There are those in the FX world who believe that the narrative in certain circles of a return to bilateral, relationship-based trading was a group of liquidity providers talking their book and hoping to push the market away from what could sometimes be costly anonymous trading venues.
Looking at the numbers in last week’s FX committee turnover surveys, specifically the spot e-trading statistics from the UK and US, I think it is fair to say that the cynicism is wrong, or the narrative is working, or both, because the last two years has seen a definitive shift in trading away from anonymous venues towards disclosed channels.
Using my renowned analytical skills* I took a look at the various spot e-ratios for disclosed and anonymous in the April 2016 surveys and the latest survey results taken in April 2018. I used 2016 because in the UK there was only $4 billion difference in spot FX turnover between the two surveys (in the US it was slightly wider but I think still close enough for comparative analysis).
In April 2016 in the UK, electronic broking systems, typically the anonymous matching systems represented 36.1% of electronic spot volume and 20.9% of all volume, while multi-dealer systems were responsible for 40.7% of spot ‘e’ volume and 23.6% of overall.
In the latest survey, the broking systems had a 29.5% share of ‘e’ and 18.2% of overall spot volume, whereas the multi dealers had 45.9% and 28.4% respectively.
Reinforcing that is a less significant, but still noticeable, uptick on spot volumes on the single dealer platforms, which saw their share of spot volume rise over the two similar surveys to 24.7% of ‘e’ (from 23.1%) and 15.2% of all volume (11.2%)
In hard numbers, the disclosed channels across two similar months in spot FX handled more than $44 billion per day more, largely at the expense of the anonymous ECNs.
In the US meanwhile in April 2016 the indirect channels, typically ECNs and other anonymous venues handled 26.6% of ‘e’ and 18.3% of total FX volume, while the direct, or disclosed venues, handled 73.4% of spot ‘e’ business and 50.6% of total spot business.
By April 2018, however the shift apparent in the UK was also occurring in the US where anonymous venues’ share of e-spot trading had fallen to 24.7% (17.3% of all FX spot turnover) while that of the disclosed channels had risen to 75.3% and 53% respectively.
Again in hard numbers single dealer platforms saw a $10 billion per day increase, while ‘other’ direct channels, normally seen as aggregation, saw an increase of $25.5 billion per day. Thomson Reuters Matching and EBS saw absolute volume drop by $10 billion, while “Other ECNs” saw activity rise by $7 billion and ‘Other Indirect Channels’, broadly seen as those venues such as retail and small institutional platforms, saw activity drop by $6 billion. It should be noted that actual spot volume in the 2016 survey was some $33 billion less, however, and that will skew the numbers slightly.
Looking across both centres there has been a significant shift towards disclosed over the two year period, 67.6% of all ‘e’ volume in the UK and US in April 2016 was via disclosed channels, in 2018 this had risen to 72.5%. In terms of overall volume the percentage has risen from 41.7% to 47.1%. Anonymous volume also fell as a percentage of total spot FX volume, from 20% in 2016 to 17.9% in 2018.
At first glance this creates an apparent paradox – at the same time as this was happening the numbers reported by the major platforms rose by just over 11%. I say apparent paradox because of course a lot of this is explained by the growth of FXSpotStream, which is a disclosed venue, as well as the influence of EBS Direct and disclosed streams at Hotspot (CboeFX).
I may be over-reading things and probably we want more data to support this theory, but to me this represents solid evidence of the shift in attitudes to liquidity provision in the modern FX market. We know the major LPs have become much more choosy about where and to whom they stream – volume does not beget success in today’s FX world – we have also seen an acceptance on the part of reasonable customers that something needed to change as the quality of their liquidity pool was being affected by some LPs behaviour.
As noted in a recent article in Squawkbox, new and better methodologies to measure LP behaviour are available, therefore we may see even more improved conditions in aggregated liquidity pools which would, presumably, lead to more activity through those channels at the expense of the anonymous venues. The latter really need higher volatility and greater uncertainty so the dealers come to hedge with each other if they are to continue to grow.
I look at this latest data set and see market participants voting for a more responsible market with their business and that can only be a good thing for responsible participants and their customers. Not only do those existing on the edge of acceptability now have (potentially) the Global Code to deal with, but their models will come under pressure as liquidity deserts those channels that support their behaviour.
Hopefully what we are seeing in this FX committee data is further reinforcement of the push for more accountability and transparency in how people act in the FX market. The FX market has had a volatile decade or so as the balance of power swung between the big banks, smaller non-bank firms and the customers, with each having a period of ascendency. With each swing has come behavioural challenges in the well-publicised cases of bank traders attempting to manipulate the market and share customer information; the onset of latency arbitrage when fast firms picked on their slower brethren; and predatory execution tactics by some customers as they realised LPs were willing to stream to them no matter what the consequences, as long as they won the trade.
Underlying the FX industry there has, I believe, always been a reasonableness – an understanding that everyone needs to get something out of the market, be it profit, efficient execution or something else. The Global Code is very much about promoting this reasonableness and balance, and this latest data suggests to me that the market is also embracing that concept by acting in a transparent manner, both as LP and liquidity consumer. If I am right then this is a really positive signal that the worst of the FX market’s challenges are behind it (for now of course, something else always comes up!) and the industry can look to move ahead with greater confidence in each other. That has to be a good thing.
*Which don’t exist of course