Smarter trading – Data and auto execution
An interview with Gareth Coltman, Head of European Product Management, MarketAxess.
How is automation manifesting on the buy-side desk today?
All the buyside firms I speak to are working to build automation into their workflow – even clients who are not traditionally prime targets for automation see the benefits that it can bring. In addition to increasing efficiency, auto-ex allows firms to systematise their best-ex framework.
At a desk level, firms are attempting to implement this framework by connecting together pre-trade decision making, execution, post-trade performance and analysis. Across the credit markets, firms are at various stages in developing that lifecycle through a mixture of third-party tools, execution venues, and trading solutions architecture. We’re helping by making it easy to plug in our tools and data assets wherever they can add value to the desk.
We launched auto execution in the MarketAxess workstation a year ago, and in January 2018 we had a small number of early clients testing the workflow and executing a handful of trades. Now, as of January 2019 we’ve seen billions executed, and have clients in Europe auto-executing thousands of trades. What’s helped is that we worked on an automation solution that allowed firms to adopt it gradually – allowing clients to start automating via our workstation, rather than manually working lots of orders. That allowed them to get confident about the process and the rule sets that they designed and to carefully analyse whether the data was telling them what they expected.
These are big steps for a lot of these firms – and what we see is a lot of clients wanting to roll out automation gradually. But what’s exciting is that now we have a small but growing number of clients who are implementing complete no-touch trading. By which I mean, they are controlling their automation entirely from their own order management system (OMS).
What does the adoption curve look like?
The growth has been very rapid, and I think it’s going to accelerate further. Active funds of all sizes are using it to help manage peaks in workload. Whether a big inflow, a transition, or a market event they welcome the assistance to manage these peaks in order flow. At the other end of the spectrum are large global asset managers who are aggressively competing in passive products and looking for a solution that can execute a substantial portion of their regular order flow.
How can you help firms to make their trade execution desk a differentiator for winning mandates?
There are two areas; one is to reduce cost through automation, and the other is how to add alpha to the execution. Both are data driven.
Adding alpha is about finding opportunities – finding liquidity at an opportunistic price. One example is using indications to proactively aggress liquidity. We are able to give clients a feed of Open Trading data, which tells them when there is another natural client opportunity out there and they can react to that.
Fundamentally it is about capturing and analysing the right pre-trade data to spot a market opportunity. For instance, there may be an ability, because of market demand, to pick up a certain bond at a cheaper price as an alternative to the bond that the portfolio manager was looking for, and which will provide similar investment factors.
The trader who is gathering this data can aggregate it and present that back to the portfolio manager. Reviewing investment decisions through the lens of liquidity can allow them to add alpha immediately through the execution process. But clearly, for traders to add alpha in this way they need access to the widest, most accurate set of liquidity, pricing and execution data. It is, of course, one of the reasons MarketAxess’s Composite+ has been so popular with clients.
In those firms who see their trading as a differentiator, is it true that many desks are taking a componentised approach to building their technology?
Yes, it is true – because they want to tailor their tech to their trading style. And we work to provide our offering using standards such as FIX protocol to support that approach. Traders want freedom to choose a provider of data that can be applied throughout the complete trade lifecycle. So it’s great if I have a transaction cost analysis (TCA) tool and I have a reference price that I can use in that TCA process, but if I can’t employ the same data usefully on a pre-trade basis and during the execution decision-making process, then it’s not going to help me. So I need something that I can plumb in throughout the trade lifecycle.
What are the cul-de-sacs you see clients getting stuck down?
Many clients are at the beginning of investing in a technology stack, which will support the execution framework that they are building. They need to think very carefully about which platforms and vendors they choose, because that’s what will dictate the path they travel down. It’s likely that that will be a very tightly-coupled, close relationship. So traders have to make sure that they are fully thinking about the ramifications of that level of integration, and to make sure they choose the right partner.
I speak to asset managers who have invested in a technology, and years later are still struggling with the implications of that decision on their business. So when clients ask me how to make a choice, my advice is always to do lots of research, and make sure you choose a partner who is really focused on your specific needs and, importantly, asset class.
How do you see traders’ roles changing in 2019-20?
The job of the buy side trading desk is becoming increasingly complex, and on top of that complexity, the pressure to simultaneously increase efficiency and drive down costs makes it a very challenging (and exciting!) place to be. The macro-economic situation is very uncertain – and looks set to be so for some time! So traders need to make very sure that whatever technology they have, it not only works well in the current market environment, but is also set up with the right checks, balances, controls, supervision and, importantly, adaptability so that when the market environment changes dramatically, they remain in control.
With our automation offering, we have thought deeply about how we provide that level of control and visibility.
Finally, AI is an exciting new area for our industry. How does AI figure in the future of automation?
We believe, frankly, that when it comes to automation, the best data wins. That means both the data and the analysis and learning that sits on top of it.
One example of this approach is MarketAxess’s Composite+. Here we’ve leveraged machine learning to take all of the data that we have, across different bonds, and learn which sources of data are likely to be the most important to predicting accurate future bonds prices. We then apply that learning, every day, to continually improve the delivery of bond reference prices that have what we see as a phenomenal level of accuracy.
I think, in the future, we may see machine learning helping traders to further improve their execution decisions by learning from their behaviour and suggesting an optimal route to market. In this way the human trader and machine can work together to continuously understand and improve the execution decisions and outcomes. It’s an exciting vision of where human and machine effectively become one.
©The DESK 2019
