Is it time to really start to get worried about the primary FX venues? I have written before about the decline in volumes at these venues but one could be forgiven for thinking the decline is now becoming a little precipitate. Both Refinitiv and EBS reported their lowest spot ADV since they started reporting (so pretty much forever) and if you include CME as a primary venue, which can be a little tricky given the volume data includes options on futures, even there the decline is visible.
In terms of numbers, looking at the annual averages since 2013, so two BIS surveys ago, EBS is down 27.5% from 2013 to 2019, while CME is down 25.3% and Refinitiv (Thomson Reuters), and they may want to look away now, is down 32.3%. There are some caveats to these numbers, namely, the Refinitv data includes FXall volumes so is not all primary venue data, and CME Group will get a boost to its ADV in December with it being a roll month, although it is unlikely to add more than $2 billion per day to the annual average, either way though the steady decline since the salad days of 2010-13 seems inexorable.
Even if we look at the last survey period, as I wrote a few months ago, the primary venues don’t come out of it well with EBS down 10.4% from 2016 to 2019, while CME is down 3.3% (which could be wiped out by a good December roll month) and Refinitiv is down 12.2%. At face value it may look as though over the past three years the pace of decline has slowed, however look at the bigger picture and you will see it is broadly the same. According to the BIS from 2013 to 2019 spot volume decline by 2.9%, however, from 2016 to 2019 – when the primary venue ADV was down less – it rose 20.3%. The broader picture still indicates platforms underperforming the BIS benchmark by between 23-33%.
In terms of why the decline continues, especially against what remains for some the mysterious backdrop of that 20.3% rise reported by the BIS, there are as many theories it seems as there are over what drove me to learn to dance the tango (one for the podcast listeners!)
Generally speaking, the decline is put down to one or a mixture of the following:
- Less control over liquidity on the part of the major LPs, who find their pricing recycled to secondary and tertiary venues (and this may be changing as I have discussed in recent months)
- Lower volatility and the consequent increase in internalisation rates by the LPs
- The rise of the non-bank market maker (typically a firm with good connectivity across the spectrum and the ability to cancel, or use last look, much quicker to protect themselves from sweeps
- The primary venues cannibalising themselves to a degree by launching different liquidity models, such as EBS Direct, Select etc; or buying an aggregation style platform that comes with an ECN (Thomson Reuters and FXall)
Even looking at 2017, which was widely seen as the low point in recent years for the spot FX market, these venues have seen further decline, with EBS down 7%, Refinitiv down 3.6% and CME down 9.8% (again with the caveat that we are currently in a roll month so ADV will be stronger than average), so should we be worried?
Unless the world has changed in the past two years to the degree that most market participants no longer rely heavily upon this data for their pricing and risk engines (not to mention the rather important factor that is TCA), the answer should be ‘yes’ because we are getting to the stage whereby the price discovery importance of these venues must inevitably be undermined by falling volumes. After all, if you buy the price discovery aspect of the argument then at face value one primary venue in EBS handles around 3.9% of spot volume and another, Refinitiv, handles around 4.4%. At what stage does using these venues as a primary source of data become too risky for the LPs and too thin for the execution quality measurers? Especially if one considers that the percentages I have quoted above are for all spot venues, not just the primary venues. If I had to take a guess I would suggest the primary venues themselves see something close to 2.5% of daily spot volume and that compares poorly to 2010 when (using the BIS survey as the benchmark) they handled around 10% of activity – and the 2010 data does not include venues like EBS Direct and Thomson Reuters did not own FXall. For the last two surveys EBS has handled around 5.2% of spot volume and Thomson Reuters/Refinitiv around 6.1%, which means their likely share of the market has halved in the past three years and has fallen to a quarter of what it was just under a decade ago.
Of course, there are people out there who will argue that this is all irrelevant, but my sources tell me we had a good instance in the past two weeks which highlights the importance of these venues. I am told that Refinitiv had to take Matching down two weeks ago today (Monday) and speaking with dealers (both takers and makers) they almost to a man and woman say that pricing in the platform’s core currency pairs disappeared on the other platforms as well – if that doesn’t highlight exactly what the BIS Markets Committee was writing about in 2018 then nothing does.
We find ourselves therefore in an interesting position – the primary markets seem trapped in a downward spiral in terms of volume traded, but they remain highly relevant to the market functioning. The latter does not, of course, help these firms trying to grapple with what must be a diminishing bottom line from their spot businesses. Personally – and I have stated this before – I think the FX market is a better place for having reasonably strong primary venues and it has always been thus. With apologies for going all “old school” on you (again I hear you cry!) even back in the voice days while there may have been five or more inter-dealer brokers in each currency pair in each centre, there was always a general consensus on who dominated was for each currency pair – and price was often formed from that “primary” source. In other words, nothing has really changed except for the speed and the name of the primary venue.
As to what can actually be done about it, that is a lot harder. I have argued that one or both of the primary venues should get rid of brokerage (subject to volume thresholds) across the board and rely upon data sales – that being clearly their most important service, but at the end of the day it is probably also up to market participants to shift their business there, and we cannot dictate where people trade.
It could be that all we actually need is a return of volatility (good luck wishing for that!) but in reality it is probably more than just that. If there is one thing that I believe has undermined the primary venues it is the growth in aggregation platforms among regional and second tier banks – this trend means that the primary venues no longer occupy the most prominent real estate on dealers’ screens. I am not sure how we get these dealers looking to post their interest to go to the primary venues again when most of the top LPs allow them to post to their internal (and generally larger) pools of liquidity for either no, or minimal cost. Equally, as these regional or lower tiers players increasingly become customers of a small group of LPs they need the primary venues less and therefore they care less about whether they exist.
Where I think the real problem for the primary venues lies is in their relationship with the major LPs. Many, many years ago I wrote about the challenge for multi-dealer venues from the growth of internalisation, specifically that their top clients – the major banks – were now becoming their competitors. Looking at the data and talking to people in the market it strikes me that what we may actually be witnessing is that theory (finally!) coming right.
If that is correct then the real turning point for the primary venues will only come when their competitors and largest customers (who are one and the same of course) decide that enough is enough and the decline is starting to hurt their business. As to when that happens – if it happens – that is anyone’s guess.