The Last Look…

The decline of volumes on the “primary” FX venues is not a new discussion, liquidity providers have been talking to me about it for some years now, but last week’s EBS volume data served as a catalyst for intensified debate.

There was genuine shock in the industry after the publication of the data for February, it is so long ago that the platform was this low ($45 billion per day), it wasn’t even reporting data – and, of course, that number is not just EBS Market, the actual primary venue, but all spot volumes at EBS.

As I noted in our report on the data, it was undoubtedly impacted by the Lunar New Year – but then all platforms were, because it was celebrated in more than just China, and no-one has reported anything near the 28% year-on-year drop. In fact, only the other “primary” venue owner, LSEG FX, saw a year-on-year drop at all, and it fell by 11.8%, but even here (it also reports data across a variety of spot platforms) the month-on-month decline of 4.3% was broadly in line with other platforms to report, unlike EBS’ 18.2% fall from January.

Any way you want to paint this, the February data for EBS looks pretty bad, but that Lunar New Year aspect should not be over-looked. EBS became the dominant CLOB in CNH when it achieved something remarkable in the first decade of the century, in actually taking a currency pair away from its rival. Of course, there is no little irony in the fact that EBS’ big win of the past 15 years was at the expense of the other “primary”, but a win it was, and clearly CNH is important to the business, perhaps more so than many of us thought. Lower trading volumes in Asia more generally may also have contributed to the decline – EBS is also the platform of record in USD/JPY of course, but there have been some eyebrows raised at the apparent importance of CNH to the platform.

A few people picked up on one aspect of the CME report – as indeed did I – that CNH futures volume was up some 90% year-on-year in February, but I don’t think this is as notable as it first looks due to the relatively low volume traded on the Merc. It executed some $300 million per day throughout February, therefore it doesn’t take a lot to register a large leap – that said, the diverse fortunes still jumped out at some observers.

Overall though, I would be worried, because yes, the Lunar New Year in 2023 was in January, so we are comparing one month with a major holiday to one without in 2023, but in January 2023 EBS data showed no, or minimal change – indeed the January 2024 data was down almost 12% year-on-year. This is a more general trend lower, not something that is just about one month – although I am confident that the data will bounce back in March.

Either way, the EBS volume report, influenced by a calendar event or not, has crystalised the debate around market transparency and data – can we still use the primary venues for price formation?

The simple answer is probably yes, but the time when that is not the case is approaching. In spite of the troubles, and taking into account we have to guess on the actual primary volumes, Market and Matching still operate the largest firm limit order books in the market – or at least among those venues to publish data. This means they remain the most valuable source of market data.

With FX volumes growing more generally, however, these venues share of the market is shrinking dramatically. In April 2010, the share of EBS and then-Reuters of spot volumes was 19.6% (using the BIS data). Of course, in 2010, both venues operated one main venue (it is unclear whether Reuters included minor platform data); in April 2022, the time of the most recent BIS survey, this share had dropped to 8%. Even worse, this is using the full data from EBS and LSEG FX, which includes non-firm venues’ volume.

Let’s be generous and allocate 80 yards a day between the two – that’s just 3.8% of global spot volume. Even worse, the volume for both in 2023 was below that in 2022, so it could be lower now. Whichever way this is painted, it is a gloomy picture, one that has been thrust into the spotlight by the EBS February data.

There are some who believe that last look venues’ data can be used, because there is a sufficient “sanity check” from the “primaries” to hone it, but this seems, looking at anecdotal evidence, a few lone voices in the wilderness. More and more people that I talk to in the industry are looking at broadening their sources of data, and the holiday-inspired shock, or not, from the EBS data is increasing the pace of that work.

A worry of specialist or regional players, however, is that this trend, where LPs have to pay for even more market data from a variety of venues, means the big get bigger and the specialists are further squeezed out of the market. This may or may not happen, but the reality is it would be up to the end customers – many of whom tell me they are concerned about the shrinking pool of genuine LPs – to maintain those specialist players, or for those players to develop a differentiator like actually becoming more of a risk warehouse rather than turning the volume for a small profit.

CME is likely to be hoping that February is indeed an Asian holiday-inspired outlier – even though the longer-term evidence points to a continuing downtrend

There is one other aspect of this debate that is lurking in the background, and that is what is happening at CME itself. Like many, I am interested to see what the Merc’s FX Spot+ service actually delivers. The sense is growing that Spot+ could be the way CME brings its FX market data together to bolster the existing value provided by EBS. It has to be concerned at the drop in headline activity, and its potential for negatively impacting its FX data business. If EBS spot trades (even at these historically low levels) are blended with FX futures, the data side at least is suddenly based upon a lot sturdier foundations.

Whether that is the plan or not will be revealed in time, but for now, CME is likely to be hoping that February is indeed an Asian holiday-inspired outlier – even though the longer-term evidence points to a continuing downtrend. It spent good money on EBS (and Brokertec), so must be shifting a little uncomfortably in its seat, but that does not mean the acquisition is a failure.

The FX market structure continues to evolve, and one of the biggest impacts has been on the primary venues thanks to internalisation and what is seen by some as the “difficult” trading environment thereon – thanks mainly to, ironically, the level of transparency delivered by the data.

We may well be shocked by the February volume data, but perhaps not surprised given the longer-term trend. It may also well be that we look back at this period and see what was a seismic shift in how the market forms its pricing. That looks at face value for CME and LSEG, a bad thing, but the fact is both remain at the centre of the industry – it is for them to change with it, rather than draw a line in the sand. That way, they remain relevant.

colin@thefullfx.com