A New Jersey-based trader was indicted last Thursday in Chicago on charges of manipulating commodities markets. It is alleged that Michael Coscia, the owner of a trading firm called Panther Energy Trading, made profits of around $1.6 million by defrauding participants on CME Group and ICE markets. Well, when you consider the amounts of money generated in the Sumitomo scam, that doesn’t really sound much, I suppose.
But this case is interesting, not because of the size of the fraud, but because of the nature of it. You see, in this case, the accusation is of using algorithms to generate spoof orders to attempt to influence other market users to behave in a particular way. ‘Spoofing’, for those not fully familiar with the slightly arcane world of high-frequency trading, involves entering orders into an electronic market with the intention that they should not be executed; they are entered and cancelled within milliseconds, their sole purpose being to convince other market users that there is volume waiting to be bought or sold. When those other market players react, the spoofer takes advantage of the spike or dip in price that his illusion of volume has created and trades at a price which would otherwise probably not have been achievable. (I should point out that Mr Coscia has previously been fined by the CFTC and the FCA for activities relating to the use of high-speed algorithms in financial markets. This time, though, it is a criminal indictment, which carries heavy penalties, including, potentially, substantial gaol terms.)
It’s difficult to know where to begin with this. If a trader enters an order, then surely he must have the right also to withdraw it before it is executed? Well, of course, but that’s not really what we are faced with here. These orders are entered and withdrawn in milliseconds, by an algorithm; in other words, the programme is set up with no purpose other than to create an illusion of activity which is entirely fictional, in order to influence the behaviour of other traders. Now, as a purely philosophical issue, attempting to influence the behaviour of others is part of trading, just as it is part of playing some card games of chance. It is the pitting of one person’s wits against another’s; nobody likes to be caught out by a spoof, but actually the practice is as old as the hills. At least, I guess that would be the justification that could be advanced; it doesn’t really hold up though, on examination.
First, financial markets are not a philosophical arena - they are a vital part of the financial system, and need to be seen to be clean and transparent. Secondly, a face-to-face negotiation, or a card game, is a particular environment, not frankly replicated by an electronic exchange. Algorithmic trading is clever - producing the programmes that the quants do requires a lot of skill and knowledge. But it’s also seriously threatening to the well-being of financial markets. I’ve directed readers before to Michael Lewis’ excellent book “Flash Boys” and I urge any who haven’t yet read it to do so. It gives a disturbing sight into the world of HFT. Lewis concentrates on equity markets in his book, but the allegations against Mr Coscia involve commodity markets, including, to highlight the significance of this for the majority of readers here, copper. Now, the accusations concern CME, but it would be a brave man to say that similar activities were either not happening or not possible on the LME’s Select system.
I don’t know Mr Coscia and I bear him no ill-will; but what he has allegedly been doing is a risk to all of us. The flash crash of 2010 was a sign of how the rise of high-frequency algorithmic trading can impact even something the size of the US equity market. This criminal prosecution may mark the beginning of the regulators looking more closely at what is going on; maybe it will even push them towards the idea of banning this type of trading altogether. It is not good enough simply to hail high-frequency and algorithmic trading as just the natural progression of markets; these are potentially pernicious influences, and it would be advisable to control them before they create major problems, rather than wait until it’s too late. Otherwise, having sown the wind, we are going to reap the whirlwind. Just because it’s clever, that doesn’t make it desirable.
I must thank @goldmatrix1; it was their comment on Twitter which drew my attention to this case.