As CME Group prepares to announce the launch of a new platform for EBS Direct, Galen Stops has the exclusive interview with Jeff Ward, global head of EBS, about what the platform will offer and how trading venues can differentiate themselves in an increasingly data-driven environment.
Galen Stops: Why did you decide that it was necessary to re-platform EBS Direct?
Jeff Ward: The idea behind launching EBS Direct in 2013 was to offer a relationship-based execution platform to complement our central limit order book (CLOB) driven platform, EBS Market. By leveraging our liquidity management know-how and our desktop footprint we built up traction on EBS Direct very quickly, but we were more successful in attracting the manual user trader base to the platform than we were with the API trading world.
Therefore, the genesis behind the rebuilding of the EBS Direct platform was to provide next generation execution capabilities and create the most deterministic and fastest trading platform in the world in order to attract larger banks and more sophisticated clients that not only trade through an EBS GUI, but primarily use an API.
GS: Ok, but what does “next generation execution capabilities” mean in this context?
JW: It means that we’ve created a much more flexible, scalable and very high-performance platform that is highly deterministic and allows clients to execute in a number of different ways.
For example, they may want to execute in a complete uninterrupted pass-through mode, meaning that we just provide a very good, fast technology bridge between liquidity provider and liquidity consumer. Or, if they’re a manual trader, they might want to continue to have us aggregate a book for them and trade on an order book via a GUI – this way, we can support new order types with integrated analytics.
What we’ve built is a next generation version of what we created five years ago and we believe that it will be best in class in terms of overall speed of market data distribution and speed of execution.
GS: You’re talking about the platform being fast, can you give me an indication of what kind of speeds we’re talking about here?
JW: The target is to be, for both market data and order processing, under 50 microseconds – and right now our performance testing is on track.
But it’s not just about speed. Speed is important because the platform being fast means that firms are always trading on the freshest rate in the market and therefore their missed trades and
reject rates will be lower. But if you’re a manual trader on a GUI it’s also about the functionality available – this is why we’re adding different order types, such as the ability to execute on a VWAP, we’re adding all kinds of different potential views into the order book that we didn’t support in the past, and obviously it already offers different types of execution, such as sweep and execution on a single ticket.
So although speed is very important for certain client types because it provides a very deterministic, pure trading experience with their liquidity providers, the rebuild is also about building out a wider array of execution capabilities as well.
GS: What stage of the build are you at now and when is the new platform going live?
JW: It’s already live in an early stage beta. We have a couple of clients live in beta, and we expect to have between six and 12 live by the end of the year.
This first phase is for clients to consume market data through the new system. In the later phases, we’ll add execution and then after that, we’ll replace all the connectivity with our liquidity providers. The various benefits of the new platform will be available to clients in phases from now through the next 18 months or so, with the aim of then fully migrating everyone off the old system and onto the new platform by the end of 2021.
GS: Moving away from the new platform itself, EBS Quant Analytics has been a big initiative for you. What impact has that rollout had so far?
JW: We’ve had good feedback from clients, but we’ve also seen some statistical improvements on the platform. For example, we’ve seen the average hold time for our top 10 liquidity providers on EBS Direct reduce 61%, coming down from 93 milliseconds to 37 milliseconds. We’ve also seen reject rates from the same LPs drop 41% from 5.3% to 3.15% over the past 12-18 months.
While I’m not claiming this is entirely due to our analytics, because there are broader drivers of these trends at play, but the fact that we’ve given tools to clients which help them look at spread retention, cost of rejects, frequency of rejects and length of hold times has certainly helped.
Giving both liquidity providers and liquidity consumers access to real-time analytics and reports online means that clients (liquidity consumers) are now able to benchmark their execution and make much more informed decisions around the type of service they are getting from the platform, while LPs are able to improve the yield that they’re getting.
Looking ahead, I think that the next generation of clients that are going to be using this type of data will be focused more on automating decisions in real time rather than looking at a report or a GUI. At some point, firms will be making real-time decisions about adding or removing a liquidity provider or dynamically changing how they price clients based on this data. I think that will be the next step around analytics.
GS: Data seems to be front and centre of everyone’s business models these days. Is data the new battleground in FX? How can data providers distinguish themselves from competitors?
JW: I would say data has been a battleground and will continue to be so as the market evolves. But you have to think about it in a couple of different ways because there’s a difference between the CLOB world, where it’s an all-to-all type model, and the relationship-based world.
On the CLOB side, I think that things like rules around minimum quote lifespan and fill ratio requirements have eased some of the concerns about the race to zero. At EBS, we then started to look at market data and reduce the market data update rates. For example, we reduced the updates for our EBS Live Ultra product from 100 to 20 to five milliseconds and that helped our clients have a better view of the book to manage their risk, or to actually use as an input to price their own client. This in turn helps them reduce hold times and last look times.
When you look at differentiating your service in the relationship–based world, where firms are providing liquidity either through a direct feed or through a venue like EBS Direct, data and execution certainty are – along with the market impact of a trade – the key drivers. Those are the three magic components of providing a really good service.
And that’s where our data analytics platform has been significant because it helps the LPs improve the yields that they retain on our platform, while also collapsing the overall cost to trade for clients.
GS: More broadly, in a world where liquidity – certainly for G10 spot FX – is commoditised, what do you view as the key areas of competition for platform providers?
JW: You have to view this question through a different lens for different types of venues. For CLOB–driven primary markets like EBS, and even the FX futures world, it’s really about ensuring that your market data and the platform you’re providing is robust, that it’s underpinned by the right rulebook and that clients are able to get the best execution possible.
This is critical when you’re a primary market like EBS, because when something happens in EUR/USD or NDF pairs or in CNH, that’s where people are looking first. It’s sort of where the pebble hits the pond in terms of where activity happens and then the secondary venues react to that.
On the more relationship-based trading platform side, I think that you need to provide flexibility but also the best performance in terms of execution, certainty and market impact, and that’s why we’re re–platforming EBS Direct to make sure we do this.
Then over time, being able to combine these execution venues, alongside the most liquid FX
listed futures and options products, through a single front-end and single API, underpinned by market data products and supported by the analytics proposition is a really unique proposition that we can offer at CME Group.
GS: So you’re talking there about being able to access all of these liquidity pools through one front-end, do you think that this is where the market is headed? That the drive for efficiency will lead to consolidation in the platform space as market participants increasingly look to access liquidity through one venue?
JW: I think that a lot of consolidation has already happened. Yes, there are still a few independent platforms out there and over time they might be assets that people will look at because in my view this is a scale business over time.
To your earlier point though, parts of the FX world – such as G10 spot FX – are relatively commoditised, there’s no shortage of places that you can execute that type of product, so I think it’s going to be more and more about scale and about bringing something unique and providing better benefits and efficiencies to market participants.
GS: And finally, keeping with the same theme, it’s been nearly a year since CME Group acquired EBS — how far along are you now in the integration process?
JW: We’re still very much in the process of bringing the two firms together. A lot of progress has been made around planning the integration of the EBS technology stack into the CME one, which is a multi-year project which we’re aiming to complete by the end of 2021.
GS: But is there clarity now on what the roadmap and final technology structure will look like?
JW: While there are still parts that are being finalised in terms of what and how we align with clients, internally we have a strong vision of what the future looks like from a technology perspective. And that is about ensuring we integrate the EBS matching technology into the CME Globex matching technology whilst also ensuring that we keep all of the market structure components of EBS that we’ve built over the years and make the platform what it is.
Above all, we think that the migration will improve functionality and performance on EBS and that our combined global customer footprint will encourage liquidity and participation across both the cash and futures marketplaces.