets. Why would that be, with the crash of 2008 so fresh in traders' minds?