Ediphy’s CEO, Chris Murphy, explains how traders can engage with the latest protocols and technology through a single access point.
The DESK: Firstly, what has guided your engagement in the development of Europe’s consolidated tape?
Chris Murphy, Ediphy: Our interest in the consolidated tape has come from being eager end users of that transparency data. A key part of the electronification of fixed income markets is having access to trade data to build the picture of true liquidity and see where instruments are actually trading. Indicative quotes have been out there and prevalent for a long time but we understand the market has been frustrated around the reliability and validity of those indications. Those concerns have become even more pressing throughout this year, as markets have started to dislocate.
How does that availability of data evolve the way that people trade?
If you’re trying to automate more trading and use the data to optimise your execution, to date traders have had to rely on human heuristics to try and figure that out. Once you have more data at your disposal, you can start to use data analysis techniques to get better results. There are a variety of elements that piece together. The base layer is access to better data and more sources of data. The next level is the capability to process that data, clean it and manage it, then build analytics on top of the data. The final piece is integrating that into your trading workflow. At Ediphy we think there’s a service that we can provide which helps our clients to bring some of these capabilities into their arsenal allowing them to increase automation, achieve improved prices and better time their market entry.
You offer both low and high touch trading support analytics; which technical challenges did you need to overcome providing that?
The nature of the asset class means certain parts of this market will not be able to just adopt a low touch approach. Ediphy ensures low touch and high touch options are always on the table, optimising execution in the low touch and providing the data that traders need, all in one place, for the high touch. Putting this all together is a non-trivial, not to mention costly, exercise requiring expert skill sets across data engineering and data science domains, but we think it is essential for the modern trader to avoid being left behind!
How are low and high touch trading triage being framed?
Ediphy’s offering blends the two. Automated trading strategies that are hard coded, processing orders beneath a certain size as low touch and those above as high touch are blunt tools. That threshold should be context specific, relative to the market. Instead of making an initial high level decision as to whether to work an order high touch or low touch, we enable you to host all of your orders in our platform and get the best of both worlds. We scan the market for liquidity in each instrument and use historical and real time data to understand which liquidity pools and protocols are best suited to work that order. Context specific data is presented to the trader for each order, with low touch orders being sent to us for active execution, whilst more difficult to trade positions can be left resting passively in the platform as it still continues to monitor available liquidity whilst the trader also works those orders manually. The idea is to have human and machine working together, augmenting the trading desk with technical capability.
Where do you see the barriers to evolution in trading at the moment?
In busy markets, traders don’t have the bandwidth to explore fresh ideas and opportunities. There are new venues and novel protocols cropping up all the time, which creates a massive hurdle for people to try and keep on top of what’s out there. But even if you are interested, the adoption cycle to get a venue or protocol onboarded and integrated into your system comes with considerable overhead. We help remove those pain points. The picture becomes way more difficult if you consider complex operational workflows such as interest rate swap (IRS) unwinds, for instance. We have created a solution to assist with that process and the resultant execution.
How can traders get around that?
They onboard Ediphy, we do the rest. As well as providing the technology, importantly we’re providing a managed service connecting to all of these liquidity pools, bringing them together and normalising all the different protocols. Our clients instruct us on what they need done and we handle the rest. We abstract a lot of the incidental complexity, so our clients get best execution and the integration of different capabilities by us, creates really real value for our clients. Furthermore, given our single technology stack approach, our clients will be able to take advantage of incremental additional services over time as they will be natively integrated with the execution layer, allowing them to reduce their legacy technical debt in the future.
Are we likely to see a greater proliferation of new offerings at present?
Fixed income is far from homogenous. Certain instruments are more suited to a request for quote (RFQ) and other instruments might be more suited to a request for market (RFM) or a periodic auction, depending on how you’re trying to execute your order; by price, time or market impact. New venues are being introduced all the time to cater for a different segment of the market as the market evolves. You shouldn’t be constrained by the venues you’re onboarded with, you should be able to access all of these protocols in a very smart, data-driven way. Ediphy aims to plug into everything, so our clients get the benefits without the heavy lift.
How can they fit into a workflow decision-tree that manages high touch and low touch execution?
We think execution management systems are a hard product to design for fixed income. A lot of venues in fixed income don’t support a clean API approach that allows RFQ negotiation through that API. The RFQ workflow from one venue to the next is different. As a result, the user experience when you’re bringing together RFQs, central limit order books or other protocols becomes very messy.
Ediphy’s approach is to bring orders into a common blotter and allow those orders to be worked manually, or just sent to us for low touch executions. We navigate venues and protocols with instructions from the client’s trading desk, on objectives – how aggressive or passive to be, and execution priorities. We take those instructions and translate them into an execution plan. That could be to wait until you’re able to get bilateral dealer liquidity on Bloomberg chat or routing directly to a central limit order book but having man and machine working together in an integrated workflow determines what is optimal.
What are the learning machine learning processes you’ve used to facilitate decision making?
We’ve built evaluated pricing, based on all of the instruments that have printed in the market. This is where a consolidated tape is very important. You can infer where a certain instrument should trade even when they haven’t traded for a while, based on where other instruments have traded more recently. We have techniques to infer where fair value should be. Other techniques model dealer axes to build a picture over time as to where we expect dealers to be stronger or whether they’re going to be axed the other way around. As we execute in the marketplace, we build that picture by using our analytics. That helps us to determine optimal liquidity sources, creating the pathway to improved decision making and better end results.
©Markets Media Europe, 2022
