What caused the “Crash of 2:45”?

The short answer is that anyone who knows the details isn’t telling.

what happened

Last Thursday, May 6th, was a generally bad day on Wall Street.  What makes this session worthy of not, however, is the Tower of Terror-like plunge that the market took at about 2:45 pm, followed by an equally swift rebound several minutes later.

visiting the SEC

Representatives of the major exchanges met with the SEC yesterday to discuss the situation.  All reportedly professed to have no idea what caused the sudden decline and rebound.  That’s not really much of a surprise.  On the one hand, no one wants to take the risk of becoming Congress’s latest whipping boy if they admit to having had even an innocent hand in the debacle.  On the other, its track record in cases like Madoff suggest that the SEC doesn’t have the knowledge or interest to pursue the issue by itself.  So anyone involved would likely be calculating that, absent a public declaration of repsonsibility, they will never be found out.


There have been two persistent stories about the trigger for the mid-afternoon drop.  One, offered by CNBC, is that a trader from Citigroup made an input error that generated a sell order that was 1000x the amount intended.  Citi denies this.  The second, more detailed–and therefore, I think, initially more plausible–comes from the Wall Street Journal. According to the newspaper, the initial trigger was a relatively small sell order placed in the Chicago options market by hedge fund Universa Investments.   However, the Journal may have been attracted to this trade as much by the fact that Universa is linked to Nassim Taleb, who popularized the idea that disruptive “black swan” events are much more numerous than academic theories allow for.

what we do know

Several things are clear:

–as the computer-to-computer selling got more intense, the NYSE turned its machines off and reverted to manual processing of trades.  Apparently, although I wasn’t aware that this was the case, this is the NYSE’s publicly stated policy.  This action effectively eliminated one of the main ways derivative holders could hedge their positions with offsetting exposure in physical stocks.  It caused trades to be redirected to other middlemen, who buckled under the added pressure.  It isn’t clear whether the others took defensive measures similar to the NYSE’s or whether they were simply overwhelmed by volume.

–ETFs were hurt unusually badly in the selloff.  According to the Financial Times, two-thirds of the securities where exchanges subsequently cancelled trades were ETFs.  ETFs themselves represent only about 16% of the total number of securities traded on US exchanges.  I don’t think there’s any great significance to this, however.  In a “normal” stock trade, the market makers matches a third-party buyer and seller–who have already decided they want to transact–and charges a bid-asked spread.  In an ETF trade, however, the counterparty is often the market maker himself, who is constantly calculating the NASV of the underlying ETF and his hedging possibilities and deciding whether to transact at a given price or not.  To let the market know he is potentially a buyer or seller, he typically has “placeholder” bids of, say, $.01, that under normal circumstances let the world know that he may be interested in transacting.

Last Thursday, a lot of those $.01 bids were hit before the market makers could withdraw them.

–as mentioned above, the exchanges unilaterally decided to cancel transactions they considered to be anomalies.  This presumably leaves a lot of unhappy individuals who had placed low-ball limit orders.  Perhaps not by coincidence, the counterparties–that is, the losing side–to those canceled trades are the same exchange members who canceled them.

what happens next

The SEC is under political pressure to do something to prevent a recurrence of anything like the near-instantaneous decline of 5% that happened last Thursday.  Some of this pressure likely comes from proponents of having a stronger “uptick rule” (see my post on this topic).  Although the current uptick rule seems to me to invite sudden plunges in stock prices of the type we had last week, no one is publicly suggesting that this is the cause of last Thursday’s meltdown.

Whether needed or not, a new trading halt mechanism is probably the “solution” the SEC will opt for.

I have two related thoughts:

–added volatility may just be a fact of life in a modern stock market, where computer-to-computer trading and monthly performance measurement of the type hedge funds undergo are prevalent.  In other words, get used to this.

–the first time something like this happens, it’s a surprise.  Even now, I’m sure that investors are beginning to devise strategies to cope with–and take advantage of–increased volatility.  Understanding what’s going on and planning for it, investors will enter the market as buyers sooner and more aggressively. By doing so, they will temper volatility.  For their part, sellers, learning that there’s a penalty for being aggressive on the downside, will adjust their behavior as well.

That’s why I think canceling money-losing trades made by people pushing the market down is a really bad idea.

