Trader Talk

'Flash crash' 5 years later: What have we learned?

The 'flash crash' of 2010 in 90 seconds

Let's not dwell on who caused the "flash crash." Can we just agree that it wasn't one person, that it was likely caused by a confluence of events?

Good. Then we can concentrate on more interesting issues: can it happen again? What changes have been made and what further changes ought to be made?

Can it happen again? Sure it could, but the chances that it would happen in the manner it happened have been reduced. Here's why:

1) Stocks cannot go to a penny like they did on that day. One of the most damaging developments was to see, for example, Accenture go from $40 to $0.01, Boston Beer go from the $60s to $0.01 (all off of the NYSE floor). These "stub quotes" were never meant to be serious bids; they were merely placeholders for market makers. One of the first actions regulators took was to ban these "stub quotes." Now, market makers have to post bids and offers within 8 percent of the National Best Bid and Offer (NBBO).

2) New circuit breakers halt individual stocks if there are sudden price swings. Five years ago, each exchange had its own rules that governed when individual stocks would be halted for sudden volatility. This confusing mishmash has been eliminated: there are now consistent system-wide circuit breakers for all individual stocks (known as limit up/limit down), as well as revised market-wide circuit breakers that will kick in when the S&P 500 drops 7 percent or more during the trading day.

Who caused the 'Flash Crash?'

There are other changes that have been made to improve market structure.

For example, regulators have adopted several rules to address technological or software failure, such as the "algo gone wild" that almost brought down Knight Trading in 2012, or the Facebook IPO glitch at Nasdaq the same year:

1) To cut down on the chances of some kind of trading algorithm disrupting the market, regulators have adopted market access rules that require brokers to monitor the flow of their customer traffic. Exchanges have also implemented a form of "kill switch" that would alert brokers when there is unusual activity.

2) Last year, the SEC voted to adopt Regulation SCI, which requires market participants to develop uniform rules to test software and reduce trading glitches.

3) There have been some efforts (not enough) to reduce "single points of failure," areas that would be especially vulnerable should a software glitch occur, and create backup systems that would immediately come online to reduce trading disruptions.

But even more needs to be done. One of the problems with pinpointing what happened in the "flash crash" was the lack of a complete trading record. Turns out the data is scattered. A new proposal for a Consolidated Audit Trail (CAT) would allow the SEC and FINRA to better monitor trading activity. That's good news.

The bad news: the data base would be massive and likely very expensive. The process of selecting someone to develop the project has been slow, but is ongoing.

Then there are the calls for changes in market structure. Some say the current system has created too many trading venues (11 exchanges, more than 40 dark pools) and should be simplified.

SEC Commissioner Dan Gallagher, for example, has been calling for a "holistic review" for several years.

Gallagher has specifically called for a complete review of Reg NMS and its core principle, the trade-through rule, which requires all trading centers to execute trades at the best possible price (lowest bid, highest offer). Gallagher believes the rule has greatly increased the complexity of the market and increased the chances of technological failure.

He has a point: the trade-through rule has increased complexity. Orders need to be routed to dozens of different venues that have different incentives and pay different rebates, which has caused a proliferation of order types.

But that is the logical outcome of trying to increase competition, which is what the SEC wanted when it helped create the modern market system twenty years ago.

Most traders support the basic concept of a trade-through rule: prices should not be executed on other venues at values that are inferior. It's part of best execution, which is a fiduciary duty.

One change may be possible: it may not be necessary to route all orders to all venues when one of those venues has a miniscule market share, but it's tricky to decide where to draw the line.

Others have focused on high-frequency traders, arguing their influence should be reduced. There have been proposals for a tax on excessive bids and offers that are cancelled; others have proposed a financial transaction tax to reduce short-term trading.

For her part, SEC Chairwoman Mary Jo White last year proposed that high frequency traders be registered with the SEC, an obvious and sensible step.

She has also floated the idea of an "anti-disruptive trading rule" that would ban short-term trading strategies when the markets were extremely volatile or liquidity was especially scarce. This would be much more difficult to implement.

Still, White and the SEC has moved slowly, as they should. It's very clear they do not believe the U.S. trading system—which they created—is fundamentally broken, and they are far more concerned that making big changes now will create some unforeseen circumstances.

There's a second problem: the SEC simply does not have the staff to aggressively pursue a widespread overhaul of the U.S. trading system. For the last several years, they have been overwhelmed with implementing Dodd/Frank financial reforms, which will continue to occupy them for the next several years.

The most that can happen is modest moves, like the limit up/limit down rule.

To provide a conduit for reform, the SEC recently created an Equity Market Structure Advisory Committee; its first meeting will be on May 13.

The first topic of discussion: the trade-through rule that Gallagher wants to change.

Stay tuned!