The Last Look…

The response to last week’s column drew an interesting picture of how some people see the modern FX industry, specifically, the impression on the part of more than a few that the LP landscape has changed quite dramatically.

There are, no doubt, vested interests at work, but it was notable that a large proportion of those providing feedback on the higher reject rates in August pointed the finger of blame at the growth of non-bank market makers. There is no doubt that some LPs are more prone to leveraging the last look process than others, but it is far too simplistic to lay the practice at one door. I recall some time ago when looking at reject rates with a buy side consumer, that one or two banks were actually the worst offenders on the panel.

That said, other times I have been shown some stats from what I would call the second tier of non-bank LPs that had truly horrendous hit ratios. As always with last look, there’s data for every viewpoint!

What also came through from the feedback, however, was the view than non-bank players have a much bigger role in FX markets than they used to have, but, surprise, surprise, I am not sure I agree with that! Again though, it probably depends upon which geography you view the issue from.

I turned, as I often do, to the semi-annual FX turnover surveys, especially those of the UK and US, that provide good granular detail, and there is no doubt that in the latter, non-bank players have taken on a more important role. The presence of a strong FX futures market in the region obviously helps, as does a lot of banks’ historical reluctance to really commit to the market once Europe has gone home.

Clearly, not all volume through a prime broker is from a non-bank market maker, but it is a decent proxy for the sector. In April 2016, when the US first started publishing the data, two-thirds of spot volume was traded via a prime broker. In the latest survey, from April 2024, it was up to a massive 92.5%. In other words, if you hit a price in the North American market, especially the afternoon, the chances are you are not hitting a bank.

It is not as clear cut as that, however, for actual notional volume from Other Financial Institutions, which includes the non-bank LPs, has actually dropped in the US over that period – from $197 billion per day to $176 billion. At the same time, Reporting Dealers, and they are exclusively banks, saw notional activity rise, from $76 billion to $136 billion. Other Banks, (and to a lesser extent, Non-Financial Institutions) saw a similar rise in activity, which suggests that maybe more banks are accessing PB services?

In the UK, a much larger market, PB-supported spot volume has also risen, to $455.9 billion or 44.9% of turnover, from $313 billion, 41.8%). This increase, however, has been matched by the Reporting Dealers, again all banks, whose collective share has risen from 37.9% ($286.1 billion), to 41.8% ($411.4 billion).

The two contrasting results are further confused by the aggregated data. Across both centres, PB-supported spot volume has actually lost market share, to 60.8% from 69.4% in April 2016, but so too has Reporting Dealers’ share, from 43.6% to 37.1%.

The demographic behind the data that so many LPs rely upon for their pricing has changed

On one hand, I would argue that, as our regular columnist Ted Holloway has convincingly, this signifies the big getting bigger, especially in the UK, but how to explain the aggregated decline, which is mainly at the hands of Other Banks?

For a start, the PB data could be the result of hedge funds getting back into the carry trade, thus requiring the services of their PBs. In April 2021, before the most recent hiking cycle began, 55.7% of spot across both the UK and US was prime broked (the US at 80%, the UK at 44%). In the US at least, the influence of the non-bank LPs may be less than it seems using this data.

Equally, what some people miss is the different nature of LPs – they are not all built the same. The non-bank players still largely prefer to quote in smaller sizes (compared to historical levels) and with a small number of exceptions, they still largely exist on the ECNs. Banks, on the other hand, especially from the Other Banks category, are still quoting their close relationship customers in larger sizes, via bilateral channels. Comparing the two types of LP, therefore, is worth very little – they have different motivations.

The interesting area is in firms like XTX Markets which, long ago, established itself in what is still a pretty unique area between the non-banks and the banks. The firm has heavily invested in technology, and is nimble like the best of the non-banks, but how it approaches its liquidity provision is very much like a bank. Other firms have tried to occupy that space, but have largely failed to achieve sustained success – even as XTX’ focus has shifted more to equities markets.

Overall then, I don’t see there being much a shift in the landscape when it comes to LPs – where they sat maybe six-to-eight years ago is where they sit now, with the caveat that some of the names may have changed. Where I think there has been a shift is one that has not been largely discussed – data.

Non-bank firms have, over the past 15 years, become a higher profile player on the ECNs and, notably, the primary CLOBs. As internalisation has grown as a concept, fewer banks push flow to these venues, and a host of what could again be termed tier-2 non-banks have stepped in (to a degree, volumes are obviously lower overall on these venues).

This means then, that the demographic behind the data that so many LPs rely upon for their pricing (and they still do in spite of concerns over the decline in volumes on EBS Market and LSEG Matching) has changed. Whereas maybe 10-15 years ago it was largely bank-generated, now it is more likely to non-bank generated, which is probably why more banks have started risk managing in the manner of a non-bank.

Where you hit the price will dictate your chances of hitting a particular type of LP (and they will use the data from that deal differently), but perhaps the thing that has changed is not that the non-banks are grabbing a higher market share, more it is that the segment now has a greater influence in the data that feeds the price you get from all LPs on different channels?

colin@thefullfx.com