–no one so far is questioning the behavior of the NYSE (admittedly, not my favorite organization).  It’s kind of like switching from big fire hoses to small ones as soon as the fire starts.  What good is that?  In fact, if this incident isn’t the accident that I think it was, short-sellers may have been counting on the NYSE to remove liquidity from the market.

Short-selling and the uptick rule

The SEC has been holding hearings about reinstating the “uptick rule” for short-selling in the US market, and has come up with a new rule.  The rule-making comes as a reaction to assertions that the Wall Street decline in late 2008 was made worse by speculative short sellers taking advantage of the removal of the uptick rule in 2007.  What are the issues?


A short seller borrows stock from an owner, promising to return it on demand, and  promptly sells it.  The debt to the owner is expressed, not in dollars, but in a specific number of shares of a certain stock.  So at some point, the short seller must buy the stock back in the market and return it.  This can happen because the short seller decides his investment idea has outlived its usefulness, or because the original owner “calls” the stock back to him.

why sell short?

The short seller may think the stock in question will go down, so that he can make money by buying it back at a lower price and then returning it to the original owner.

Or he may have a hedging strategy in mind (this was the original “hedge fund” idea).  If so, he will reinvest the sale proceeds in something else he thinks will do better than the stock he sold short.  He might, for example, short Toyota and go long (i.e., buy) Ford.  In this case, he will make money if Ford goes up more, or down less, than Toyota.  He doesn’t need the stock he’s betting against to be an absolute loser, just a relative one.

Short selling is highly institutionalized on Wall Street.  Many financial firms and most large institutional investors have stock lending departments, which deal with each other to facilitate short selling by locating and arranging for stock to be borrowed and arranging for collateral to be provided.  (As a portfolio manager, I understood why my firm would do stock lending.  It didn’t thrill me, though.  It always irked me on those occasions when, despite their promises to the contrary, the borrower refused to return a stock I wanted to sell.)

Then, of course, there was the incredible disaster at AIG–and apparently other middlemen as well–where that company took the collateral it was holding, typically Treasury bonds, sold it and replaced it with sub-prime mortgage securities so it could earn higher interest income.  After sub-prime mortgage securities tanked, short sellers couldn’t close out their positions because AIG couldn’t give them back their collateral.  AIG needed another $40 billion+ from the government to clean up this mess.  But that’s another story.

the uptick rule

The uptick rule was instituted for exchange-traded stocks by the SEC in 1938, in response to a market swoon in 1937.  The SEC believed the market’s fall was caused, or at least made considerably worse, by rampant speculative short selling.  In response to brokerage industry lobbying, the rule was rescinded in July 2007–just at the beginning of a sharp market contraction made considerably worse, in the view of many, by rampant speculative short selling.

The rule was that you could only sell a stock short on an uptick.  That is to say, you could only sell a stock short either:

–at a price higher than the immediately previous trade, or

–at the same price as the previous trade, if the last price that was different from the previous trade was a lower price.

Put in less precise terms, the rule says that if a stock is declining you can’t hold it down or push the price even lower  by selling it short.  You can only sell short into a rebound (and thereby prevent that rebound from advancing), but if the stock turns lower again, you have to stop selling it short.

Remember, none of this prevents a “natural” seller (someone who owns the stock) from selling.  The uptick rule only applies to short sellers.

the new rule

The SEC has just instituted a new, limited anti-short selling rule.  It applies only to stocks which have fallen by 10% in price during a trading day, and applies only to the remainder of that trading day and the following trading day.

During that time, a stock may only be sold short if the sale price is higher than the highest bid price maintained by any market maker in the stock.  In other words, the short sale trade has to establish an uptick.

The main untested feature of the new rule is that it only kicks in after a stock has fallen by 10%  Before that, it’s a free-for-all.  It may well happen that market makers decide to move the market in a stock that’s being sold short down as fast as possible to the 10% mark, where they receive temporary protection against short sellers.  In my experience, that’s what happens in foreign markets when market makers encounter any sort of concerted selling.

“naked” shorts are much more important, I think

I don’t find anything particularly wrong with short selling–I should mention, though, that I worked on, and later ran, a (very successful) short portfolio in the early Eighties.  I’m also not sure that the uptick rule is a significant issue.

If anything, there may be unintended negative consequences.  My experience outside the US with markets that put arbitrary, or commodity market-like, restrictions on selling is that they end up with declines that are longer in time and deeper in extent than similar markets that don’t have such restrictions.

I think that colorfully named “naked” shorts are a much more significant concern.  More on that topic in my next post.