As well as the obvious areas around liquidity and spreads, the pandemic appears, if the anecdotal evidence of many of my recent conversations is anything to go by, to have reignited discussion around what constitutes best execution? I’ll be honest and say that I thought this subject was as settled as is it ever going to be this time last year, but as always seems to happen in markets, it only takes one event to overturn the apple cart.
When people talk to me about how they evidence best execution it does elicit a wry smile on my part, for earlier this year any price was a good price and people hit them freely without thought to the consequences. Market impact, as a concept, was very much relegated to the back of the bus as everyone ran for the comfort of the first price. As I have mentioned previously, however, there has been something of a mean reversion back to the status quo, with many LPs who ran for the hills returning (bullishly, not sheepishly it should be noted – they are quick to claim they were “always in” even when the evidence suggests otherwise) and this has prompted further introspection.
The problem is the short-term nature of modern thinking – yes there were LPs who genuinely stayed in and did their best and in a different world they would rewarded by being part of a smaller liquidity panel at most of their clients. The problem is, as one buy side execution trader put it to me recently, under the best execution policy currently in operation at their firm, they were unable to reward that loyalty when the other LPs returned.
There are going to be times when an LP has genuine issues and has to step back for a period of time, but to my thinking, as long as they are honest with their clients, they can be forgiven that – if the analysis of the relationship is for longer than one month or quarter. The problem is that at most firms loyalty and robustness of service cannot be measured – how do you put a number on the potential cost of another potential liquidity event and your best LPs potentially not giving you their best stream because they have seen volume and wallet share diminish? There’s a lot of “potentials” there and that word is the anti-thesis of the modern, data-driven market.
I seriously question whether a customer should be concerned about the odd tenth of a pip in pretty quiet conditions if the chase for that undermines relationships that are vital during times of stress.
I think another factor that has prompted this reassessment of best execution is peoples’ genuine surprise at how well the algos worked during the peak liquidity crisis. The opportunistic, liquidity seeking strategy in particular “had a good crisis” and this has prompted several firms to reconsider how they approach markets – suddenly spread capture is a reality to many of them. This does, of course, come with more buy side firms assuming market risk, and sometimes you wonder if they truly understand that in FX, but overall, the current thinking seems to be ‘use more algos, assume more market risk, capture spread’ (and as an aside, this assumption of market risk surely means more buy side firms need to adhere to the FX Global Code?)
Such a trend also, it seems to me at least, provide a much-needed boost to the agency algo industry, although again, I cannot see how third-party firms can really compete. The reason the opportunistic algos worked so well is, in part due to the sheer size of spreads back in March and early April, but more pertinently, because a lot of providers were able to offer access to a private, internalised, liquidity pool. You can work it any way you like, but when it comes to best execution of anything other than the basic market amount of one to three million units, keeping it away from the public gaze is the best option.
Sadly, what seems to me to be the ideal solution to the issue of best execution is beyond the abilities of the vast majority of firms in the FX market, and that is flexibility. If the past year has taught us anything it is that, once again, liquidity conditions can change very quickly and in those circumstances an execution desk needs to have flexibility to change best execution policies. Yes there are times when a bigger liquidity panel is better, but generally speaking that is when nothing is happening because, funnily enough, a lot of the flakier LPs are very prominent when nothing is happening!
I seriously question whether a customer should be concerned about the odd tenth of a pip in pretty quiet conditions if the chase for that undermines relationships that are vital during times of stress. For – and this is a regular feature in this column – the LPs are taking a much closer look at the value of their relationships and are dedicating their streams accordingly. Yes, on the odd occasion a customer may be able to grab a tenth of a pip in one million (and probably signal to the whole world at the same time), but how does that gain stack up against the potential hit from trying to get 10 million or more done in stressed conditions when most of the good LPs have them on their ‘C’ stream which is either as wide as the Atlantic Ocean or is turned off?
Yes, they can go to the algos, but do these clients really want the market risk? There was one crucial aspect of the March mayhem that was overlooked by the champions of the algo (and, I stress, I remain with them in spirit) and that was the general uncertainty of market direction. Opportunistic algos and spread capture are good in times of confusion, but what about when a market is taking a dive or soaring? In those circumstances there is little confusion, just a heck of a lot of bids or offers – none of which are any use. These conditions offer nothing more than a parade of optimists hoping someone hasn’t been paying attention.
So by all means we can continue to have the best execution conversation but can we make it about more than just spreads, market impact and depth of book. Is there a way to solve for those “potentials” I mentioned earlier? In our data-driven, risk-averse world, we need a way to quantify the relationship because if there is one thing of which I am convinced going forward, it is that even more focus will be placed upon this aspect by the service providers, and that means the customers have to be ready.