Five Observations on the FX Turnover Surveys

The FX Committee Semi Annual Surveys offer an insight into market participant behaviour, the headline act in the latest round being FX swaps providing the driving force behind the increased activity. Here are five other observations on the data that you may have missed.

1: Dealers are (naturally) internalising more

A notable feature of the spot reports (where the data is available), is a drop off in activity between Reporting Dealers from April 2025. In the UK and US reports activity fell by 14.8% and 10.2% respectively, largely contributing to the 3.1% and 9% declines in spot volume in those centres respectively.

In the FX JSC’s report for the UK market, spot volumes were up 6.6% with Other Banks, +6.7% with Other Financial Institutions, and +2.3% with Non-Financial Institutions, suggesting that the main players’ internalisation rates probably returned to what has been “normal”. In April 2025, amidst the “Liberation Day” carnage, risk was likely held by these players for much shorter duration. That month also saw two traditional venues for dealers externalising risk, Matching and EBS, see their third busiest months since the onset of the pandemic in February 2020.

The case for internalisation was not as clear cut in the US, where all counterparty groups saw a decline in activity, however other anecdotal evidence can be found in the drop in the prime brokers’ share of volumes. Speculative accounts, from whom dealers are less likely to internalise the risk, are large users of PB, and year-on-year the PBs’ share of spot volume in the US fell from 96.6% to 89.7%, the lowest it has been for two years. There was a similar move in the UK, where the PBs’ share of activity dropped from 51.5%, to 45.7%.

To a degree, there is an element of April 2025 being an outlier, at the 45-46% level, PBs’ share in the UK is in line with much of the last decade. In the US, the share has been above 90% since October 2023, but averaged around the 85% across the previous decade.

2: NDFs continue to soar

Unfortunately, only the UK and US give any real insight into NDF trading – a similar view from Singapore in particular would be welcome – but in both centres NDF activity continued to rise, with both hitting new record highs.

In the UK, ADV for NDFs was almost $286 billion, meaning activity has nearly doubled in just three years, and risen nearly five-times over the past decade. This is a 27.9% rise just since April 2025, with the growth driven across all sectors. Reporting Dealers remains the biggest counterparty segment by far, and saw activity increase by 33% to just under $164 billion per day, while Other Banks rose by 39.4% to $33.3 billion. The second largest segment is Other Financial Institutions at $71.4 billion, and although growth here was more modest at 15.2%, this is a new record for NDF volumes with the group. Elsewhere, Non-Financial Institutions rose 21.6% to $17.3 billion per day in the UK.

In the US, a counterparty breakdown is unavailable, however NDF volumes rose above $100 billion per day for the first time to hit $114 billion. This is up 45% year-on-year and is more than double what it was three years ago.

This growth has been reflected in the data released by platforms for NDF volumes, but there is perhaps an idea of the opportunity set for these platforms in that the UK report notes that electronic broking systems account for about 28% of NDF volume, with a smaller balance being traded on multi-dealer venues. Just under 10% of NDF volumes are traded via voice brokers, and while this is historically low, clearly there is still an opportunity to automate more NDF trading.

3: Singapore is getting closer to the US, but…

Although, as we noted in our initial report on the turnover data, there has been one instance when Singapore’s activity was closer to the US, there is an undoubted trend whereby the city state is edging towards the position of the second biggest FX centre in the world. Just 10 years ago, the US was responsible for nearly twice as much as Singapore and while both have since grown, the latter has done so much quickly.

This highlights Asia’s increasing importance to the FX industry, but also, at a headline level, hides some countertrends, or perhaps a slowing, or stalling, trend. Looking at the pre- and post-Covid April surveys, since April 2019, Singapore has seen spot ADV grow a massive 170%, while the US has grown an impressive, but inferior, 72%. In outright forwards (including NDFs) Singapore is up 163%, the US 60%, while in FX swaps Singapore has risen just shy of 150%, while New York has grown 80%. Albeit in the lowest volume product, Singapore’s growth in FX options has outstripped everything else at +175%, while the US has grown just under 53%.

There is, therefore, a clear narrowing of the gap between the two, and Singapore has, for the last three reports, handled more FX options volume that the US (there was a similar crossover in FX swaps in October 2021).

