Opening Cross: When is Fast Too Fast?

As the financial markets have sped up and become more dependent on low-latency market data to power algorithmic and high-frequency trading strategies, they arguably need oversight that can move at the same pace-not necessarily stricter oversight, but one that can follow and react to the markets in real time, even as the definition of "real time" evolves.

The so-called "flash crash" - perhaps poorly named since it had nothing to do with the much-vilified practice of flash orders - of May 6 still appears to have everyone baffled as to the cause. The Securities and Exchange Commission, which last week approved rules for a new market-wide circuit breaker to prevent a reoccurrence, says it is still investigating the causes of the sudden market drop and rebound.

The fact is that sell orders left hanging in a market with no buyers will depress a price, and returning bids will restore a price upwards. But did the increased speed of today's markets cause that to happen much more quickly than ever before? The SEC's new rules will implement a pause in trading across all US equity markets for a stock whose price changes by 10 percent over a five-minute period. At time of writing, the rules are set to be implemented immediately for a pilot phase running from June 11 to Dec. 10, covering stocks in the S&P 500 index.

But in today's fast markets, five minutes is a long time. So why impose a time constraint when the important factor is the 10 percent price move, whether it takes five minutes or five microseconds? And in this low-latency environment, these movements can add up to a large amount in a small time - after all: the events of May 6 did not trigger any of the existing circuit breakers.

But who actually monitors the markets in those timeframes? Well, the markets themselves. Every consumer of low-latency data uses complex algorithms to scrutinize the markets to exploit any tiny imbalance that could represent a trading opportunity. From an old-school price chart to an engine built using complex event processing technologies, all of today's data and tools are designed to help professionals monitor, predict and react to market movements. But these technologies don't exist in a vacuum: they are available to regulators as well as to traders, and perhaps regulators need to become more like traders themselves if they are to stay on top of the markets without alienating participants by being overly prescriptive.

Most industry participants seem to agree that legislating innovation - for example, by trying to artificially level the playing field to appeal to the lowest common denominator - isn't the way forward. Fast firms will continue to be fast, and will spend a fortune on co-location and hardware acceleration and direct feeds, while slower firms will continue to have to differentiate themselves in other ways.

Larry Tabb, CEO of Tabb Group, has said that speed is a good thing, since it reduces risk, while more than half of respondents to a recent Thomson Reuters poll believe that high-frequency trading has increased liquidity.

While regulators' focus on protecting retail investors is laudable, retail investors are never on par with the big firms, just as in the FX markets where there is a clear difference between retail prices and interbank rates. But that's not to say they don't benefit from low-latency and high-frequency trading: retail orders get rolled up and brokers trade off those positions using their algorithms. Thus, retail investors get a better price because their broker can trade more efficiently, and everybody's happy.

Instead of legislating innovation, perhaps regulators should use those same innovations to their advantage.