Spoofing is a strange thing; some could argue it is part of a best execution policy when trying to fill a very large order, others that it is criminal activity, and both would have valid points – but is it fraud?
I have no wish to get into the legal technicalities, especially around US law, but it’s an interesting question, and brings another dimension to the whole issue. Last week Merrill Lynch Commodities agreed to pay $25 million to the US authorities over charges that two of its traders had conducted spoofing strategies in precious metals futures – the two men are on trial in Illinois over the issue – that is ongoing. Part of their argument for dismissing the case is that spoofing has its own legislation in the US and therefore should not be considered wire fraud.
The charges brought against Edward Bases and John Pacilo are part of an umbrella action by the US authorities brought in 2018 against banks, traders and a technology provider and, it has to be said, the results so far as mixed at best for the US government. Jitesh Thakkar, the software provider was acquitted a short while ago – undoubtedly the right decision for as Chuck Mackie, who many of you will know, wrote in this blog in April, traders don’t share their secret sauce with anyone. Elsewhere, the traders are either awaiting judgement or have, in the case of former UBS trader Andre Flotron, paid a fine to the CFTC but been acquitted by a court.
An interesting aside in this issue that some of the traders were on the receiving end of disciplinary notices from CME, but not all. Jiongsheng Zhao, who was found guilty of spoofing in a US court, but is awaiting the final outcome because his sentencing has been postponed pending an estimation of how genuine his offer to help authorities really is, was fined $35,000 by CME for his activities on the exchange and barred from trading…for 10 days. Elsewhere, former Tower trader Krishna Mohan, was fined $10,000 and barred for three years, none of the others received sanctions.
In other words, the jury is, pun intended, still out on whether it is fraud, although I would have thought that an exchange group would come down a lot harder on someone they found acting fraudulently than a 10-day ban. Either way, the decisions that will be made by various judges over the coming months could have a significant impact on the authorities’ ability to clamp down on spoofing – whether it be by technicality or not, a ‘not guilty’ would not help their efforts going forward.
To go back to the start of this column, and I think I have teased you enough by now, is it valid to state that spoofing could be part of a reasonable best execution policy? Well, this is where I get controversial and say, yes it could…within reason.
Let me explain/dig myself in deeper (delete as appropriate). In the cases of Bases and Pacilo, the US government is arguing that they spoofed the market by placing a genuine (often iceberged) order into the market and then either placed very large and visible opposite orders or a series of smaller but frequent opposite orders into the market to create the illusion of a lot of buying and/or selling interest. The critical aspect for me was that the opposite orders were not placed at top of book, therefore one can question the all-important “intention to trade”. The fact that they were flashed into the market and rarely in play longer than a couple of seconds reinforces that sense, although, of course, only the traders themselves will really know whether that intention existed.
All of which brings me to my point about best execution. If a trader is executing a very large order, likely to have market impact and create signalling risk, should that trader be able to “throw off the hounds” by reversing their order at times? As an example, if a trader has an order to buy that will take up several hours, should they be able to sell at various times during the execution to perhaps make people think they have finished buying?
We should not kid ourselves that there are not algo trading participants out there in all markets that seek to sniff out large tickets and (legally) front run them – and that becomes a problem for the hedger. If you look at the case discussed above, anyone who may have lost money due to the alleged spoofing activities is likely to have been a professional trader and not a hedger – the latter don’t tend to move their bids and offers because interest appears against them. As one of the chats in the Bases and Pacilo charge observes, it was really about “spoofing the algos”, and if that is the case, they were influencing short term traders looking to grab a pip here and there.
To go back to my example, if a trader is executing a large ticket on behalf of a customer, they have a duty of care to that customer, and that includes not letting the market run away by making the order too visible. If the “sniffing” algos pick up on the order, they will jump in front of it, or, if they are a market maker, skew their price away from the order – in either case, the customer gets a worse fill. If that same trader could occasionally stick an offer into the market at top of book, or just hit the bid, then that would muddy the waters a little and, hopefully for the customer’s sake, reduce slippage.
I cannot see any compliance officer in the world accepting this as a strategy, of course, but I wonder if that is really a question of them not understanding how markets really work? Because to me, there is one absolutely crucial difference between to two examples, in the latter the trader would do just that – trade. There would not only be an “intention to deal” there would be actual trades and the counter bids/offers would be at top of book, not two or three levels down. The intent would not be to force the market into a counter-order the trader had out in the market, it would be to help reduce market impact from their client’s order.
Pre-empting some of your feedback, yes, I accept the challenge would be what happens if it fails to stop the market and the customer order takes longer and costs more, but could the executing firm not actually calculate that and recompense the customer? It needs to be stressed that the executing party would have to fully explain to the customer how and why they would execute such a strategy and get implicit approval to do so before trading, thus making the real issue one of the reputation as far as the skill of the trader and depth and quality of that firm’s analytics.
The risk/reward for the client could be very positive in terms of real points saved from slippage (on large orders remember). Unfortunately, the risk/reward for the executing firm is, in the current environment, commensurately negative because it will likely result in long hours with compliance and, at the worst extreme, with the regulators.
I am not sure exactly what, if anything, can be done to resolve this situation, but what I would argue is that an open, reasonable and thoughtful industry debate around what really constitutes spoofing could help clarify the picture, for as long as there is transparency of action, there is a real opportunity to reduce transparency of order – and that helps one participant above all others, the customer.