2017 in review: FX market structure

I’m always surprised when I hear folks say that 2017 was an uneventful year in FX.

Volatility and revenues

It certainly feels like we are living in ‘interesting times’ from a macroeconomic perspective with Brexit, Trump and China’s continued ascendency. Yet it is absolutely true that volatility in FX (as in all asset classes) was rock-bottom again this year. The VIX is currently hovering near its lowest recorded levels.

As everyone knows, FX revenues are highly correlated with volatility. Accordingly it has been a down year for most FX divisions. Those in growth mode or with a large corporate franchise or exposure to bespoke FX risk management products have fared relatively better but, across the market, revenues have been challenging. On the other side of the accounting ledger many mature institutions find they have an already pretty optimised cost base that cannot be meaningfully reduced from here without pulling out of core activities.

If history is any guide, volatility cannot stay down here indefinitely but those who designed their business on the assumption it might are looking smart. Exacerbating this has been the work required for MIFID II. This has been a huge focus at most institutions and has sapped resource from new business initiatives. This is likely to continue into H2 of 2018. After that life should be a bit more fun as focus returns to new ideas and products. 

Regional banks

One bright spot has been the improvements that banks outside the traditional top 10 have made. Often boasting strong corporate franchises and excellent credit ratings, these institutions are in many cases using third-party vendors to outsource non proprietary aspects (connectivity, credit checks etc.) and are thus able to launch high quality eFX client offerings at a fraction of the upfront or maintenance cost that a DIY approach would’ve required, even just a few years ago. Anecdotally this seems to be bearing fruit.

The Global FX Code

One defining feature of the year, unfortunately, was a number of high profile fines for historic FX conduct activity. It will be a relief when the last of these is out of the way and the industry can focus on the future instead of the past. Promisingly the zeitgeist of 2017 was animated by one such forward-looking initiative: the Global FX Code. 

To get hundreds of private and public institutions across the world to work together and agree on best practices is an achievement in itself. The most contentious topic was pre-hedging in the last look window (‘Principle 17’) with a robust debate on whether this was or wasn’t acceptable best practice. Following a comment letter period, it was agreed that this was not in line with best market practice and the Code’s language will be updated this month to reflect this. Although it may seem subtle, this is a huge victory for FX market consumers and will put an end to one of the biggest rent-seeking activities in today’s market structure. The big challenge now is adoption. It seems certain that all major banks will sign up but a question remains on how to deal with major HFT who market make on anonymous venues and may wish to remain unencumbered by industry best practices.

In their comment letters a number of institutions argued that the Code should go further and ban last look entirely. It seems unlikely that this will happen since, unless there is a third-party matching agent in-between the LP and its client, the LP will always need the right to reject a trade. Why? Because it is the LP who must determine the contract is within credit and position limit thresholds. Nonetheless it does seem that last look’s overhang on the market is set to reduce.

TCA

One reason for this is the increased use of TCA in the market. There are now a number of high quality independent providers and the uptake amongst the buyside in 2017 has been unprecedented. This can only be a good thing. Certain ECN’s are also bringing a new level of transparency to their platforms with several introducing analytics that show their clients their cost of rejects and market impact for the first time. What gets measured gets managed and it seems inevitable that once consumers see the cost of rejects they’ll vote with their feet in cases where the cost is disproportionate. 

For various reasons, systematic approaches to portfolio management - whose practitioners have always been at the forefront of the market in terms of execution analytics and quality - appear to be in favour with investors at the expense of discretionary global macro traders, who were typically a huge part of G10 FX client volumes over the last decade. Due to the vastly different execution styles and business requirements of these two types of participant this is quite a pivot and the sell-side is cautiously adjusting its personnel to accommodate this shift.

Interbank venues and market data

The second headwind for last look is the faster data from primary markets. This changes everything. Over the last year primary market updates have sped up meaningfully: in some cases from 100ms to 5ms. Now LPs looking at the primary market to perform an ‘on market check’ in EURUSD or USDJPY can do so in 5ms or less. Just as a drop in oil prices must eventually result in cheaper petrol at the pump, so last look times are being slashed to reflect the step change in the underlying interbank market that powers price discovery. Incidentally one interesting development in the interbank market is the conscious tying of market quality to market data: in some cases, one must contribute (i.e. provide liquidity) in order to receive the fastest market data. 

The faster interbank market data has had another effect: the number of quotes generated on secondary venues or single dealer API has exploded. Some executives have gone on record as saying quotes are increasing at over 30% compound each month! This puts extraordinary pressure on systems at a time of low revenues, since an increase in quotes does not result in an increase in matched trades. Expect to see more focus on this in 2018 as end-users (whose systems may also struggle) and ECN’s begin to evaluate the cost of processing these quotes. It seems likely that the business approach will be to measure quotes generated:traded volume ratios for each LP and offboard those who generate a lot of quotes without adding much value by being top of book and winning trades. This is likely to spell hard times for liquidity recyclers who generate a huge number of quotes whilst rarely trading.

Clearing and NDF trading

Credit has always been a defining feature of FX market structure. Clearing in NDF continues apace and next year might reach the tipping point of >50% of dealer-to-dealer volumes being cleared. Increased adoption is important as it drives unit economics with the relative cost of clearing decreasing versus the alternatives with each additional % gained. Dealer-to-client activity is different entirely and is likely to remain bilateral for the foreseeable future. The infrastructure required to trade NDF electronically has also blossomed over the last 12 months with the market rapidly transitioning from a manually risk-managed and priced product set to an electronic one. 

There are a variety of clearing initiatives beyond the existing NDF market. Whilst compression services and repapering have been effective in alleviating some of the pain faced by banks’ forwards desks, there is a gradual path of regulation-mandated deadlines that mean that each year we are likely to see an increase in clearing of FX products. As before - once economic tipping points in terms of unit costs are reached, expect to see a rapid acceleration in uptake.

Bitcoin

I can't even.

Transactional FX

With business conditions poor in institutional FX markets it is unsurprising major players have invested in and focused on consumer payments i.e. physical FX. Valuations (often based on relatively small notional investment rounds) look seriously optimistic for tech entrants, who are having the effect of repricing consumer FX from circa 300bp to 50bp (or less) from mid. It is likely these tech entrants will never reach meaningful profitability but their existence will nonetheless be a drain on retail banking franchises over the next decade since they’re likely to have to reprice defensively. This transactional FX business is of course totally benign in profile and largely uncorrelated with institutional FX revenues. With cross-border retail increasing at circa 25% compound annually and 85% of transactions still involving physical cash the received wisdom is that there’s plenty of room to grow the payments pie.

Anyway...

If you got this far, congrats, and I wish you a relaxing and enjoyable break. Look forward to catching up with you in 2018!

The views expressed on this blog are the author’s personal views and should not be attributed to any other person, including that of their employer.