A few years ago I wrote what I thought was a clever article about the fragmentation of a consolidating market. In it, I tried to explain the contradiction of how more trading venues were springing up in the FX market at the same time as widespread consolidation was reducing the number of banks active in it. I argued that fragmentation was a major reason why the FX market revived and started to grow after a period of what appeared to many as almost terminal decline around the turn of the millennium.
Looking back, it now seems clear that the predictions of many – that the market would move towards an exchange model – were and remain wrong. The FX market has refused to centralise and it has continued to thrive in an extremely fragmented structure. Now, there seems to be a growing consensus that the market will never move on to a single exchange; but rather than talk of fragmentation, participants are starting to describe the market as being segmented. In other words, the numerous venues that already exist and which have yet to see much of the light of day may well continue to thrive because they all provide slightly distinct services tailored to meet the needs of a particular market segment.
The acceptance of segmentation was perhaps the main theme to emerge at a recent seminar I attended in London hosted by BT Radianz. It was meant to be about the impact of algorithmic trading in FX. It had a well-balanced panel consisting of Deutsche Bank’s Ian O’Flaherty, FXall’s Mark Warms, the CME’s Derek Sammann, EBS’s Steve Toland and Damian Mitchell, an old mate of mine from Midland Bank who has gone on to better things and is now running Dsquare, his own trading boutique. I did joke after the event that I taught him everything he doesn’t know.
The seminar was good but sadly it developed into what I can only describe as a ‘love-in’. The moderator was far too nice and the panel lacked a rottweiler-type character to rip things apart. Naturally, I asked a question to try and rev things up a bit, but it was pretty tame. I don’t know if my incisiveness was blunted by the superb setting at the Law Society, or whether I got caught up in the overall feeling of niceness that had pervaded the atmosphere.
Basically, I asked if there was really any room in the market for multibank portals. I was going to pose this to FXall’s Mark Warms, but for some reason he looked really unwell and I thought that if I picked on him too much, he’d have a heart attack. So I took a circuitous approach and asked Ian O’Flaherty if, given the fact that the banks are trying to drive client business on to their own portals, he saw the need for the likes of FXall.
Ian had earlier explained at some length how difficult it is to define liquidity and he admitted that pricing it properly was even harder. I was hoping that he’d say that the only way the banks would be able to do this would be to limit where they streamed their prices to. But he is far too savvy to fall for such a blatant attempt to stir things up and he fended the question off easily – and offered Mark a pretty easy way out of it too. Horses for courses was the conclusion, which is a colloquial way of explaining segmentation.
Afterwards Mark asked me why I was always so nasty about FXall. Me, nasty? Surely not? I think he was joking, but by this time I had begun to think his ruddy complexion was caused by anger, so I beat a swift retreat to find myself a tonic water.
FX and the equity model
Naturally, if you’re going to a talk about algos, you will hear about the impact they have had in FX. But as ClientKnowledge stated in the paper, e-FX at critical speed, it produced for the seminar, there are a couple of couple of ‘facts’ that people seem to have accepted about FX that are not necessarily true. These are that banks are about to be disintermediated and that FX is about to embrace an equity-type model and move on to an exchange.
I agree with ClientKnowledge’s rebuttal of these facts and think it is worth making a few more key distinctions about the equity and FX markets. First, the equity market is minute in comparison, and what liquidity there is split between thousands of issues. It is a fallacy to argue that the market is more efficient than FX. Spreads are much wider, liquidity is shallower and dealing costs are far higher, even though activity is often concentrated on a single trading venue.
The first electronic trade may have happened in the equity market before it did in FX – I’ll have to look that up some day – but it did not embrace this new paradigm as early or as whole-heartedly as the spot FX market did. In fact, in many cases the equity market has remained anachronistic and many exchanges have had to be dragged kicking and screaming into the 21st century. Maybe the reason why algos have had such an impact in equity markets is because they were needed in such an inefficient environment – users have had to search out often hidden liquidity to get a better fill. FX, in comparison, is a relative model of efficiency. It’s not perfect, but it’s not bad.
As for FX imitating equities in the future; in Europe at least there is a very real chance that the opposite is going to happen. The talk is all about the equity markets fragmenting, not consolidating, in the post-MiFID market place, because the bigger sell-side players are not happy with how the markets developed after the incumbent exchanges became for-profit organisations. Of course, there is also the odd moan and groan about various trading venues in FX, but it all seems to be handled so much more amicably. The advent and growing sophistication of algorithms means that the pooling of liquidity can be done virtually, so why bother concentrating it on a venue that is widely loathed and could develop into a monopoly?
