The Last Look…

Last week’s column elicited a lot of feedback and conversation, so much so that I can move the issue on a little this week with some more information and, inevitably, my own theory.

For those of you with short attention spans or the inability to search back one week, I wrote about the sense amongst major LPs that one of their brethren had changed how they hedge out of risk and this was having a knock-on effect throughout the market.

Notwithstanding the different views I received upon publication, and I have to provide a tip of the hat to a friend who seized upon my “economist strategy” in the column when I closed out by suggesting that it may all be nothing more than a push for Euromoney votes – what would feedback be without criticism? – there seems little doubt in my mind that something has changed. This means it’s really just a question of what?

Interestingly, in contrast to what I suggested last week, liquidity consumers do seem to be seeing increased market impact. Several firms who measure this got in contact over the course of the week to confide that in April they typically saw more market impact than in previous months. This could, of course, be the result of other major LPs responding to any change they perceive to have occurred, if mark outs are being squeezed then naturally players are going to hedge quicker, with the result being more market impact.

If that is the case then this has consequences for those firms dealing with each other on a “full amount” basis. Putting aside the paradox that full amount trading in the modern sense is nothing of the sort – it is merely a matter of letting liquidity rebuild between a series of larger clips – if LPs are going to take longer to rebuild liquidity back to, for example, 20 million, then consumers are naturally going to be less keen to execute a trade for 200 million in 10 “full amount” clips. This is not the fault of the LP – they are merely looking at the analytics associated with providing liquidity and hedging out the risk. If market conditions change they alter their hedging style and therefore we should expect them to do the same if a style change at a rival firm tilts the playing field.

What is happening could have consequences for those firms dealing with each other on a “full amount” basis

There were quite a few comments about how this change is affecting those “LPs” who are in recycling game. Apparently a few are struggling to make money in the new environment. Pausing briefly (sarcasm alert) to wipe the tears from my eyes as I ponder these firms’ struggles, I feel the urge to point out that any firm claiming to be an LP needs to be able to react to a changing landscape. If one firm has found a different way of hedging then these recyclers should react the same way as everyone else – with a change of their own. I have little time for those complaining that the world has changed for them, for this highlights a sense of entitlement that permeates too much of the market currently – LPs should not have the right to make money on every trade and customers should not be interested, certainly not complain, if the LP makes money out of their flow. Perfect doesn’t exist, deal with that fact!

Anyway, the sum total of all my conversations of the past week and more, has allowed me to come up with a theory as to what has happened. I should stress, this is a theory only, and as my track record suggests, it could be right…and it could be wrong! It goes back to the futures aspect.

I noted last week how hedging into futures markets could have a slightly higher impact due to transparent nature of those markets and also the fact there were a lot more speculative, tech-savvy, accounts on those venues who will jump on the flow and exacerbate impact. The question is, has a major LP started using futures more? I would answer, they probably feel they have had to. This is not about a firm suddenly “discovering” futures markets, most of the major firms have been connected for some time, more it is about those markets playing an increased role in their hedging.

Consider this. Taking the BIS Triennial Survey for 2019 as a base, spot volumes are 33.4% higher than in 2010; 2.9% lower than in 2013 and 20.3% higher than in 2016.

Over the same time frame (and using April as the benchmark month), the combined spot FX volumes of EBS and Reuters/Refinitiv have almost halved since 2010; are down 43.2% since April 2013; and are down 17.3% on 2016. With volume of $148.3 billion in April 2019, $141.4 billion has disappeared from those venues in less than a decade.

Volumes on EBS Direct rose 51% year-on-year in April. Overall EBS volumes in that period? Plus $700 million per day or 1.1% higher.

The actual situation is much worse, however, for the above data includes all spot FX volumes handled by those firms, crucially including the non-primary market venues. If you go back seven-to-10 years, I would suggest that Matching would be seeing volumes in the $60-80 billion range and EBS Market in the $100-120 billion range. Now those numbers hover around, probably under, the $30 billion mark on each venue, meaning ADV one-third what it was a decade ago.

There is actually evidence of this in the latest data from CME Group, owner of EBS. In its release for the volume data for April, CME says that volumes on EBS Direct, it’s relationship platform, rose 51% year-on-year. Overall EBS volumes in that period? Plus $700 million per day or 1.1% higher.

Comparable data from LSEG’s Refinitiv is unavailable, but anecdotally speaking, as someone who watches this type of thing, the predilection of Refinitv people to “talk up” FXall stands in stark contrast to the silence around Matching, suggests the same has happened there.

Faced with this, what is a major LP to do? They are clearly seeing more volume and while internalisation diminishes market impact greatly, there are still significant exhaust flows that need to go into the market. Some help has been provided through LMAX Exchange and CboeFX, both of which have no last look, firm liquidity, venues that are doing decent volume, but if those venues combined are seeing ADV in the $25 billion per day range, it still doesn’t make up for the shortfall in the primary markets.

Over the same period, incidentally, CME’s FX futures and options volume has also declined, but by less, some 25-30% and, assuming 10-15% of FX volume is in options, CME is still supporting $50-60 billion per day (non-roll months).

At face value then, this seems a case of an LP getting first mover advantage and using futures more than they previously did, hence the knock-on effect across the market. Of course, I should point out that generally speaking CME’s FX numbers are not registering a significant increase – they are steady at best – which brings me to the point.

I don’t actually think this is a case of more overall flow going to the futures market with the host of firms ready to jump on it, rather I think this is a case of a greater proportion of the exhaust flow going there. Anecdotally, I am told that internalisation rates are steady at the major LPs, so it’s simply a case of more of the hedging in the public market being done on a venue with more than its fair share of opportunistic traders.

There is an irony in how the continued diminishment of the primary venues (to the stage where we really have to question how long they will retain that status), has led the banks to rely more upon a venue they once tried to avoid 

The challenge for the entire FX market is going to be reacting to this change, and I can only really see shorter hold times for risk resulting. There will, no doubt, be a healthy debate as to whether this is actually good for the market, personally I am unconvinced, but it is just the new reality.

Overall I think this episode highlights the huge impact internalisation has had on the market. Around the end of the first decade of this century, the major banks had a real issue with the trading environment on EBS Market in particular, complaining that too many HFTs were spoiling it. The irony is, EBS’ response to the threat represented by, most publicly, ParFX, coincided with a surge in internalisation rates as banks and then non-banks upgraded their technology stacks, with the upshot its (and Matching’s) volumes have dropped.

Compounding this irony is how the continued diminishment of the primary venues (to the stage where we really have to question how long they will retain that status), has led the banks to rely more upon a venue they once tried to avoid due to the presence of so many HFTs and prop trading firms.

What’s happening is not new, therefore, it’s just different, and probably signifies yet another turn in the evolutionary cycle of FX markets – and it’s something LPs and, importantly, their customers, have to come to terms with.

colin@thefullfx.com

@colinlambertFX