The Last Look…
Posted by Colin Lambert. Last updated: August 30, 2021
The Global Foreign Exchange Committee’s Three-Year Review of the FX Global Code more or less formally came to an end with the release of the recent paper on last look, which means this is probably a good time to take stock of what has been achieve, especially as we look ahead to a change of leadership.
Some readers took my recent column on the last look paper to be a criticism of the GFXC in general, which it was not meant to be. I remain disappointed at the limited progress made on last look, but some progress has been made – and that is something. It has always been stressed that the Code is a living, breathing, document, I suspect that last look will feature quite heavily in the next three-year review, such is the slow progress on this subject.
Without wishing to get into yet another last look conversation, I would also remind readers, as I did to some of my correspondents – of my inherent distaste for last look, full stop. The current debate is about latency buffering and asymmetrical hold times, but I am remaining true to my initial argument first made more than 15 years ago, which means to me these are hideous toppings on an already horrible cake.
Back to the GFXC, however and a perspective on the work achieved (or not) in the latest review. Overall I think a good job has been done, the sections on settlement risk needed to be updated to reflect both the growing concern of non-PvP settlement in FX markets and the march of technology. Likewise, certain critical areas around multi-participant platforms and anonymous trading were strengthened.
A really good result of the GFXC’s work is the release of the disclosure templates, to me they reinforce transparency, give customers a level playing field on which to judge their service providers and do not intrude too much on a firm’s IP. One or two people in the algo world have expressed concerns about giving too much information away; as I pointed out to them, the difference between what they and their peers offer is minimal, the real differentiator – access to liquidity that will minimise market impact – is not part of the template.
Latency buffering and asymmetric hold times are hideous toppings on an already horrible cake.
If there is a downside, inevitably it comes in the two most difficult areas – pre-hedging in order management; and the aforementioned last look. I am not sure the guidance around pre-hedging was helped when the messaging from the committee shifted slightly with the removal of the phrase from anything relating to the Fix. This was due to an inconsistency in the Code, which was identified, however I am not convinced eliminating the prefix and re-labelling it “hedging” has helped. The fact remains, we have trading going on ahead of the window that materially affects the ultimate fill the customer gets and that to me remains dangerous territory.
On last look, a fair few people wanted a clearer message, especially around latency buffering and asymmetric hold times, but, aside from my earlier point about not liking last look at all, we do have to accept that the Code is a best practice document, and therefore cannot be too prescriptive.
If marking the work, I would, therefore, give the GFXC a “credit” for its efforts of the past three years. The level of engagement has been good, typified by the statement put out in March 2020 highlighting the risks associated with a month-end Fix in such volatile markets. Previously there has been a sense that the GFXC worked on a six-monthly cycle around its meetings, and that is natural, but demonstrating the ability to step in and highlight potential risks to a segment of the market that maybe wasn’t previously listening, has eroded that.
This is, indeed, a challenge for the next chair and co-vice chair of the committee as Guy Debelle and Neill Penney step down at the end of their terms. It is yet to be revealed who will take over, but it is to be hoped that they are senior officials, with credibility across the industry and a voice of authority that will be heard. Richard de Roos, as the existing co-vice chair, has had a year to get his feet under the table and be able to help the transition.
The first year or so will not be easy, however, because so much has been achieved over the past 18 months, culminating in the three-year review. There is, however, a need to maintain momentum. This could come in the form of further analysis papers, such as those published last year on market conditions, or perhaps be focused on the semi-annual turnover reports from the regional FX committees.
It is time for more of the buy side to assume some responsibility and police their service providers according to the guidelines established by the Code.
There is one other area that I believe the GFXC needs to continue its messaging – assuming I have read the tea-leaves correctly. My sense with the two trickiest areas of the Code, pre-hedging and last look, is that the committee has done as much as it can for now. Yes, there will be opportunities for further adjustment as we move to the next review period, but as things stand, unless there is a seismic shift in stance from the deeply entrenched positions on the committee (which I very much doubt), little more can be done.
This leaves the focus very much on the consumer side of the ledger. I fully accept that even one year ago it was an incredibly difficult task to directly compare disclosures. Equally, last year, pre-hedging ahead of the Fix was seen as standard, now it is just hedging ahead of the Fix. This means the liquidity consumer, the customer, is in no doubt about what is going on and this is where I think the messaging from the GFXC has been heading for some time now. It is time for more of the buy side to assume some responsibility and police their service providers according to the guidelines established by the Code.
I acknowledge that the Code does not provide legal certainty, but proper buy side analysis and policing of the service they receive would do a lot more to bring providers into line. This means, however, making decisions and taking positive, proactive, action. If you are happy with the activity ahead of the fixing window, explicitly say so. If you want your orders executed in the window only, then explicitly acknowledge the warning of your executing party that it may cause a liquidity event. Both are fine, they demonstrate that you have assessed the risks, received the warnings and want to continue.
On last look, assess your LPs round trip times on both accepts and rejects and question large differences. Equally, question those LPs who take significantly longer to execute or reject trades, if LP A can do it in 30ms, why does LP F take 80 or 100? More pertinently, as a piece published earlier this year highlighted, if your LPs are playing games with asymmetricity or higher rejects, take the business away from them. It’s no good complaining about your treatment if you continue to trade with those you feel are not being 100% fair.
This leads me to an early request for the next review of the Code. If we are to reflect the modern FX industry, then I believe the FX Global Code should more clearly recommend as best practice that the buy side conducts proper execution quality analysis. This should look at price action ahead of a fixing, or other, order; it should also explicitly calculate the cost of rejects. The technology exists where perhaps it didn’t three or four years ago, so use it!
As the leadership of the committee approaches renewal, I feel that conduct in the FX industry has progressed thanks to the efforts of the GFXC over the past three years – we are never likely to have a perfect outcome that suits everyone in some areas. Of course, as a living document we should have moved on, but I feel, as my “credit” mark earlier suggests, that the GFXC has perhaps, exceeded many peoples’ expectations in 2017-18.
The next stage will be trickier, a lot of the low-hanging fruit has been picked, but if my assessment of the messaging is correct, then further progress can be made. Whether or not it is, depends upon whether the buy side is willing to step up and play its role. It won’t be easy, the buy side is a hydra, but if leadership can emerge it will do the sector, indeed the whole industry, a great service.