An AI Revolution Is About To Transform Corporate Bond Trading

Originally Featured in The TRADE by Jim Kwiatkowski, CEO, LTX, on February 19, 2025.

Across Wall Street and around the world, traders and portfolio managers are rushing to reposition corporate bond portfolios in the wake of recent rate cuts from the US Federal Reserve and reduced expectations for additional easing in 2025 as inflation lingers. The trades they are making today will play an important role in determining how their portfolios perform in a new geopolitical and interest rate environment. These trades could end up being remembered as important for another reason as well: For many of these market participants, this may be the last time they make major changes to their portfolios using the manual, time-consuming and inefficient methods that have defined the corporate bond market for decades.

The reason for this is that new artificial intelligence solutions available today can transform the entire process of assessing liquidity, identifying bonds, selecting counterparties, trading bonds, and post-trade reporting. These new efficiencies have the potential to eliminate hours of work and improve results for individual traders, while creating new efficiencies that could facilitate more market-wide trading volumes, liquidity, and transparency, reducing spreads and lowering costs for investors.

The Corporate Bond Market Today

To understand the potential impact of these next-generation tools, it’s important to look at how market participants trade corporate bonds today. The corporate bond market is a heterogeneous mix of instruments, each with its own tenure, coupon and other characteristics. Traders and investors must navigate a maze of data streams and analytic tools designed to help research and analyze the expansive universe of bonds and their own portfolios. Many bonds trade infrequently, if at all. According to our analysis, in December 2024, 75% of total US credit trading was concentrated in only 16% of bonds¹ . The sheer size and diversity of this market makes it challenging to access. As a result, less than half of investment grade bonds and only a third of high yield bonds are traded electronically.

By contrast, in markets like foreign exchange, upwards of two-thirds of total trading volume is now executed on electronic platforms. For US Treasuries, that share approaches 80%. What those and other markets have in common is that they are composed of a relatively small number of similar products that trade regularly. However, corporate bonds have not proven as amenable to electronic trading as most other asset classes. At least not yet.

Market participants frequently advocate for continued electronification because it improves market efficiency by providing transparency, increasing liquidity, and reducing costs. Electronification evolves market structure by enabling new types of trading and supporting higher trading volumes for the benefit of the entire market, and our belief is that AI can help to accelerate this process more quickly.

The Data Science Revolution

One of the biggest frustrations for corporate bond market participants today is the fact that they possess a huge and rapidly expanding volume of data, but it is nearly impossible to gather and process all that data in time to help inform a trading decision. That is a problem almost perfectly designed for artificial intelligence. AI-powered applications can quickly analyze the entire universe of available data to create a robust profile for every bond in a portfolio, and for every bond in the market. Instead of manually scouring market data terminals for pricing and other bond data, a trader, investor or dealer can use natural language to simply ask the application about characteristic-based bond discovery, relative value, price and volume history, counterparties, available liquidity, and reporting on trading history. In a matter of seconds, market participants can receive responses to complex bond-related questions, incorporating data from many different sources, that would have taken hours to obtain manually.

The market is seeing innovations every day, with new platforms employing data science capabilities like machine learning and large language models to aggregate, process and analyze data. For instance, on the LTX trading platform, our BondGPT tool allows users to ask questions and identify corporate bonds based upon the user’s criteria and much more. These capabilities can help answer critical questions like: “What are the most liquid bonds that meet my investment criteria, at what price should I be transacting to get a particular size executed, and what counterparties are likely to be willing to trade on those terms?”

Saving Time and Making Money with AI

The benefits to market participants are obvious. As they execute trades, they will save time, while ensuring that their trading decisions are driven by the best and most timely data available.

The application of AI to corporate bond trading will also deliver other benefits that might take longer to arrive but could prove even more important in years to come. AI-powered platforms will unlock market liquidity by expanding the universe of bonds that participants can analyze and consider when making trading decisions, and by revealing new potential counterparties in the position to buy or sell bonds. Over time, these enhancements will lead to a boom in secondary trading volume. By making it easier and less time-consuming to identify, analyze and price corporate bonds, AI-driven applications could create positive market structure change, enhancing overall market liquidity and lowering costs for all participants.

It’s difficult to predict when these benefits will materialize at a market-structure level, but our buy-side clients continue to show eagerness in embracing these changes. In the coming weeks and months of 2025, we look forward to sharing more about the innovations we’re working on at LTX to help. We believe that rate of adoption for these innovative tools will continue to accelerate quickly as portfolio managers and traders using traditional methods to analyze and trade corporate bonds experience the transformative potential of incorporating artificial intelligence into their workflows.

Read the article in The TRADE.

¹ Analysis based on TRACE volume by CUSIP vs. total traded volume in December 2024.