The Last Look…
Posted by Colin Lambert. Last updated: November 20, 2021
When GFXC chair Guy Debelle stated clearly what many of us already thought – that the FX Global Code update frowned upon latency buffering or additional hold times in last look – there were still plenty of FX market participants grasping at the straw “well, it was only at a conference and isn’t official”. That straw has been removed and should, surely, be the one that breaks the camel’s back.
Before I start, however, thank you to those eagle-eyed readers who picked me up on my errant maths last week – I did think about the excuses of auto-correct etc for putting an extra nought in the numbers, but then had to ‘fess up. At least you all now know why I am no longer trading!
The FCA’s explicit statement in support of Debelle’s comments makes it clear that the regulator of the largest FX market centre in the world believes AHT is not best practice – that should be a wake-up call for a lot of people. My understanding also, is that as regional FX committee meetings are held around the globe, representatives of the local central banks are reinforcing Debelle’s comments and making them part of the record of the meetings – again that is hard to ignore.
As I wrote just after the Code revision was published, my view was the update made AHT poor practice, but the language used was a little wishy-washy, hence why some chose to believe nothing had really changed. The weight of public statements now being made would suggest those rose-tinted spectacles need to come off now.
The FCA also explicitly linked adherence to the Code with the Senior Managers’ Regime in the UK, a regulation that can hold the relevant person responsible for conduct in this area. SMR has caused a few problems, and I am not convinced that it really targets the right people in a business, but if someone senior enough thinks they’re going to be in trouble, they are going to change things, which is what was intended all along.
So, with the FCA adding its considerable regulatory weight to the matter of latency buffering, where does this leave the industry?
I actually think that while some LPs may suffer, there will also be a detrimental effect on some clients in the industry – personally I don’t think it’s a bad thing either. Some clients continue to abuse liquidity and are content with 70%-plus reject rates, I doubt they will remain that high, because inevitably the spreads to these accounts have to get wider. If that happens, what should have occurred in the first place may actually eventuate – the market will self-regulate these clients’ behaviours by either widening spreads considerably, or cutting them off altogether.
Either widen tricky clients or don’t quote them – you’re probably not making money out of some of them anyway, so the FCA, via the GFXC, may have done you a favour.
I have spoken to some large hedge funds in recent years who drifted away from FX because they found their spreads were getting too costly because some LPs simply weren’t quoting them. This wasn’t the result of the LPs enforcing some sort of standards upon them, it was merely that an increasing number of LPs didn’t see the point and competition diminished. These funds’ solution was to revisit how they traded, and return to the market with a totally different attitude, that recognised their behaviour contributed to the problem.
Not all clients did this of course, but I suspect if – and it remains a big “if” – the LPs do adhere to no AHT, then they too will have to change how they interact.
For some LPs, this represents a huge challenge, after all, they might lose ground in the Euromoney survey if their volume numbers go down. What are they to do, the atmosphere of crisis must be palpable? Just to confirm that I am indeed being sarcastic, I think the LPs should find a transition fairly easy. Either widen tricky clients or don’t quote them – you’re probably not making money out of some of them anyway, so the FCA, via the GFXC, may have done you a favour.
More seriously, I think this will help strong LPs come to the fore. There are some big players in the market who believe in the value of latency buffering, this may mean a change in process, but I suspect that six months down the road (assuming they change of course) they may find that they are winning more business that currently is going to smaller LPs who are, to put it politely, less into the risk-warehousing game.
This could also bring a few regional players back into the fold as well. There are several regional players who operate with zero AHT, but they are often losing out to players who are buffering. If the latter are forced to change, does this make the regional player more competitive? If it does, I think it’s good news for the market ecosystem overall, and, importantly, for clients.
I don’t think we will ever see the day that executing on last look liquidity is not seen as best execution, but on a stream with AHT?
A final thought on the impact of this statement. Is the next big question needing clarification from the GFXC and regulators like the FCA whether executing on AHT streams is in fact best execution? It’s not as clear cut as it seems, because the fact is we still have last look – it’s important to remember it hasn’t gone away, merely how it is used should be changing.
I don’t think we will ever see the day that executing on last look liquidity is not seen as best execution, but on a stream with AHT? There is definitely a case to be made that AHT liquidity is likely to result in higher reject rates, this being the case should the executing party actually include these streams? The FCA statement questions whether executing with a party that has not signed the Code is best execution, but that is a broad observation – it may well be we have to wait and see what happens on AHT streams and best execution.
All is rosy in the garden of those who believe latency buffering is wrong then? Well, yes, but – and you all knew a “but” was coming – this could also represent a major challenge for the Code, as I observed a few months back.
First of all, clients have to support the push to abolish additional hold times. I have already observed that some are content with 70% reject rates – are they going to check up on their LPs? They may well see nice tight spreads disappearing (I don’t believe it will by the way, competition will ensure they stay – but they won’t be at current razor-thin levels), so are they going to metaphorically shoot themselves in the foot by holding their LPs to account?
There are also those clients (still in the majority I would argue) who simply don’t know or don’t care. A recent survey found only 40% of managers were really bothering to track their true cost of trading, are they going to get even more into the weeds with last look hold times?
I actually believe that the FCA statement will be a wake-up call for UK-regulated entities, who will actually start to look at this stuff, but the rest of the world? Perhaps Europe, but that will take years for ESMA to get its act together. Certainly Australia, where ASIC already supports the FX Global Code, and areas of Asia. The big question is what will the attitude of the US be? It is an open secret that many in the industry believe the Fed feels this is a minor issue at best and is unlikely to act strongly, and no other regulator has the necessary power to endorse the Code and its principles.
Being an optimist, I do think this will close the book on latency buffering, which brings me to the second “but”, something I raised a few months ago – what happens if the people just ignore it? Luckily the UK market is so big that few players will be impacted, but we will need to look closely at the jurisdictional powers of the FCA. Can it hold a trading firm in, for example, Texas or California, to account and bring direct action? In the US the answer would be yes, but in the UK? I’m not so sure, especially if the matching engine being used is not in that jurisdiction.
There are enough big players backing the end of latency buffering to form a consensus, but it probably does need one or two large players who have thus far not changed their hold times to reflect the credit and market risk check, to switch sides before it becomes irresistible.
One more note on the FCA statement, which was a Lambert Dreamland given it also talked on another old favourite – pre-hedging. On that subject, the FCA states, “Pre-hedging practices where market participants do not communicate their practices to clients in a manner that allows the client to understand the potential impact on the execution of their order are not consistent with the Codes.”
I know that a lot of executing parties do raise this issue with the clients, especially around the Fix, and that many clients continue to ignore advice. That should change after the FCA’s statement, especially for those funds using any window that cannot cope with the volume. If this isn’t a call for a longer Fix window I don’t know what is, and as such it is probably beholden upon managers in the UK especially, to either seek out an alternative solution, or press for change from their current supplier.
And with all that excitement out of the way, I’m going back to my reading matter, Basic Arithmetic!