So you’re a software programmer, looking to make a few bucks in these trying times, and someone comes to you asking you to programme a trading strategy for them to use. No problem, get the specs, do the work, walk away with the money – or do you?
We may, finally, have had closure in the Navinder Sarao “Hound of Hounslow” case with the settling of the CFTC’s action against Edge Technologies, the firm Sarao employed to programme his spoofing strategy which, ridiculously, the US government blamed for the Flash Crash in US equities in 2010. The programmers involved have been largely left unscathed by the settlement and the soap opera around the US authorities pursual of Edge CEO Jitesh Thakkar, which also appears to have reached a conclusion, but a couple of lines in the settlement would make me think twice if I was a programmer helping to build such a strategy in the future.
Firstly, the Consent Order states that both parties, in other words Edge and the CFTC, agree that the software firm knew Sarao planned to use the software to spoof. More crucially, it states that Edge “knew, or should have known and understood that market participants would react to the false signals communicated by the Spoof Orders Trader A intended to place with the Back-of-Book function and would use that information in making trading decisions.” [my italics]
Trader A is, as is widely known, Sarao, so the Consent Order now puts on the record that the US authorities expect software programmers to fully understand market functioning, regulation and the appropriate laws. This requires them to proactively gain this understanding and while I have no intention of getting “generationalist” over this, I have met plenty of young software programmers who exist in dark rooms and have little or no real understanding of the broader issues in this world like the law or market regulation. As a speaker once famously said on one of my panels, these people “geek off” in dark rooms and rarely come out!
So programmers need to understand this stuff better and I think, generally speaking, those employed at trading firms do – it’s part of their training by the institution. The challenge is for the smaller, private firms, who are springing up thanks to job cuts in the institutional world – all of sudden you see these firms having to employ compliance specialists which changes the economics of the model drastically.
This is not necessarily a bad thing, of course, we do need to ensure that the market ecosystem cannot be easily undermined and while spoofing is a nuisance, the same neglect of market rules could allow someone to do something much more serious.
What Sarao was doing was obviously dodgy, as highlighted by the programme automatically cancelling the balance whenever an order was hit for a portion and his insistence on adding or subtracting one lot from the order every time he was joined on the bid/offer to move to the back of the stack. For fans of the TV show Little Britain, this was a “computer says no” strategy.
By achieving an Order against the firm, I suspect that the CFTC feels it has kept the door open to other actions.
There is one other, more broadly philosophical, factor here as well and it’s about transparency. I think people are starting to understand the downsides of too much transparency in markets – don’t get me wrong, we need total transparency of action, but the very fact that Sarao’s strategy worked so well illustrates why so many people like me have an issue with transparent order books.
Clearly the market does react to the placement of large orders in the stack, and participants are not to know whether it is a spoof or not so they act as though it is a genuine order. What this tells me is that every time someone tries to place a decent sized bid or offer in a totally transparent market their chances of actually getting a decent fill diminish significantly.
All of which begs the question, have we heeded the real lesson of these spoofing cases – that best execution is highly unlikely to be achieved in a totally lit book?