The Last Look…

There has been a rise in the noise level recently around the growth of non-bank firms in FX, and how they are grabbing more business from the banks, including that from end-user customers – but is this chatter misguided? Does it reflect what is happening in a relatively small sector of our industry? Or is there a storyline being fed that has little basis in the bigger picture?

I am a little bemused as to where this narrative has come from, because my sense is actually the opposite – aside from some of the ECNs where non-bank firms have long had a solid presence. A few buy side firms tell me their algos are hitting more prime-broked liquidity, which suggests a non-bank LP, but when it comes to direct trading, it has not come up on my radar to any great degree.

There is little doubt that non-bank firms are making a fortune in markets at the moment (so are the banks just quietly), but a huge chunk of that revenue is in other markets, where there is a lot of dumb flow to feed off. One or two non-bank firms have “solved” FX, but the majority struggle due to the idiosyncratic nature of the market structure with speed bumps, randomisation and, importantly, the relatively low percentage of aforementioned “dumb” flow.

The data – which I accept can be shaped to make multiple cases – also does not back up the narrative of non-banks taking over in FX. Looking at the last five BIS turnover surveys, the share of hedge funds and PTFs (non-banks in the BIS’s parlance) is higher than 2022, but lower than the previous reports. The latest Triennial Survey has 12.1% of spot volume traded by this sector – and that includes hedge funds remember – and while it was 10.6% in 2022, it was north of 13% in the previous surveys. Clearly, as a broader sector, the numbers are going up but the ratios aren’t – unless hedge funds have abandoned FX, which is not what my sources tell me.

Equally, while the FXC semi-annual survey in the US has long shown a market dominated by PB accounts, in the UK, the share of spot volume that is prime broked has also gone down. In 2013, as the GFC still had an impact on credit, this was at 35.3%, but since then the share of prime broked spot volume has been in the mid-40s percent range. In the latest survey, It was down to 41.5%. Again, hedge funds may be doing less, but not to the degree that would have a huge impact on the broader data.

If non-bank firms do come to dominate, there will be one group to blame – the banks themselves

One final data point I look at when judging these firms’ influence is the data from their favourite playground, the exchanges – a place where the banks’ credit advantage largely disappears. I absolutely accept that banks, asset managers and hedge funds also use CME for FX trading, but it is still, whether the firm likes it or not, largely associated with the non-bank prop firms. That means that CME volume data could be a proxy for non-bank enthusiasm for FX – and if that is the case, then interest levels are not growing. Not only did CME’s FX volume flatline in 2025 – a year when so many OTC platforms had record years – ADV was at its lowest since 2021.

This all suggests to me that the narrative is wrong. Yes, as prime brokerage re-emerges from its regulatory-induced slump, non-banks can probably operate more on OTC venues, but they do like the exchange arena, and given CME regularly mentions growing asset manager participation in its FX markets, the inference is clear – non-banks are doing less there. If you look at the bigger picture, it seems obvious to me that FX remains a largely bank-driven business – and these firms are still dominating in the one area that really counts – revenue generation from FX.

Could this change though? Could the market structure shift to the extent that non-banks become an even-bigger presence? It would have to be quite a shift, but it is not out of the question. What would be required is more buy side firms changing how they execute, partly through more algos, which is certainly possible, but also through taking on more market, or execution, risk, as an extension from their algo use.

These firms interact with non-bank liquidity in equity markets especially (although they still largely favour dark pools or large blocks), so it’s not a huge leap to embrace this in FX as well, however there is a problem. As noted, the FX market structure is different, and liquidity is deeper, and the FX trade is largely all about the admin for these firms, rather than Alpha. Why get involved in trading that induces slippage when the risk can be handled much easier in a private, internalised, environment?

So I still see banks as the dominant force in FX, and given the credit bottleneck is not really an issue in spot, it is hard to see how that changes. If it does, however, there will be one group to blame – the banks themselves. At a micro level, their business models have brought them into a technology war with leaner, more-nimble, firms, who can always be faster. The banks differentiator was always credit and risk absorption – the first remains but has been diluted, the latter remains crucial. For if banks stop acting as a risk warehouse in FX, and become even more of a broker (through internalisation admittedly), they move into even closer competition with the non-banks – and that is a contest in which they don’t have a significant advantage.

That said, the banks are making more money than ever in FX, so do they care? Perhaps they will if some of their blue-chip clients start doing more with non-banks?

colin@thefullfx.com