I am struggling to find someone who was not surprised by the BIS data on Monday, I think most of us who study the FX committee semi-annual data closely expected somewhere around the $6 trillion mark, but obviously it blew through that by another 10%. This very much reflects the growth in emerging markets to me – volume that is not easily captured in the committee data – and was obviously heavily driven by swaps volumes (although the BIS report makes a point of mentioning the growth in NDFs under the Outright Forwards category, but does not break NDF volume out). Either way, exchange rate volatility may be suffering but interest rate volatility certainly isn’t – the surprise was, as noted, that the increase wasn’t flagged in the semi-annuals.
As is my wont, and it does seem to get trickier every three years, I like to look at the relative performance of the public trading platforms through the prism of the BIS survey – I say trickier because so many venues report volume differently that it is hard to really match up the data with the BIS because we don’t know how much is spot and how much other products.
One thing we do know, of course, is that the primary venues do report spot only data – and they have really suffered over this three year snapshot. In many ways this could be what has thrown so many people over the high number overall, because April was the worst month for several years for the primary venues – yet over the same period spot volume globally went up. According to the BIS spot volume rose 20.3% from April 2016 to April 2019. At the same time, EBS saw an 18.2% decline; Thomson Reuters/Refinitiv a 16.5% decline; and CME activity was 3% lower.
So the picture isn’t pretty for those venues and, as you might expect, some of the newer venues look a lot better, mainly, I would argue, because they were still in their early growth phase during this period. FXSpotStream is up 78%, while Fastmatch/Euronext FX is up 48.1%. Probably the more impressive performance comes from Hotspot/Cboe FX and 360T because both were well-established ventures in 2016. The former saw a 23% increase in activity across the period concerned, while adding GTX’s figures into 360T’s for April 2016 highlights an 18.7% increase for the combined venture.
I don’t think we should be surprised by 360T having a strong showing, because it has historically been very much an FX swap-dominated venue, indeed Thomson Reuters/Refinitiv saw non-spot volumes rise by an equally impressive 23.6% over the same three-year period.
What is interesting to me, however, is the overall data of mainly spot (or equivalent) ventures. The combined notional volume traded by EBS, Refinitiv, CME, CboeFX, EuronextFX and FXSpotStream actually fell over the three-year period, by 2.6% to $301.8 billion per day. Making the (unsafe) assumption that all of this volume is in spot, that means these platforms handle around 15% of activity every day – which means there is a heck of a lot of activity going on elsewhere.
The obvious assumption is it is going direct, probably through aggregation, although we will not find this out until December when the BIS Quarterly Report publishes more in-depth data and analysis of the survey. I would wager, however, that the picture to emerge will reinforce the value of internalisation as a differentiator in the FX execution world.
There are another couple of initial observations I would like to make around the survey. Firstly, we all know that the FX Global Code is struggling to have an impact on the buy side and certain other sections of the market such as retail aggregators and some non-bank market makers who exist lower down the ladder. We can, making yet another assumption of course, put a number on this now and it’s $2.37 trillion – that is the amount of turnover reported with institutional investors, hedge funds, retail aggregators and corporates. Assuming that all reporting dealers and non-reporting banks have signed up to the Global Code, this means that 36% of volume is executed by non-adhering participants. The hope is that in the 2022 survey, this percentage is a lot, lot lower.
The other – and last for now – observation, is that the huge growth in prime brokerage – it almost doubled – accurately reflects the diminishing role of the banking industry as risk warehousers. There are, of course, exceptions at the top end of the tree, but this growth in PB is probably explained by the rise of the non-bank market maker and the share those firms are grabbing at the expense of the regional or second tier players.
We can debate long and hard whether this is a good thing or not, either way, however, it is a reality. It might be interesting for the senior management teams at the banks to look closely at this data, for it reflects the erosion of their position in the foreign exchange market at the hands of firms with generally an easier regulatory regime to which they have to adhere, firms with better technology, and, importantly, firms that are likely to be customers of their prime brokerage business.
I remember thinking many years ago at a Managed Funds Association conference how strange it was that hedge fund managers in the listed space thought of the banks as brokers, and named them as such. In FX that hasn’t been the case to date, but I wonder if the BIS data is signalling the start of that shift here also?
I happen to believe that there is a vitally important role for banks willing to assume risk in foreign exchange markets, however it often feels as though the senior management at these institutions don’t see it the same way. That is a shame because it weakens the fabric of the market, because if there is one thing I am confident of in this every-changing and chaotic world it is that if more and more banks go the broker route then the people to suffer at some stage will be the same clients they purport to exist for.