Games of spoof can be very expensive. Just ask Mike Ashley, who reportedly lost £200,000 when playing against his advisors at Merrill Lynch. Or ask Peter Beck, the Canadian day-trading evangelist who has lost £8m to the FSA.
These are, of course, different types of spoof. While Mr Ashley used the drinking game to settle a legal bill, Mr Beck’s SwiftTrade equities trading network was — in the FSA’s judgement — spoofing the market:
In the FSA’s opinion, between 1 January 2007 and 4 January 2008, Swift Trade’s manipulative trading caused a succession of small price movements in a wide range of individual shares on the London Stock Exchange (LSE) from which Swift Trade made substantial profits. It has not been possible to measure Swift Trade’s profits precisely; however, they were in excess of £1.75m.
SwiftTrade stands accused of operating a “layering” scam, whereby fake orders are dropped into a stock’s order book to improve the price ahead of a trade in the opposite direction.
According to the FSA, SwiftTrade’s clients were prolific spoofers:
The decision notice states that the FSA believes that this was a particularly serious case of market abuse. It was widespread and repeated on many occasions involving tens of thousands of trading orders by many individual traders sometimes acting in concert with each other across many locations worldwide. The trading led to a false or misleading impression of supply and demand and an artificial share price in the shares they traded which was to the detriment of other market participants. The FSA believes that such conduct, if unchecked, could undermine market confidence.
Such manipulation is usually quite easy to spot, as the spoof orders all have to come from one place. The FSA alleges, however, that Mr Beck was able to call on a personal army of spoofers dotted around the planet.
The FSA explains that SwiftTrade ran “a network of over 50 customers based in more than 150 trading locations worldwide” (how those numbers work we’re not entirely sure; perhaps they all lived on international borders). These key customers in turn engaged over 3,000 traders as “independent contractors”.
Not only was this zombie army told to follow Mr Beck’s directions, the FSA says, it would pay him 27 per cent of trading profits. So, rather than being an intermediary to its clients, SwiftTrade was said to have been acting as a proprietary trader.
So how much could a zombie day-trading army mess around with the daily wash of liquidity in individual stocks? Not much, the FSA admits — though it could do not much quite regularly:
While the price movements which Swift Trade caused were short term, they secured the prices of the shares in question at an artificial level. Further, the trading activity had market impact for a prolonged period in a variety of FTSE 100 and FTSE 250 stocks. The volume of the manipulative trading was large enough to misrepresent the overall liquidity on the order book for the shares in question.
Any similarity to high-frequency trading is not entirely coincidental.
But like all stories in this genre, we are left wondering if the zombies were truly evil or just misunderstood. SwiftTrade argued that it had not influenced the trading strategy of any of its clients. And, perhaps surprisingly, it had an expert witness on hand who agreed.
(S)he told the panel:
“I do not believe that this activity was in any way planned, organised or orchestrated or indeed even discreetly encouraged by Swift Trade. … Co-ordination or even simple co-operation between traders alleged by the FSA is also implausible given Swift Trade’s financial model.”
SwiftTrade’s expert argued that, rather than being centrally co-ordinated, the apparent zombification of SwiftTrade’s army could be blamed on Darwin. Yes, really.
“The principle of these businesses is highly Darwinian, they seek to have a large number of traders each evolving their own trading strategy, those that are profitable will survive, those [that] are not will leave. Inevitably traders will copy each other and the more successful strategies will become more prevalent. As new traders are brought in they will also tend to copy the more successful traders’ strategies. This is a perfect example of an eco system.”
And questions were raised over whether spoofing via a distributed trader network would actually work.
The witness argued that big orders were quite likely to be filled, and market makers were smart enough to disregard outliers. (S)he described as “fundamentally illogical” the idea that a market could be moved by unfilled orders some distance away from the touch price …
… In fact, it was claimed that there would be so little impact from such trading that it was questioned whether it could be fairly assumed that Swift Trade had engaged in layering to generate any profit. It was speculated that if there were to have been any negative impact it would have been upon algorithmic traders and that this was of general benefit to the wider market.
But the FSA wasn’t buying any of it.
Whilst the ‘Darwinian’ hypothesis provides a theoretical and abstract explanation as to how traders could have come up with the same trading strategy independently of each other, the FSA considers that in the circumstances of the case this hypothesis is not a plausible explanation. Instead the FSA considers that Swift Trade disseminated this trading strategy and directed the trading activity.
Mr Beck is challenging the FSA’s decision and in June obtained a High Court injunction against the FSA publishing its judgement. That expired on August 26, meaning you can now read for yourself the full judgement on the FSA’s website.
Update: the first cut of this story gave the impression that the FSA had called the “Darwin theory” witness. We should have made it clear that this expert witness had been put forward by SwiftTrade.
Zombie maths – FT Alphaville