The Last Look…
Posted by Colin Lambert. Last updated: April 3, 2023
“Unique liquidity” is a phrase I really don’t like, and it has to be said I often view it as one used by those venues or providers giving away their desperation to be seen to be at the top table – generally speaking, the major platforms rarely talk about it. Equally, I have never really accepted the “unique flow” argument from platforms, typically viewing the issue as two-dimensional between bilateral disclosed flow, and anonymous ECN flow. I may, not for the first time, have under-estimated, however.
There is good reason for my distaste of the “unique liquidity” mantra of course, in many cases the LPs on the venues are from a select group – another favourite pastime of mine used to be totalling up the number of LPs claimed by platforms, all I can say about those numbers is the qualification bar must be pretty low – and increasingly clients seem content to connect to more than one venue either for aggregation purposes or technology redundancy. This dilutes any uniqueness.
One area where platforms are, to a degree, differentiated, is in the client base, however, as noted, I typically have segmented it into the bilateral relationship venues and the ECNs, accepting there are some crossovers.
I raise this, apropos of nothing, because I was looking at the first, and most comprehensive, set of data available from the platforms in March and saw they threw up some interesting juxtapositions.
I would have thought that similar venues, for example in this case Euronext FX and Cboe FX – and even CME – would have had similar volume patterns, but it is anything but. March was clearly a busy month – Cboe hit its second highest ADV ever – but what struck me about the daily data was how the peaks and troughs between the three venues were so different. As an example, taking the daily data available from all three, they all had their busiest days on different dates. CME hit its FX futures peak on 8 March with more than 2.8 million contracts (probably between $255 and $260 billion notional), while Cboe FX peaked on 21 March (at $73.2 billion), and Euronext FX hit $36.3 billion on 13 March. Interestingly, 360T’s peak came on the same day as Euronext FX, the platform hit $37.4 billion on 13 March.
The same was true of each platform’s quietest day, CME’s FX futures had just over 580,000 contracts on 22 March, while Cboe FX hit $33.3 billion on 28 March and Euronext FX $16 billion on 3 March (360T was 6 March at $20.8 billion).
Looking at how busy markets were, both EUR/USD and USD/JPY one week realised vol peaked on 16 March and second highest days were on the 13th and 24th of March, while Cable peaked 13 March and was again on the 15th and 16th, as well as the 7th. In rate terms, EUR/USD had a decent dip on 6 March, was up and down about two big figures from 8-15 March and then climbed some three big figures 15-22 March.
So what does this tell us? One observation is that when markets were volatile with two way moves (over the course of a day or two), CME’s FX futures hit their peak. Likewise, Euronext FX’ top three days were in this period, which could suggest there was a lot of short-term speculative interest taking advantage of short-term volatility. Cboe FX on the other hand, had its three biggest days during the euro’s mini-bull run from 15 March, suggesting, perhaps, that more trend seekers are on the platform.
It will be interesting to see what the data looks like for another month when we have volatility and volume spikes as we had in March
It is a shame that other platforms don’t provide such granular data, it would be fascinating to see, for example, how EBS Market and LSEG FX’ Matching compared, and how FXall and Currenex compared to 360T? Looking at the latter, there is no real pattern suggested, and of course it is impossible to know where the extra volume was traded – GTX or the relationship platform – although it could be that the spike in volatility prompted more hedging activity on the 23rd, a day when other venues were close to their average ADV for the month.
It will be interesting to see what the data looks like for another month when we have volatility and volume spikes as we had in March; for example is, as the data may suggest, CME’s FX futures something of a leading indicator of activity, peaking as it did in the days before Euronext FX, 360T and Cboe FX did so? Certainly on the Merc’s busiest day, Euronext FX had only its 12th busiest day of the month (with ADV slightly below the monthly average), Cboe FX its 10th (slightly above its monthly average) and 360T its 16thbusiest.
Overall, if this is part of a pattern (it’s something I haven’t really looked hard at previously), then once again I can be disabused of a notion about the FX market. I have always accepted that there are many different reasons to trade FX, but I have generally split platforms into active traders and hedgers in terms of their client base. This may be wrong, even within the ECNs there may be different characteristics to the client base that I have previously dismissed.
I should therefore, probably apologise to every speaker on one of my panels over the many years, who have, many times, claimed to be unique – perhaps there is something in it? I won’t of course…nobody turns up at a conference wanted to be “sold” at!