The Last Look…
Posted by Colin Lambert. Last updated: April 8, 2025
Last week I noted in our weekly newsletter’s Any Other Business… column (exclusive to those who receive the newsletter – sign up now!) that a new generation was coming to terms with the phrase “liquidity mirage”, but how much does it exist?
The story that was referred to me by a friend was quite generic, but may, or may not, have been inspired by a recent paper published by Deutsche Bank looking at the FX market, within which the question was asked, “the FX liquidity mirage: myth or reality?” The paper does not answer the question, rather it lays out some aspects of the debate, meaning yours truly is free to have a crack at answering the question.
The simple answer, as I hinted at last week, is “yes, and it’s never gone away”. In a fragmented market, liquidity will always be multiples of its reality because, frankly, it all stems from one or two prices. Naturally, the issue is more complex than that, but the starting point has to be the reality check that for all the volume that has left the primary markets, the only true liquidity in FX markets is on firm venues – and even then its availability can come down to speed of cancellation versus speed to execution.
At a simple level, this suggests that, taking an estimate of Cboe FX’ firm stream, EBS Market, LMAX Exchange, and LSEG Matching volumes (and adding a bit for other smaller venues that exist), about $100 billion of spot turnover daily is firm in “normal” times. That’s about 5% of volume, which suggests the liquidity mirage could be up around 20-times.
Obviously, there are streams for certain clients that have hit rates of 100%, that will boost the number (although technically the liquidity could still be subject to the mirage), there is also a fair amount of business done via the phone and chat, which, given its nature, is not subject to last look – it’s a question of if someone says yours/mine before the LP says “off”! So the actual number is hard to estimate – in the 2022 BIS Turnover Survey, voice business was responsible for about a third of all FX volume, so $2.5 trillion or thereabouts, but that does not differentiate between spot and the voice-heavy swaps and options markets.
The analysts at Deutsche note that a trade on one venue can trigger a simultaneous reduction in liquidity on other venues, and they are right of course, I have seen too many instances of a price, or deal, on a platform, being reflected elsewhere, either through “pipping” or cancellation, depending on what happened. This is the inter-connected nature of the FX market.
The thing is though, none of this is really new. The mirage has existed ever since e-trading saw the introduction of last look, and it has grown commensurately as volumes traded on prices subject to last look have grown – and will continue to grow. What is new is my sense that there is a frustration in the quant ranks over their collective struggle to identify the mirage, or even measure liquidity accurately.
Different metrics that can help measure have been thrown around in my conversations on the matter over the years, including percentage of aggregated liquidity, slippage, probability of fill, probable price impact and many more – and they’re all pretty useless, because the quants are finding out what any voice trader could have told them years ago; any liquidity analysis is, by its nature, historic. What happened over the last minute, day, week, month, decade, millennium, is largely irrelevant, which is anathema to people used to dealing in hard data and certainty.
Facts are facts: the liquidity mirage exists and has never, and will never, go away
And this is the problem, we are trying to bring certainty to an issue that will always be vague and ambiguous (unless we have just one venue and get rid of last look, but don’t get me started…). The fact is, a new piece of information changes the dynamics of the market, and can emerge on a second-by-second basis. All can be rosy in the FX garden and then, for example, Typhoon Trump unleashes, and all is chaos, leading LPs to adjust their settings.
The current generation needs to understand that they exist in a tech-driven market where not only price but depth changes by the millisecond, and that includes regime change. The only real difference between now and 20 or 40 years ago, is the speed at which change occurs. What existed 10 milliseconds ago, when everyone was content to stream their 20-21 across multiple venues, changes in an instant and it can very easily be five pips away in not a lot, on not a lot of venues. All the analysts can do is make a best guess – and that, of course, comes with the disclaimer of “past performance etc etc”.
I said to my friend who referred me to the article last week, that a new generation of FX people are waking up to reality, but I am genuinely surprised if people are worrying about this – because, as noted, nothing is new. It could be that clients, who have been drip-fed promises of deep, robust liquidity for a decade and more, are struggling to come to terms with the “new” market, in which things can turn ugly in a millisecond. If they are, then welcome to the club, it has your predecessors as members and all you need to do it pull up a chair and listen to the tales of woe when they couldn’t get a “decent” price in 50.
The crux of this matter is, and remains, perception. Too many people have become too comfortable with tight spreads and low volatility, but the nature of the beast is that when the smelly stuff hits the fan, two things happen; liquidity thins out, and prices go wider. It has always been the case, and frankly, I can’t see it ever changing.
Looking at markets and speaking to people over the past few days when things have got truly chaotic, three things have become clear to me. First, there has always been a price for whatever amount someone wanted, but the “problem” was they didn’t like the spread. Secondly, in “normal” amounts (1s probably), the market traded at frequent and regular price points – there were no huge gaps, at least in G10.
The third observation is one for which I have to apologise (again), but facts are facts: the liquidity mirage exists and has never, and will never, go away. The only true definition of liquidity remains…you can get as much done as you want when you’re wrong!
