So, do we need to talk about the primary venues? The ADV data from CME Group (which obviously includes EBS) and Refinitiv makes for pretty grim reading with the latter two reporting the lowest daily turnover since they started publishing data in 2007 and 2009 respectively, and the former being in the fifties of billions for only the fourth time since I started tracking the data in January 2007 (April 2020 was actually the lowest for CME in that period).
Is there a problem here then? Obviously if you work at a competitor firm you are going to say ‘no’, but bearing in mind the BIS Markets Committee paper three years ago that highlighted how important the primary venues were for price discovery, perhaps the answer for the broader industry is ‘yes’?
Of course, as data, technology and analytics become even more sophisticated, the days of pricing off the primaries is being diluted, but talk to enough e-traders and they will tell you that a bid or offer on a primary venue is very quickly matched, or improved upon by the smallest incremental tick, on the other venues. I sat with someone earlier this year who showed me how it works – and this was not the first time I had been shown it – and it is, quite frankly, obviously.
These venues are, therefore, still important and it’s because they represent the purest form of the price available. Most banks and even one or two of the larger non-banks now, are running strong internalisation regimes, what filters down to the primaries is the net of that flow, making it a true representation.
Talk to people about the primaries and often I am told they have been badly hurt by competition, but I am less sure. To me, this is a story of internalisation – as the banks especially got better at it, there is less need to post interest to these venues, and volume drops. To go back to the BIS paper, this decline in activity doesn’t make them any less valuable as a resource, it just negatively impacts the revenue stream from brokerage.
The decline in numbers does make for stark reading, however. In this week’s podcast I briefly mention that in May 2011, so before Thomson Reuters bought FXall, those two venues, plus EBS and CME handled ADV of $557 billion. Now that included non-spot products in the case of FXall, so let’s knock it down by 57 yards (it reported $86 billion that month) and that’s 500 yards – half a trillion dollars a day.
In May 2020, CME, EBS and Refinitiv reported spot ADV of just over 187 yards. So more than $300 billion has disappeared from those venues. Other platforms have come into being in the meantime, but none are handling the type of volume to make up that gap, and none of the existing platforms have grown by anything like that. Look at the BIS data that nearly matches that time horizon – from the 2010 survey to the 2019 report, spot ADV grew by $500 billion – so it’s certainly not a reflection of the market cooling off.
People do genuinely ask me what I think has gone wrong with the primary venues and my answer has been the same as it was almost a decade ago when I had a rather loud “discussion” with the head of one of these venues in the middle of their (open plan) office. I don’t think anything has gone wrong per se it’s just the market structure, and the advance of technology, means that most players in the market are internalising more, which means, inevitably, less volume goes there.
Last year (it was maybe the year before, I write a lot of these things!) I wondered aloud about the primary venues going brokerage free to see what would happen to the competition. I also mentioned ramping up data prices if more volume was attracted, because it would, commensurately, be more valuable. I am, not for the first time, starting to question my own thesis and actually pondering what would happen if they increased brokerage?
At first glance it would be a silly thing to do in the face of such a fragmented market, but given the aforementioned pure nature of the beast, the fact that the data is cleaner than venues where there is last look and recycled liquidity to pollute the water, maybe it’s not so silly? If nothing else, it would be interesting to see what would happen to ADV on these primary venues – would customers vote to exit with their business? Instinctively you would say yes, but who really knows until it’s tried?
Of course, making such a brave move in the current circumstances is, as I mooted with going brokerage free, unlikely, but the very fact that the numbers are hitting new lows means that surely something has to be considered? I understand when people say to me ‘look at March’ when their numbers were impressive (but by no means as impressive as they were in, for example, 2011), but equally I respond with ‘look at the percentage increases’. At EBS the March ADV was 57.8% increase year-on-year, seriously impressive and better than the 42.9% increase recorded at CboeFX, but it was lower than, for example, Euronext FX’ 76.2% increase (although those that like to see that firm’s Tape as a benchmark would observe reported volume only rose by 28% across that period).
I understand the numbers are bigger at the primary venues – Refinitiv was up 45.3% year-on-year (360T was +70%) and CME was up 31.3% incidentally – but either way, they even seem to be losing the advantage they once had when markets get really busy.
Ultimately, I am not sure there is much that can be done to reverse the decline, and as long as the data retains its high quality the venues will probably still be important. I continue to believe that internalisation is the biggest problem for these venues, but also thrown into the mix should be the fact that there is, thanks to the FX Global Code perhaps, more awareness of how difficult it is to monitor behaviour on these venues. Perhaps greater clarity over behaviour would help, but all have pretty high profile behaviour monitoring resources and are well aware of their responsibilities.
One final aspect to consider is another old favourite of this column – the decline of anonymous trading generally. The 2016 BIS report had electronic-indirect volumes (so generally speaking anonymous) at 22% of turnover; in 2019, it was 16% – during which time, of course, overall volume went up, by around 20%.
So there are multiple factors at play here, all of which may be concerning to the owners of these venues, but the reality is they are the result of fundamental shifts in how the FX markets operate. The bottom line, as far as I am concerned, is that while competition may be playing a role, it is nowhere near as much as some people think – if indeed it has a role at all.