P&L Report Card
If the Best FX Platform Award was not the most competitive in the 2020 Digital FX Awards, it is because competition in the execution space has become white hot. While some players may not have the time and perhaps the resources to develop their single dealer platforms fully, they have most certainly had time to build their execution services.
Partly this competition was empowered by some institutions deciding in 2016-17 to go the route of agency, thus making their development path a lot simpler – build some algos, roll them out, connect up to the market and away you go. Of course it hasn’t turned out like that – and at this stage we would like to engage in a little self- congratulatory nonsense for predicting this – because, as noted in previous years, an algo is only as good as its liquidity pool so why would a strategy ignore the biggest liquidity pools in the market, the major dealer’s internalisation programmes?
It is noticeable over the past two, maybe three years, that some banks have started investing in their principal businesses to bring them up to speed with their execution services. BNP Paribas did this three years ago after a short period in which it looked like the bank was going fully agency and over the past two years, the last 12 months especially, we have seen Credit Suisse engage in a solid effort, backed by the latest machine learning techniques, to build its principal business. Nowhere more is this effort noticeable than in how the bank likes to tout its streaming capabilities in large amounts – up to EUR 500 million in EUR/USD for example.
Of course, this is not a one-way street, because we have also noticed how the effort on the part of some banks to deploy algos developed in their equities business haven’t worked, or that those strategies that just say “hand the order over to us and we will execute using our own algos” haven’t gained traction. We suspect the latter instance is a case of customers wanting to know more about how the strategy operates and the banks not wanting to reveal any of the secret sauce. That is their right of course, but it is hard to see that approach working in today’s transparency obsessed world.
That all said, if you are a client looking to execute large or small tickets in this market then there is not only a solution for you, there is a range of choice, although it is interesting to note that the one area of feedback that is not universally positive over Citi’s Velocity is around algo execution. The analytics work – and will work better in HTML5 – but user feedback suggests that many of their customers looking to deploy an algo, go elsewhere. The pricing capabilities of the bank are beyond doubt and have re-established Citi as one of the true top players in FX, it is surely only a matter of time before this spills over into the algo execution piece.
Elsewhere, Credit Suisse’s AES remains tremendously popular, although as we noted earlier, in the tremendously competitive space that is execution services, the bank is suffering from a couple of product gaps, namely pre-trade analytics. The past year has seen the bank enhance its execution dynamics, but again, to highlight the problem of a bank seeking to make up for a period of under- investment, the work it is doing on, for example, using AI and ML for directing the child orders, has also been done elsewhere. AES is excellent, but unlike eight years ago perhaps, it is no longer a clear leader in the field – pausing investment or having to invest elsewhere will do that.
To us, BNP Paribas has taken on the role of something of a thought leader in this field. It was very early to develop adaptive algos of the type now rolled out by several of its competitors, it also led the way in taking a portfolio approach to post-trade TCA. Do the immediate measurement via an independent provider by all means, is the message, but if you really want to learn about your execution framework, step inside. Last year saw us tease the rollout of Alix, a virtual execution assistant, which actually occurred a couple of months after the awards were judged. This has proved to be very popular with busy execution desks because it allows them to engage the algo but retain the ability to interact with the order in flight, but on an exceptional basis.
We have already noted Deutsche Bank’s move forward in FX and its analytics suite, Market Colour 2.0, continues to offer ground-breaking analysis tools to inform the execution process, both pre-, intra- and post-trade. As the bank invests further and its confidence grows accordingly, we would expect to see even more participants turn to the bank for their FX needs. If there is one area that feedback suggests the bank can enhance its execution experience further, it is one or two more algo strategies. Currently the bank relies, as it has done historically, on the algos its own trading desks use and while there is no doubt this is the best the bank can offer, as noted, some clients prefer to have a little more control. The suggestion from these parties is that perhaps Deutsche can offer one or two “plug and play” strategies with the ability to alter the path if the client so desires.
Other banks that are very much part of the competitive landscape here are Goldman Sachs, thanks to its portfolio algos and work with Bloomberg, (another tease for that!) as well as Morgan Stanley, which continues to innovate in the execution space thanks to its QSI team, and a relative newcomer to this segment, HSBC. The latter is an interesting one for us because we have yet to see the algo offering for ourselves; however, feedback from users is universally positive and who are we to argue with that?
