The Last Look…
Posted by Colin Lambert. Last updated: January 24, 2023
One of the challenges facing anyone trying to launch a new FX platform, or indeed to seriously grow an existing business, is that it is just so hard to get clients to move. The desire is rarely there for change, but even if it is, the logistical load is way too heavy for most firms. Generally, then, the FX platform business is a stable one, so it is interesting that two firms have apparently entered 2023 with changed brokerage structures, albeit for different reasons. What does this tell us?
For the first, FXSpotStream, the rationale seems fine because, I am told, the firm has moved to a brokerage, rather than a subscription model. The model where everyone pays an annual fee worked when the platform was in growth mode, but as things settle down, there were clear inequities. A couple of LPs absolutely loved the system – they had such a tremendous market share their bro-per-million was almost in cents, while others, who were winning less business, were paying a huge amount per-million (one LP suggested it was around $60).
This clearly couldn’t continue, because at those high levels the value of the business won is seriously diluted and, inevitably, the flow coalesces around a smaller and smaller group of LPs, which is not, importantly, what the clients really want.
The other platform to change its brokerage levels is, again I am told by several LPs, FXall for its RFQ LPs. The platform has apparently raised spot brokerage to $10 for RFQ. And it is charging $5 for fixing orders and “automated send orders” (direct to an LP) and $4.50 for algo and resting orders. Discounts are available as more brokerage is paid.
This is an interesting time for both platforms, because one, FXall, is in the midst of a long process of re-platforming, while FXSpotStream is widely considered to need a technology upgrade if anecdotal evidence from my e-FX sources is anything to go by. It could be that both see this as a way to increase revenue to help that process along, but it is more likely a reflection that the liquidity consumers are not going anywhere – and in the case of FXSpotStream, that the way it charged simply needed changing.
For FXall, with competition amongst platforms as hot as it ever has been, it looks like a counter-intuitive move to raise brokerage, but it has clearly been thought out, and there has to be an underlying confidence that little will change.
It is interesting, however, that the increases come against a background of volume stagnation. The decline of volumes on the primary venues has been well-documented over the years, but from 2016, ADV on Refinitiv’s platforms has dipped, before recovering in 2022. There is the frustration that the firm does not break out Matching and FXall volumes in its monthly data, but the sense is that the rise and fall of ADV on the firm’s venue is largely linked to the fortunes of Matching – FXall has a steady profile.
In 2016, Refinitiv printed ADV a fraction under $99 billion per day; in 2022 it was a fraction above. In between there have been highs ($99.7 in 2018) and lows (sub $88 billion in 2019, 20 and 21). The broader picture says little has changed and efforts to grow volume in spot haven’t been that successful – thus, if you want to increase revenues, there is only way.
Of course, it is not only in spot that the pricing structure for FXall is changing, LPs tell me the changes go out the curve in forwards and FX swaps. Again, the re-platforming is bubbling along in the background, and again, while the last few years have seen volumes continue to rise, it has been at a much slower rate. In 2015-17, Refinitiv was printing around the $280 billion per day mark in non-spot products, this jumped to the low $330 billions in 2018-19, rose to $341 billion in 2020, and has been $356 billion and $353 billion per day respectively in 2021 and 22.
Perhaps the biggest differences in the spot and swaps world are the existence of competitors who seem to be gaining serious traction, and the broader growth in FX swaps and outright volumes generally. Spot volumes have risen by some 27% from 2016 to 2022, according to the BIS Triennial Survey, something few platforms have managed to match (no pun intended). In non-spot, the difference is even more stark, with volumes rising over that period by more than 61%, or to put it in notional terms, by almost $1.9 trillion in outrights and swaps.
Probably the biggest challenge to FXall in this area is 360T, which has seen impressive growth in its non-spot business over the past years
On the competition front, things are also heating up. CME’s FX Link is showing year-on-year growth, although the notional numbers are still low; while the Merc’s EBS also talks of year-on-year growth in forwards and swaps, but does not provide actual numbers, and while it could, reasonably, be anywhere, the chances are the notional values remain low.
Probably the biggest challenge to FXall in this area is 360T, which has seen impressive growth in its non-spot business over the past years. In 2022, 360T saw several record volume ADVs in its non-spot business and ended the year with an average ADV of EUR 119 billion. To put it into perspective, in 2016, when it started reporting data, ADV was EUR 57 billion, it rose to the mid 60 billions in 2017-18 and to the mid-80 billions in 2019-20. In 2021, it was just over EUR 93 billion, meaning the 2022 increase was in excess of 27%.
