By Chris Rice, Senior Managing Director, Global Head of Trading, State Street Global Advisors
New data mining and architecting techniques are helping to bridge the transparency gaps that once challenged fixed income traders.
Best execution in fixed income markets has always been more challenging than in equity markets in which traders have a greater line of sight into securities trading at higher volumes across automated markets. By contrast, fixed income mostly trades over the counter and is characterised by fragmented price sources and less accessible data. However, advances in data accessibility and processing through machine learning and other artificial intelligence technologies are transforming fixed income traders’ ability to build a new generation of smart order routing platforms. These new dashboards are bridging many of the price discovery and pre-trading liquidity gaps that once challenged fixed income traders.
The advent of new fixed income trading platforms has created more information than ever before about transactions and counterparties. Fixed income market participants are increasingly trying to harness the power of these expanded data sources to gain greater insights into market liquidity and enhance trading decisions.
As a result of new infrastructure improvements, dealers now provide liquidity indications of interest (IOI) to the buy-side much more frequently throughout the trading day than before. Even just a few years ago there was often just one IOI in the morning. This provides buy-side traders with a far better and timelier snapshot of actual dealer liquidity.
This improved IOI data, combined with information from other bond pricing sources and platforms, as well as liquidity cost scores, furnishes enhanced insight and the potential to negotiate a better price. Buy-side firms are leveraging this information and in varying forms have been able to consolidate much of this data within their order management systems to make access easier for traders to digest and exploit. This data is also helping fixed income portfolio managers build more resilient strategies that take liquidity conditions into account at the design rather than the execution stage.
There have also been other more formalised initiatives to improve fixed income liquidity and market transparency. For instance, Project Neptune, a collaboration of large banks and buy-side firms, has created a powerful centralised source of information around available fixed income securities. While it is not a trading platform, it nevertheless helps solve some of the opacity and liquidity challenges in the broader industry context of lower inventory and poor information transparency.
The explosion in pre-trade and post-trade data will increase even further with the implementation of the European markets in financial instruments directive (MiFID) II in January 2018. Managers with the technology and expertise to mine and architect this new data into actionable insights will enjoy a competitive advantage.
At SSGA, we are creating the next generation of data processing and visualization techniques to improve our ability to harness new data sources and incorporate those insights into dynamic dashboards that help our traders and portfolio managers achieve better execution.
Better data visualization and dashboards
New data visualization tools have improved the collaboration between trading desks and portfolio managers. The ability to dynamically filter and interpret large data sets in differing visual formats has created an important feedback loop for trader and portfolio management teams aimed at constant process improvement. The dashboards enable more precise data customisation for targeted audiences (for example, traders versus portfolio managers versus group CIOs) and allow end users to detect anomalies in trading costs, volumes or counterparty activities.
During the past year, we have rolled out dynamic dashboards to traders that enable them to hone in on their trading activity over a given time period. Counterparties can be analysed by volumes traded or associated transaction costs. Traders can review the performance of counterparty over time or narrow down performance in certain markets or trading styles. As a predictive feature, this may flag counterparty strengths or weaknesses as a guide for future trading activity. This is particularly important information for those fixed income sectors where liquidity is scarce.
March of the machines
As advances in machine learning and other artificial intelligence (AI) technologies are applied more broadly to automating and refining pre- and post-trade fixed income data, we expect that both traders and portfolio managers will benefit. We are quickly moving to the state where we will be able to automatically capture datasets from across the entire lifecycle of fixed income trading activity – from cash flows to execution. This expanded universe of data (RFQs, liquidity, market data, execution data, qualitative commentary, historic cash flows, etc.) will help to incorporate trade execution directly into the portfolio management process and improve speed to market. We envisage a not-too-distant future in which our traders will routinely be using AI technology to improve pre-trade price discovery and optimize execution.
This iterative process of improving the volume, quality and representation of pre- and post-trade fixed income data has several important benefits for investors. Most immediately it helps fixed income traders augment the inherent price discovery challenges of trading fixed income securities in fragmented and less transparent venues, thus improving their ability to provide best execution. More importantly it provides a powerful advantage to fixed income portfolio managers by equipping them with far more reliable liquidity data across fixed income sectors as they design their strategies to minimize market frictions and maximize efficient execution.
We’d love to hear your feedback on this article. Please click here
