James Rundle: That Golden Touch

The recent UK court ruling against Swift Trade, a Canadian day trading operation convicted of market manipulation through a complex process of entering and cancelling orders to move share prices, was damning. “Left to ourselves, we might well have concluded that £8 million was insufficient,” said the Upper Tribunal of the UK Tax and Chancery Chamber, referring to the case brought by the Financial Services Authority (FSA) against the now-defunct operation and its approximately $12 million penalty.

The operation involved battalions of day traders, who would input thousands of orders on the London Stock Exchange and cancel them swiftly, taking advantage of the pre-planned price swings in the process. Concerns were initially raised by the exchange and Swift Trade’s brokers, and a remarkable story of evasion unfolded. Swift Trade assured its market access providers that it had updated its risk, compliance and control systems to ensure that the process—known as layering—didn’t happen again. It switched brokers when said providers stopped believing the firm, and did more or less everything in its power to undermine existing prohibitions on the practice.

Digi-Collar Criminals
Given the size and sheer level of activity involved in the layering operation, the fact that the FSA was able to gain a conviction and provide analysis-based evidence of trade-data samples to support it, shows how modern regulatory efforts are striving to keep up with electronic markets. It’s too little too late for the FSA, which will be broken up into two separate entities come April 1. Indeed, many have unfairly dismissed its efforts in this area as going after a soft target in the form of a defunct company, and finally growing some teeth when it least matters.

Regulatory cousins across the pond, though, are also making this move into data-driven analysis, and the prosecution of more technically sophisticated crimes. The US Securities and Exchange Commission (SEC), for instance, has enlisted the services of proprietary high-frequency trading (HFT) firm Tradeworx in order to build its trade-monitoring system. Named “Midas,” presumably after the Greek king who could turn objects to gold with a touch, the system will allow the SEC to analyze market activity in real time. The Commission is clearly taking it seriously, ramping up its analyst hires and spending to $2.5 million, even if some in the US Congress accuse it of using the fox to guard the henhouse.

Regulatory agencies have become increasingly stymied by the use of algorithms and the complex processes required in unpicking data associated with HFT. The report into the May 2010 US Flash Crash, for instance, took months to emerge despite collaboration between the SEC and the US Commodity Futures Trading Commission (CFTC). Likewise, the errant algorithm that nearly destroyed Knight Capital last summer proved unwieldy to analyze on the part of the Street’s police.

The fact that agencies such as the FSA, the SEC, the CFTC and others, are beginning to engage with statistical analysis such as that seen in the Swift Trade case is encouraging.

Top Dog
It is easy to argue that this should have been sorted out years ago. Computerized trading is hardly a new development, and for too long, the oversight burdens have been falling on the banks, exchanges, and brokers, while regulators have behaved reactively rather than proactively. Some of this is down to structure—the SEC can’t hire non-attorneys to permanent posts, for example, and I’ve yet to come across a lawyer specializing in the closely related fields of corporate legislation and Big Data analytics. But the fact that agencies such as the FSA, even if it won’t be here for much longer, the SEC, the CFTC and others, are beginning to engage with statistical analysis such as that seen in the Swift Trade case is encouraging. It should, ultimately, lead to safer markets, a fairer playing field, and an end to criticisms of “Wild West” financial practices.