Looking at the post-Covid data from April 2022, however, and a different picture emerges, where growth is either comparable, or actually in favour of the US. In spot, Singapore has grown by 24.6% since April 2022, while the US has risen 39.6%. In outright forwards, the growth is slightly in Singapore’s favour at 46.2% to 44.1%, while in FX swaps it is 51.5% to 49.6%. In FX options, where Singapore has vaulted into second place, however, it has grown 27.1% over the past four years, while the US has grown 51.3% – a narrowing of the gap in the opposite direction.

So while the long-term trend is very much in Singapore’s favour, the US is showing decent pace of growth as well, so the last mile, so to speak, will be the hardest for the City State in its push for second place.

4: Can Japan rebound?

A recurring storyline in recent years has been the lack of volume in Japan, which has seen a prolonged slump, while its neighbours in the region have soared to new heights. In part the zero interest rate environment was to blame – there was so little to get excited about unless you’re a retail carry trader – but there also seemed to be structural issues at work.

The headline numbers are not great. Using the same criteria as above, in Japan, spot volumes are up just 4.2% from April 2019, more concerning is they are down 23.2% since April 2022. The overall volume is slightly more encouraging, but still underperforms with a 19% rise since April 2019, and 2% drop since April 2022. (for context, Hong Kong has seen spot rise 66% and 38% respectively and overall volume rise 48.8% and 39.3% over the two periods, while Australia has witnessed a spot rise of 9.6% since 2019 and 6.1% since 2022 and an overall volume rise of 71.9% and +31.5% respectively).

So what is happening in Japan? Well one storyline is something the country’s authorities don’t want to hear given their predilection for moaning about unwanted volatility – the yen actually isn’t moving that much. It is trapped between a market wanting to get on the carry trade following any dip and the fear of intervention that causes that dip. This means volume only really goes up when the authorities intervene.

There is more to it than that, however, and it could be the lack of real progress in automation in the centre itself. Compared to April 2022, the share of electronic trading has actually gone down in Japan from 50.8% to 48.8% with financial institutions (it has also dropped 1.1% to 66.6% with non-financial players).

Singapore, on the other, has anecdotally (sadly the SFEMC report offers nothing on execution methods), been on a successful drive to increase e-trading ratios. This has seen yen volumes in Singapore rise some 10% since 2022 (and more than triple since April 2019) while in Japan they have dropped 10%. This suggests that traders want to trade electronically, but that perhaps the majority of local banks have not established adequate enough facilities for them to do so?

Either way, while Japan has stopped the rot in terms of its volume decline, it is hard to see it keeping pace with growth in other centres until more local providers get fully onboard with electronic trading.

5: Non-banks are building US market share

There is a divergence in the trend of spot market share held by the top banks in the UK and US, suggesting that non-banks in the latter are gradually building market share.

In the UK, it is notable how relatively stable things are, in the latest JSC report the top 12 held a 95.3% share of spot volume, 77.66% is with the top six, compared to 93.1% and 76.5% respectively four years ago in the April 2022 survey. Roll back another four years and the growth is more pronounced, the top 12 held a 92.4% share, the top six was at 72.16%. This suggests that these banks are not only maintaining their hold on the market, but growing it – the top six all had shares greater than 7%. For comparison, in April 2022 at least one of the top six had a share of just 4.76% and in April 2018, one or more was at 5.5%.

In the UK, therefore, the story of the top group of banks growing their share remains intact – in the US, however, the opposite seems to be happening – at least to some degree.

The share of the top 12 banks in the US was at 88.6% in the latest FXC report, this compares to 88.9% in April 2022 and 93.7% in April 2018. The top six has also seems its share decline, from 75.4% in 2018, through 73.3% in 2022, to 72.4% in the latest survey. The non-banks are most likely to be picking this up, however…

In terms of volume, the top 12 banks are grabbing more, and the minimum share of a top six bank is growing, from 7.1% in 2018, to 9.1% in 2022 and 9.17% in the latest survey. Equally, the proportion of spot business in the US handled via a prime broker is steady from 2022 to 2026, both surveys coming in at 89.7%. Compared to 2018, however, there is a sharp rise in PB activity, in the latter survey it was just 65.9%.

On the face of it then, something we have known for some time is continuing, non-banks are busier in the US market than UK, which may have something to do with why so many buy side firms prefer to execute during European hours. It will be interesting to see if the decline for the top group’s share in the US is arrested in coming surveys.