Commentary: Market Madness

Why the core of the problem is an old-fashioned order type

It had to have been a fat finger. Like many trading professionals, I assumed that the historic 1,000-point bungee jump in the market on May 6 had a simple explanation–a “fat finger,” Wall Street parlance for a mistake. Most traders went to bed that Thursday night looking forward to reading about the unfortunate schlemiel who hit the button. I imagined a 22-year-old assistant who keyed in “billion” instead of “million,” who then absent-mindedly clicked through five increasingly dire warning screens while in a rush to make the afternoon Starbucks run. I thought we’d all know his or her name within 24 hours, and the story would become a great source of jokes for years to come. If we were really lucky, Mr. Fatfinger would turn out to actually have big fat fingers, after which giant keyboards would be mandated for traders, and we’d all move on.  

But sadly, we were denied getting a simple explanation, and we were denied getting cool new enlarged keyboards, because almost two weeks in, no clear culprit had been named. I had to finally acknowledge that the blimp-fingered blunderer I had dreamed up did not exist, and the answer was far more complicated than a simple error.

To review what happened on May 6: Starting around 2:40 p.m., stocks began sliding rapidly, accelerating a few minutes later into a total freefall. A handful of stocks were momentarily wiped out, trading all the way down to a penny. Then the momentary lapse of reason ended as quickly as it started, and the affected stocks soon rebounded. Four hours later, the exchanges announced they were arbitrarily busting the trades they deemed most egregious, while allowing the merely preposterous ones to stand.

The lack of a simple answer had all of Wall Street pondering–if it wasn’t an error, what was it? Theories were flying as to what triggered the initial wave of selling. The conspiracy theory held that an evil genius options trader who was long volatility purposely caused this “Black Swan” event, with the Wall Street Journal even dangling the enticing possibility that the author of The Black Swan himself was involved. Then I heard an even more fun theory when a radio show caller proposed it was financial terrorism from abroad. Yes, rather than a trader with a fat finger, it was Bin Laden giving us the finger.

By a week into the investigation, the convergence theory seemed to be winning. The theory held that unusual conditions combined to form a financial perfect storm capable of generating gale-force market winds. In short, the explanation went something like this: The Greek riots created nervousness, volatility spiked causing quants to reduce their maximum position sizes, and then an initial selling wave from a large money manager hit the S&P futures knocking it to a discount.  As stocks followed futures down, the NYSE began to hit its “Liquidity Replenishment Points,” removing its bids from the public quote, and simultaneously retail stop-loss orders began getting triggered, generating a fresh deluge of market orders. Electronic market makers began getting long tons of stock, and with futures at a discount, they couldn’t hedge effectively, and therefore stopped buying. Bids then rapidly faded, and with no one left buying, the market collapsed.

Regardless of the explanation, the results of the day revealed a serious flaw in our market structure. We all saw the reality–that there are no mechanisms in the current market system to stop panicked sellers, or fat fingers, or just plain old stupid orders from whacking a stock all the way down to zero. While politicians reflexively blamed high-frequency trading, and some blamed any traders who pulled their bids in the midst of the maelstrom, no one pointed out the obvious: that a stock can only go as low as the lowest-priced sell order. The people at fault for driving stocks to a penny were not the traders who made a rational decision to avoid buying in a moment of great uncertainty. The true culprits were the sellers who were willing to pound these stocks down to any price by using a dangerous tool called a market order.

An order to sell “at the market” is a limit order with a price of zero, while a buy market order is even more frightening, being a buy order with a limit of infinity. The order type is based on faith that the other side will materialize at a reasonable price, which is unlike how people buy or sell almost everything else–no one bids for a house, or even for a pair of Yankees tickets without putting out some maximum price. Yet the majority of orders in the equities market from mom-and-pop investors are sent at the market. On normal days, this isn’t a problem, since professionals are around to absorb this steady barrage. But if the pros momentarily step out, as happened on May 6, the constant flow of retail market orders guarantees that stocks will lurch to ludicrous levels.

There are a few ways to attack this problem. The exchanges and regulators are currently focused on installing “circuit breakers,” automatic trading pauses that trigger in the event an air pocket is hit. The latest plan calls for trading halts triggered by any trade outside of a designated price range, followed by an opening auction a few minutes later. As long as the pauses are long enough to give people time to examine their orders and make trading decisions, the circuit breakers should be effective. But one problem with halts based on prints is that they don’t trigger until after there’s already been a bad print–the equivalent of having the troops put on Kevlar vests only after the first soldier has been shot. 

More elegant would be a futures-style limit down, which simply doesn’t allow trades below a certain level, until some amount of time has passed, and then the level is reset. With limit down, the bad trade would be stopped before it happens, making “clearly erroneous” trade breaks announced at 7 p.m. a thing of the past. A limit-down system is the most conservative, the equivalent of wearing bullet-proof vests before anyone’s been hit.

But while circuit-breakers or limit-down mechanisms can certainly minimize the damage caused by erroneous bullets, if the bad order is still not cancelled, the stock may just plummet again after the re-open. Wouldn’t it make sense to try to stop the bullets from being fired in the first place? The best way to do that is to remove the gun: Eliminate the market order. Make retail investors think about the lowest price where they would really be willing to sell, including mandating price limits on stop-loss orders.

Investors in Germany, Brazil, Hong Kong, and many other major markets are already limited to limit orders. Without market orders, investors can still aggressively trade–they can type in a price that’s a dollar below the current price, or even $5 below. But make all orders have some boundary, so that in the event the bids all fade in a perfect storm of market madness, professionals will be less likely to have a career-ending fat finger, and investors will be less likely to shoot themselves in the foot.

Dan Mathisson, a Managing Director and the Head of Advanced Execution Services (AES) at Credit Suisse, is a columnist for Traders Magazine.  The opinions expressed in this column are his own, and do not necessarily represent the opinions of the Credit Suisse Group.