Before we move into the winner of this award, one thought bubble about how this segment will look next year. This publication’s managing editor has long had a problem with excessive liquidity recycling and is now starting to wonder about algo execution services. We live in a white label world the minute we step outside the top group of banks, so at what stage do we start to ask who actually operates and manages the strategies and infrastructure?
During the demonstration part of this process it was noted on more than one occasion that on platforms such as Bloomberg, there are 25-plus algo providers. Using all the brainpower available, we are struggling to come up with half that number in terms of serious providers, although of course it should be noted that algos do come in many flavours and some of those are for executing smaller orders in niche markets.
It is hard, however, to believe that there are more than 25 providers of unique algo execution strategies – some of them have to be white labelled surely? If that is the case, then does the debate shift to some form of certification process wherein providers have to detail who owns the IP behind the technology and the strategies, and perhaps most important of all, who is accountable if and when things go wrong?
No doubt this is part of the work of the Global FX Committee working group on algos, but we suspect it will become a bigger issue in the coming months, especially if market conditions continue to be as challenging for people trying to execute hedges as they currently are.
Winner – JP Morgan
Three in a row for JP Morgan in this category as the bank has taken the hard work of the past three years and built on it. The key to this win, probably the factor that kept JPM just ahead of the field, was how it took one of last year’s outstanding products, Algo Central, and folded it back into the main eXecute platform to help inform the micro decisions.
We remain big fans of the Algo Central concept, the canister from which a client can build a portfolio of strategies, but the reality of the world, as JPM has discovered, is that there is probably more growth potential in using algo strategies for what would normally be called mundane orders of modest size.
We should stress that the need for a flexible and granular approach to the execution of large orders is not going away, rather what JPM has done is take the logic involved in Algo Central and provide it for smaller tickets. Still available is the ability to break a ticket down across strategies and the ability to interact with the various strategies at any time. Equally the vitally important “get me out” button, more accurately using JPM-speak “Quickfill”, allows a user to execute the balance of an order against the bank’s streaming price for the appropriate amount.
The bank has also, as part of the reintegration, made the entire order entry process much simpler and therefore quicker. Still available are the superb pre-trade and real-time TCA tools, the bank has also added NDF algos to its existing limit order functionality, using internal and external (on and off-SEF) liquidity pools.
Finally, when it comes to execution tools, and not forgetting the aforementioned StratX, is DNA, the bank’s Deep Neural Network for Algo Execution. DNA applies machine learning to optimise every child order within the larger ticket. Where most algo strategies continue to offer a pre-ordained execution path, DNA is all about optimisation at micro order level using its own observations of the market and experience.
Underpinning everything is a truly excellent analytics suite, something that has been a strong suit of JP Morgan for some years now. From ECN liquidity estimation, through predicted execution paths, to an alerts-based service that reacts to changes in market conditions, eXecute offers clients the fullest possible picture of FX market conditions in an easily digestible format.
Part of that suite is Aggregate TCA, the bank’s performance analytics package. From high level statistics such as volume traded and number of orders, Aggregate TCA provides a holistic view of clients’ execution performance, thus filling another gap in the bank’s service. The service provides overall and key metrics for this performance over a period of time (peer comparison is very close to roll out as well) and clients can drill down into the data to great depth and analyse performance by things such as time of day, strategy, currency pairs, market impact and liquidity sources. What we especially like about Aggregate TCA is how it offers great graphics for the high level performance data, but also provides the really detailed data for clients’ own quants to dive into.
Unlike some other of our top awards where we see change coming slowly, the execution services space is very fluid, innovative and fast-developing. It is a testament to the commitment of JP Morgan that not only has it maintained its leadership position in this crucial area, but it has, if anything, extended it a little. That may not remain the case, there is a lot of work going on in this field at this time, but something we have identified in recent years about the bank’s FX business – the absolute refusal to accept or condone complacency in any form – means for the third year in a row, if you want the best execution experience, you go to JP Morgan’s eXecute.