This week sees yet another change in the platform world with, as discussed during our first Dial-in Day in April, Refinitiv Matching releasing a 5ms data stream, bringing it in line with the top feed at EBS Market. This is the latest in a mini-rush of changes at ECN/platform level, following Euronext FX’s change to how tags are shared and Cboe FX launching a CLOB based on firm liquidity.
Aside from rearranging the deckchairs on a sinking ship – I have long held that the anonymous channels will struggle to grow, or even maintain current levels, as disclosed channels thrive – I am trying to work out what is driving (apart from coincidence perhaps) these shifts.
The cynic in me says it’s amazing what a bit of noise around banks studying their connectivity network to platforms can do, especially when volumes are not going up to any serious degree, but then there is also the suspicion that some out there feel the primaries are more vulnerable than they have been for some time – even though it may only be grabbing volume from one platform to go to another.
Just as the changes at Euronext FX need to have the foundation of adequate monitoring – and punishment of breaches – so too does Cboe FX’s move require careful application. Several people I have spoken to have raised the question of fairness over some market participants getting market data quicker than others – it is creating an unbalanced playing field at a time when most of the industry’s efforts are aimed at the opposite – but is it really any different from, for example, EBS providing access to a premium feed once a certain volume threshold has been surpassed?
What is interesting to me is how Cboe is using market data as its primary weapon of choice – that is something of a shift in how platforms have traditionally approached matters. My concern is that a CLOB rewarding players for being top of book (with a minimum quote lifespan) but with that TOB able to be in just 500,000 units, how much real liquidity is actually on offer? An LP can layer bids and offers to get around the MQL, keep them in small, and rake in the market data for what is, in reality, a pretty small effort at improving overall market liquidity.
Don’t get me wrong, anything that removes or reduces last look to a simple credit check is a good move in my book, but this seems to speak to the smaller, nimble players, more than it does what you would term a serious liquidity provider willing to post serious interest to a CLOB.
Cboe’s FX volumes have proved pretty stable over the past year, take out the surge in February and March and it continues to print around the 30 yard mark each day, below what it was in 2018 but slightly above what it was in 2017. Within that data the sense is that firm liquidity has been growing, from around 30% of volume to 35-40%, hence why the platform is formalising this model in an all-to-all venue, but it probably is also a function of it sensing vulnerability elsewhere.
With Refinitiv and EBS CLOB volumes in what seems terminal decline, perhaps Cboe sees the opportunity, however it is worth noting that at the $10-12 billion per day mark, its firm liquidity is nowhere near close enough to really challenge those venues, indeed it probably lags LMAX Exchange as a firm liquidity venue – and perhaps therein lies the reason for the move, it knows it needs to do something to kick on now.
Looking at the broader picture, in spite of the fact that I think these channels are never going to regain their former glories, it is heartening to see (albeit small) shifts in how participants are expected to behave. The Refinitiv move to 5ms should (and I am not holding my breath) trigger a broader shift in the use of last look – with data available from the primary venues at such a rate, why would anyone need to hold a price for longer that 10ms? If nothing else, if I were a liquidity consumer (who did not abuse liquidity I should add), then I would be asking my LPs why I am being held longer that 10ms anywhere? The answers could be enlightening.
Cboe FX is looking to reduce the reliance on last look, which is also good, the more firm liquidity we have as an industry the more accurate picture of market conditions we can build. Equally, Euronext FX is building upon its hosted last look service with the removal of pre-trade IDs – again, a positive move that represents a small shift towards a cleaner market.
I would hope, unlike their attitude to pre-hedging and the Fix, that the buy side is fully engaged with these changes and will ask the questions of their LPs about how they behave. What all these moves have in common is the background of rewarding “good” LPs (definitions vary of course) – which we should all welcome. What is needed then, is for the customers to get onboard by challenging not only their LPs, but also the multi-dealer platforms they are using, about exactly how good their ecosystem really is.