P&L Report Card
Although several firms have been focused on this for years, it did seem like a year or two ago, a shroud was lifted from so many in the foreign exchange market and they were able to see what best execution really meant. Gone were the days of ringing your mate to get the deal done, or directing a trade to a provider as a “reward” for business elsewhere – suddenly FX execution was a big deal.
If we are being honest we still think there are too many clients out there who don’t take their FX execution seriously enough, but for those that do, the sell side has come up with solutions that offer plenty of choice.
This award is for all round excellence when it comes to getting the deal done. We look for good analysis of market conditions through the pre-trade analytics package; a range of execution options including algos, risk transfer, vanilla orders and, yes, a voice desk. This needs to be wrapped up with excellent post-trade functionality not so much around the TCA (in this area the availability of a third party service such as BestX is ample), but in the ability to allocate, roll and amend trades.
We are not suggesting for a minute that a bank’s own TCA tools are inadequate, they are anything but and can be used to analyse future execution policy, it’s just that in this world where the accused (if they are a bank) is still viewed as guilty until proven innocent, suspicion means a third party has to verify everything. That said, it was interesting the number of providers who believed that their services had been improved by the advent of a dedicated third party service for FX TCA – proof again that FX and equities are not joined at the hip.
As we have already alluded to, the big growth last year was in the pre-trade analytics, although it is interesting to note that a number of providers are seeking to offer real-time analytics on how the algo is performing, along with the ability to interact with the strategy as this information feeds in. If we had one thing we would like to see that wasn’t shown to us – it probably isn’t too far away – it would be a chart plotting the expected execution, overlaid with the actual executions. To return to our theme of visualisation, this would be an easy way for an execution desk to see how well, or not, the algo is performing compared to expectations. As things stand, BNP Paribas and JP Morgan are the only two to offer this currently.
Generally speaking, when it comes to the whole package around execution the field is crowded, but not to the degree it is when it comes to just algo provision or just analytics – several banks have one or other of those products in place but have not quite pulled everything together just yet.
Of those that have, we would highlight, BNP Paribas, Citi, Deutsche Bank, JP Morgan and Morgan Stanley as being at the top of the tree as things currently stand. Others are working on adding to, or finishing, the execution “canisters” – most notably Goldman Sachs which is in the midst of a major piece of work here – but for now, the aforementioned stand out in FX terms, with UBS very good across asset classes, especially equities.
If pushed to narrow that field down further, we would say that BNP and JP Morgan have the edge, although as already noted in this article, Deutsche Bank rolling out analytics and algos (including a build your own analytics suite which we think will gain serious traction), and Citi re-vamping its pricing and risk engines, both mean serious competition is on the way.
2017 was a good year for BNP Paribas, which continued to add to an already excellent execution package. The bank added more client specific controls to its strategies, as well as the ability to STP orders back from Cortex iX to the client’s OMS.
Pre-trade the bank’s ‘what if’ type analysis methodology remains excellent, allowing users to see what the expected execution pattern is for their order by different strategies. BNP can do this (as can others of course) because of the limited number of strategies available – increasingly we hear from clients that one obstacle to algo adoption is the number of strategies around – as well as the controls around them.
BNP has also gone all in with independent TCA provider BestX, a sensible move for a bank that has historically been focused on the agency model for its algos. This move has also, however, allowed the bank to do more with its own data and analytics and it has created a liquidity lab that provides empirical studies for clients and goes into their execution style and quality in greater depth. This has allowed many clients to take what is an important step regarding their execution analysis, moving from a box ticking exercise to a real effort to better understand and nuance execution style.
The bank has not been totally focused on the plumbing, however, for it has added a strategy that, for us, fills an important gap in the bank’s offering – one that accesses internal liquidity. We have already noted the importance of internalisation to good execution and even though BNP is not the biggest house on the street, it has a robust internal pool that it’s algo clients probably need to see.
Winner – JP Morgan
In the equities world the last two years have been about the “algo wheel”, where a client has a range of strategies available to it and chooses one on the fly, guided by analytics as to which is best for the current circumstances. This works in equities markets, but we are less convinced – again – that it will in FX circles, which means going to a provider that has it all – deep liquidity, great execution functionality, real time analysis of conditions and, importantly, flexibility around which strategy or strategies to use.