Traders I speak to in the FX swaps space are conflicted by both Refintiv and 360T’s offering, both seeing features they really like in one, that they wish was with the other (building in Refinitiv and mid-match on 360T are the most common). There is a preference for 360T’s tech, however, as noted, Refinitiv is in the midst of a re-platforming; so assuming that doesn’t take too long and doesn’t offend dealers’ sensibilities (which is hard to do) there is no clear differentiator.
This puts the brokerage in the spotlight to a degree, for when LPs are talking to buy side firms, especially those looking to trade swaps in a multi-dealer venue for the first time – and there is potentially well over $500 billion of volume up for grabs – they are going to inevitably push them towards the one that is cheaper. So how does that look now?
Looking at the FXall MTF rate cards provided to me by LPs and comparing them with the 360T rate cards, there are differences, but they are nuanced. What is clear, though, is that FXall has removed a potential point of difference in that it was cheaper.
In FX swaps, 360T charges EUR 0.15 per day up to a maximum of EUR 4.50. This means that one-week bro is EUR 1.05 per million euros and three-months at the maximum 4.50.
FXall, meanwhile, is now charging its platinum LPs $0.25 per dollar million for one day, $0.50 for two-to-three days and $1 for the one week. In the one-month brokerage is $2.25, in the three months $4.25 and over six months (the structure is in tenor buckets) it is $6.25. At the time I was first made aware of the FXall changes (early November 2022), EUR/USD was trading around parity.
Any time something becomes more expensive, competitors have an opportunity and will not waste it to remind participants that they are cheaper and (according to them anyway) better
This suggests that Refinitiv is going harder for the high volume short-date business, which makes sense given its lack of mid-match functionality, which becomes much more important further along the curve, but the difference even at the short-end is pretty negligible ($1 to EUR 1.05 in the one week for example), whereas further out, there is a much wider gap. It is also of note, that those FXall prices are for platinum LPs, my understanding is that silver and gold LPs pay about 20% more.
This is, therefore, an interesting time for FXall to be raising prices, given it is facing increasing competition and its technology is, frankly, creaking (I have been told of at least three outages in the past six months). To go back to the original theme, however, this probably also signifies the company’s belief that it has a very loyal customer base, and it probably does – as do its peers.
Is this loyalty going to disappear? I very much doubt it, because it hardly impacts the buy side at all – collectively they don’t pay brokerage. As happens with its immediate peers in the B2C space, Refinitiv is relying upon the fact that the LPs want to be in front of their clients, and will continue to pay to be so. Some of those LPs are, I think it is fair to say, very unhappy with the increase in fees, but there is an air of resignation among them that nothing will change.
It is, indeed, hard to see how things could change, however there are a few things to keep an eye on.
Firstly, more buy side firms tell me they are more conscious of how much it costs their LPs to be on a venue than ever before – the analytics delivered by those LPs offers some interesting data apparently.
Secondly, as a group, the multi-dealer platform market, with the exception of FXSpotStream which was very much in growth mode, and 360T’s non-spot business, has failed to keep pace with the growth of the wider FX market. In other words, new business is successfully being diverted away from these venues by LPs convincing their customers of the value of cheaper access.
Thirdly, connectivity with downstream systems continues to improve and the more ‘open source’ the world gets, the easier it is to trade in one place and process elsewhere. The workflow aspect of the business is becoming more important, because most buy side firms recognise spreads are about as tight as they are going to be for some time. This has been an important differentiator for FXall (and FX Connect) in the past, it may not be quite as much in the future.
Fourthly, this could, if one of my predictions for 2023 comes to fruition, mean that clearing becomes a key factor. Both Refinitiv and 360T have access to in-company clearing houses, so could they become a factor later this year? Equally, does Refinitiv have an advantage in LSEG’s recently-completed acquisition of Quantile and its compression services?
Lastly, there is the influence of technological change itself. How long before processes emerge to enable clients to switch platforms? Broader society is all about changing providers on a regular basis to get the best deal, if technological processes advance to the stage that it is little more than a flick of a switch to add or remove a platform, how long until clients listen to their LPs and take the plunge?
For now, in the short-term, there is unlikely to be major upheaval in the market, therefore raising brokerage is a relatively low-risk move. That said, any time something becomes more expensive, competitors have an opportunity and will not waste it to remind participants that they are cheaper and (according to them anyway) better.
FXSpotStream is clearly in a state of flux, with a change at the top and different views on how the business should progress from here, so there is obvious risk there, but the move makes sense.
For FXall, the pros and cons are a little more nuanced, but the sense is, even with 360T definitely growing, that it will be OK, unless there is a problem in the re-platforming. If that happens, then all bets are off, and it will no longer have the strong support of its LPs to fall back on, therefore the pressure has just ratcheted up on the technology delivery teams.