Tim Monahan, senior director of product management at ICE explains how ICE continues to develop services and products that better support traders in fixed income markets, even as those markets evolve.
What is ICE doing to support better pricing for traders?
Tim Monahan: ICE has heavily invested in the evaluation process, it’s our foundation. We’ve accomplished that goal through a multi-step, multi-year process of reorganising our quant, data science and methodology team then pairing that team off with our evaluations team.
By combining the talent of all those individuals, it’s allowed us to leverage new techniques, to synthesise market data and to break down large datasets. We can use the information in a more effective way. Pairing up the subject matter experts – the evaluation team – with the right set of quants and data scientists, is what led to the noticeable improvement to pricing.
What drove you to make that change; have you seen a new generation of traders who put all their numbers into Python?
That’s exactly the shift. Graduating from Excel to Python has been the shift we’ve seen with our clients and we’ve experienced the same progression within our evaluation process. I didn’t know what a gradient-boosted tree model was until recent years, but you get familiar with those terms and techniques.
Have you seen stylistic changes to the way people invest and trade in fixed income markets?
Clients are much more focused on attribute-based trading or investing, where they identify the components they are looking for in an asset class, such as time to maturity, amount outstanding, liquidity profile, yield, coupon, industry and whether an instrument is part of an index or exchange-traded fund (ETF), then try to find the right investment that accommodates all of those attributes via a list of bonds. They will then go out systematically or electronically, to try and purchase those assets. When someone goes out to purchase those assets, instead of buying a block of US$25-50 million of one of those names, they will instead, buy a small position across many instruments. The reason that works well for an asset manager today is because the sell side is so well equipped to provide liquidity on a wider range of bonds, at an average smaller size, which reflects how well aligned they are with the ETF create-redeem process. I think this new style, or approach, to investing only works because the quality of the underlying data and analytics has improved.
What effect has the expansion of trading protocols in electronic trading had?
It has had a positive effect without a doubt. Everyone’s doing list trading. Some of that is portfolio trading, or ETF trading, or some can be classified as attribute trading, but it all comes down to being able to get a list from a fixed income order management system and executing it electronically.
How are you presenting information to buy-side firms?
We recently published a white paper called ICE Trading Analytics: Serving fast-evolving fixed income markets, and it talks about the evolution of fixed income trading and how it was so important for the sell side to get a firm order, to work that order. They used to get that order by picking up the phone and requesting the order. When they didn’t have to pick up the phone, they had to work their magic through chat. Now it’s evolving to communication that you don’t explicitly see, travelling via messages by APIs (application programming interfaces). It becomes a function of the ecosystem traders want to live in. We work hard to make sure our data is available within whichever ecosystem that is, so we’re open to the platforms that are building out execution and order management systems.
Several years ago, we collaborated with Aladdin and Charles River to include continuous evaluated pricing (CEP). Since then, a host of other platforms have popped up, like Trumid, which now has CEP. Our data is seamlessly integrated within the ICE Bonds platforms as well. Additionally, we’ve done collaborative work with MarketAxess and Tradeweb on our end-of-day evaluations. A lot of clients are interested in end-of-day list trades, and they want to base the list trade on our end-of-day evaluation.
More recently we’ve collaborated with FlexTrade. FlexTrade is an EMS bringing together liquidity from a variety of different sources including our CEP and trading analytics. That helps clients identify trading risks and trading opportunities which works especially well for clients who use us for reference data and for pricing – it creates continuity amongst the front, middle and back office for their organisation. Another more recent collaboration has been with InvestorTools’ Perform, which again provides efficiencies to the front office, so having our CEP and trading analytics at the traders’ disposal creates a rich environment that helps them manage their fixed income order flow.
How is the need for different analytics and different data evolving?
Clients will require more real time analytics and data, but high-quality pricing is still by far the single most important element that they’re interested in. But that doesn’t tell you the whole story. To understand that you need to understand the pricing, the liquidity profile of the bond, the expected execution quality of the transaction, the cost of executing that order, the market depth, and the general tone or market sentiment for that issue on that day. We bring all those components together into trading analytics. Those have been the elements that clients have asked for more data on and as a result, clients can now use our data to feed those components into a model that will use that content to frame out the market for each unique order they look to execute.
How do you break that down in services?
We have evaluated pricing, best execution, transaction cost analysis (TCA) and liquidity. Those four are the corners of the picture frame, to get a clear view of the bond that you’re about to trade. Now we’ve added market depth and market sentiment to that picture. You need all the components that we’re creating which are a direct result of client feedback. Clients may use these components to build a set of rules to automatically execute or automatically bid or offer against a pool of assets.
How do you see automated trading progressing?
In the corporate space they are making good progress because market participants in that asset class, through the evolution of the ETF, has moved from being equity players into fixed income market makers. They are already well acquainted with the electronic trading of assets, and they’ve brought that skill and drive over to the fixed income market. There’s less risk involved with automating your fixed income trading today because the technology is there, and the data is better, so market participants are all doing similar things. If everyone’s trying to accomplish this goal, the current is in your favour.