In FX terms at the moment, we believe that house is JP Morgan. Three or four years ago, we referred to a dramatic upgrade in JP Morgan’s pricing capabilities that had seen it soar to the top of the tree on several of the multi-dealer platforms. This success story has seen the bank grab more market share, which has enabled it to paint a better picture of the market for its clients.
The bank’s algo package has long been respected, but the past 18 months has seen the development and roll out of a new algo ‘canister’ (we prefer that in FX terms to the ‘wheel’) and enhanced strategies.
Within the canister, a client can select strategies, they also have the ability to apportion an order out to different styles, such as 50% TWAP, 25% IS and 25% Peg, as well as access the market data analytics. The latter provides real-time data on volatility, spread and liquidity and the user hovers the mouse over the individual aspects to see whether this is up or down on the historical average. A fourth tile in the window offers a measure of liquidity over the past 24 hours and again, this can be compared to the longer term average.
This ability to offer critical market data on the same page on which the strategy is created and managed is very workflow friendly (yes we do mean efficient!) although if we could see one change it would (again) be enhanced visualisation in terms of alerts if any of the four market data suites are at extreme levels, both positive and negative.
JP Morgan has rolled out an adaptive algo that would take advantage of liquidity events, however we think there are a few execution desks around that would like to be able to use the excellent interactive tools provided by the bank to change the strategy according to what they are seeing in the data, without having to hover the mouse.
One thing of note from JPM is how the data analytics buck the industry trend by including some of the bank’s internal liquidity – this is a real differentiator in our view.
The aforementioned new strategy, Aqua PoV, calculates the market impact from a particular order and then adjusts the strategy accordingly, meaning, for example, that if it believes there will be impact, the TWAP order is front-loaded. This strategy is also, interestingly, not bound by time – if the data tells it conditions are not good, it will slow down or, presumably, stop. In a world seeing more mini-flash moves this is another line of defence beyond the price limit box (something, we are told, few customers actually use!)
All executions can be changed on the fly, including the ability to switch volume from one strategy to another if more than one is being used and we cannot stress enough how this speaks to a concern of many on the buy side, especially the execution desks. These are trying to add value and simply pushing a button and watching the algo do its work inevitably doesn’t do that – and probably leads to a tap on the shoulder some months later.
What we really appreciate about JP Morgan’s execution package is how it offers the user a suite of products, on one screen, and allows them to change almost any aspect of the order execution process. It is hard to explain without making it sound wrong, but there is a feel of the Game experience in that the player (user) can adjust just about anything on the screen in front of them to gain an edge – if nothing else this should stand the bank in good stead when it comes to the next generation! We are not suggesting for a moment that a user plays around with the strategy as much as a gamer would, more we are saying that if something fairly dramatic does happen, they have all the tools at their fingertips, on one screen, to help them react.
Although we are loathe to use the phrase when discussing anything best execution related, it has to be said that the algo canister on eXecute ticks all the boxes. It is integrated with third party solutions, has excellent analytics graphics and data, shows real-time execution analysis (graphically) and can be easily managed on the fly. It also has all the parameters necessary for high quality execution when building the strategy and allows the client to set a limit not just on outright market price, but by the all-in rate.
This latter point is helpful for many clients who have expressed some frustration in the past with providers only able to provide an absolute limit for a strategy rather than a “smart” limit that takes into account the average fill.
There is still some work to be done to convince many compliance functions at clients of the value of algos – that is ongoing and is very much about the controls, but there is also work to be done to convince some execution desks. JP Morgan has partnered with the ubiquitous BestX to help alleviate the concerns of the former, for the latter – the people at the coal face – it has developed an outstanding product suite that allows them to demonstrate their worth and their skill. That is what customer service is all about – a partnership that is mutually supportive.
The broader platform, eXecute, remains outstanding but if there is one thing JP Morgan has managed to achieve over the past 18 months it is to set a new benchmark for execution services.