Over the years the most powerful criticism aimed at e-commerce and its potential impact on markets has not been about volatility, or market behaviour generally, it is its lack of flexibility – why else, for example, has the FX swaps market not become more automated in recent years?
Partly the answer lies in what is a pretty powerful group made up of the voice brokers and their banking counterparts – whenever I raise the issue (typically from the other side of a locked door – they can take unkindly to the question) this group tells me that there is so much negotiation involved, not to mention credit obstacles to overcome, that a computer simply couldn’t handle the complexity. They also like to point out that the most frequent outcome in the FX swap market when using a voice broker is a trade at mid-market, more often than not for the full amount.
These are all valid arguments, but can they really not be overcome by technology? It was striking to me during my recent visit to London, how many bankers in the e-FX business segment were not only targeting the FX swaps market, but also talking up their chances of raising e-ratios significantly.
The low hanging fruit here is the client portfolio – with best execution requirements extended to swaps and rolls from spot, the easy route for clients is to execute these trades on a platform, and the banks all have something to offer them. The client can get a bespoke date and amount, full STP and, crucially if required, a nice TCA report that says they hit the best price of those offered by their panel of banks.
It is there, however, that the e-commerce trail largely goes cold. Banks typically manage buckets of risk in these products and hedge flows with each other, but aside from a few bright spots at the short end, very little business between the major dealers is executed electronically – it remains the domain of the voice brokers.
I accept that the last BIS survey is very out of date as we are pressing up against the sampling month for the next triennial survey, but looking at 2016 the data show that just over $2.3 trillion was executed in FX swaps every day, with more than $700 billion per day being in the one week-to-one year bucket. Recent data from the UK indicates that about two-thirds of FX swaps volume in that bucket is executed in tenors between one month and one year, so we could easily be looking at close to half a trillion dollars of daily volume in the one month-to-one year tenors.
How much of that is automated? It’s hard to tell, but the answer is, I would suggest according to anecdotal evidence, not a lot.
Even across the entire FX swaps spectrum almost 80% of volume, so some $1.8 trillion per day, is between banks, or between banks and hedge funds – the point being that a large amount of these trades do not require settlement. Not only does this make this market ripe for clearing, as I have argued for some time now, but it would also seem that they are ready for automation – after all, who has been leading the e-commerce charge over the past 20 years? Banks and hedge funds.
With more clients choosing to execute their FX swap requirements electronically, it would seem inevitable that the counterparties involved – the banks – would want to extend that automation further up the chain. The problem is no solution seems to exist that has everybody onside – so this has to be the next frontier for the platforms.
It seems to me that the focus on the client has gone as far as it can, or should. It could be argued – as I have often in this column – that this focus on the client has led banks into dangerous territory morally, ethically and behaviourally, but it could also be argued that it has blinded the platform industry to a real opportunity.
The mantra from any platform you talk to is often nonsense – this “unique flow” doesn’t really exist, it’s just a question of how smart their liquidity consumers are and how far down the latency spectrum (in terms of when they receive the information) they reside. This is a spot only issue, because, as noted, the customers are very happy to RFQ or maybe RFS their FX swaps activity in a disclosed environment with counterparties they know who will tailor their price to tenor and amount.
So the platforms’ target should not be the customers (in the traditional sense) in this instance, they should be looking to the major dealers – and making their life easier. The mantra in forwards should be about the efficient mechanism that facilitates trades between the big risk houses – as well as the connectivity to clearing houses or novation mechanisms – and how they can have a hugely positive impact on the balance sheet. There, in balance sheet relief, is the big selling point, it’s not about how many customers you have.
There is a genuine mystification in parts of the FX industry that both the credit and technological barriers to greater e-FX swaps trading have yet to be broken down. Obviously there are a lot of non-bank firms eying this space with no little interest as they seek to bring their technological expertise to bear, and this itself is occasionally cited as a reason the banks are reluctant to support greater automation, but should it be? I would argue that most dealers (as well as customers) want to execute in big clips – they don’t want to break trades down the way they do in spot.
Given the slower pace of the market and the fact that immediate offset trades are unlikely to be available, would non-bank firms have the appropriate framework to warehouse large risk in longer tenors for any reasonable length of time? Technologically speaking of course they would, but regulatory-wise? I am less sure.
I don’t see the compelling market structure change argument in FX swaps from these firms that I did in spot markets – especially if the end user customers are content with how things work currently. We would probably find that the current market structure would stay in place longer than they could sustain what could be an expensive business model.
Of course, some non-bank firms will see the value in entering an FX swaps market with the appropriate structure, but we will not see the liquidity recycling we do in spot – and that means less non-bank players. What we could see is the next iteration of the speed game, but I am not sure if that would be welcomed by anyone. The platforms would have to up their game, the banks would lose value, the non-bank firms themselves would find themselves in a dog fight and spreads and margins would be compressed, as they have been in spot, to levels where size wins and, more importantly even the cost of even cleared business would be too high for many.
I find this a genuinely interesting conundrum – what model would gain traction in FX swaps? I tend to lean towards a dark trading environment with discretion. We all know the basic price curves so price discovery shouldn’t be an issue, but if a dynamic credit model is available that allows the banks to adjust their pricing should they face off against certain counterparties that would be a start. Throw in the ability to say, “if I am within a certain range of mid-market and can get a match against the right counterparty in the right amount, execute at mid” and you have a very compelling solution. After all, that is pretty much what the voice brokers do.
I don’t think the technology challenges to create such an environment are overwhelming, the connectivity would be complex and latency would have to be accepted in the process, but how many trades in FX swaps, especially over one week, go through per minute? I am sure the time can be found to ensure all the ducks are aligned.
It feels strange to me that we keep hearing about initiatives to automate the options and (especially) NDF markets but in the big scheme of things these are small fry. The 900 pound gorilla in this particular room is the e-FX swaps market and for some reason, no-one seems willing to approach it. That is the current reality, but I suspect if a platform approached the right people in the right banks now, they might find less resistance to the idea than they would have just two years ago – and that would be a significant step forward.