AxeTrading To Roll Out More Automation for Fixed Income Platform

London-based fixed income trading software provider AxeTrading is leveraging open source technologies to support more automation on its platform, AxeTrader.

“We are focused very heavily on rule-based automation, which gives a high degree of predictability but also allows us to plug in numerous different datasets and drivers into that automation,” says Mark Watters, co-founder and chief commercial officer at AxeTrading.

As part of its move towards better automation the vendor is gradually integrating open source event streaming platform, Apache Kafka, into its application stack. Event streaming is the capture of data about events, such as transactions—from databases, cloud services and software applications—and disseminating that in real time.

“Kafka is a high-capability messaging infrastructure that has the ability to plug in through things called ‘topics’ [the concept under which events are categorized within Kafka], different plugin points and different APIs,” Watters says. “We are purposely building the most high-performing open system that we can in order to allow for automation in multiple ways to be plugged in and out of that system.”

Watters says the vendor has so far implemented Kafka for only part of its infrastructure, but is gradually integrating the software across its systems. “It is available for cash bond trading on certain [fixed-income trading] venues. There are some venues where the workflow is quite different, and we have not got it across all venues yet,” he says.

AxeTrading is also planning to automate new workflows, including repo trading and futures hedging. The company recently added futures into its software for hedging purposes. “We do not have that [automated futures hedging] already. Some of our competitors do, and they have been around for a long time, but we are building that in a very new agile way,” Watters says.   

He says the firm is also extending the range of systems that can be connected to its platform by creating more APIs and plugin points.

For these kinds of trades, the vendor has built what it calls a rule-based “Can I Trade?” API. Depending on the information provided about a specific trade—such as the trade size, customer name, instrument—it will either allow a trade or block it. “That API already exists, but there are customers who have a wider array of those kinds of systems we have to cater to,” Watters says. 

Watters says market abuse capabilities are increasingly becoming important services to offer customers in fixed income markets. “We are now beginning to have discussions with market abuse tools [providers],” he says. “We have formed a partnership with a market abuse analytics system that our customers have told us about, but there are more out there that we need to potentially be ready to integrate with. So that is a new area that we have not yet integrated with, but expect to see the system moving towards.”

Fixed Income Automation Lagging

Watters says fixed income has lagged behind other sectors of the financial market in terms of electronification. “When you look at say the equity markets and stock markets, they moved off stock exchange floors 30 years ago and became electronic back then. Foreign exchange markets have followed. Bond markets are still very much a people business, but it is also quite a work-intensive business in [that] the amount of information that flows between the participants is quite complex,” he says.

Watters sees a macro-trend towards electronification in the fixed income space, driven by increasing regulations in the last few years. He says when the firm first started in 2009, its focus had been on less developed and emerging markets which had been very slow to adopt electronification for fixed income. “We also have customers in more developed markets and bigger banks who have more complex needs,” he says.

He says users of AxeTrading’s software can switch off and on the degree of automation. Usually for very large or complex trades a customer might prefer to do the process more manually, while for smaller size trades they would prefer automation. “That degree of separation and triage of trading is very much what is going on right now with our customers,” he says.