Trumid Enhances Predictive Pricing Tool for 22,000 Corporate Bonds

Trumid has announced an upgrade to its Fair Value Model Price (FVMP), a predictive pricing tool designed to deliver real-time two-sided bond prices every 30 seconds for approximately 22,000 U.S. dollar-denominated corporate bonds. The announcement marks a significant step in the company’s broader strategy to embed data-driven automation across the fixed income trading workflow.

The enhanced FVMP draws on data from Trumid’s own trading protocols—including RFQ, Attributed Trading, Portfolio Trading, and Swarms—as well as external inputs such as TRACE prints. The model applies machine learning and statistical methods to produce price predictions that are responsive to current market conditions.

Pre-trade analysis, trade automation, market-making, and portfolio pricing

Jason Quinn, Chief Product Officer and Global Head of Sales at Trumid, commented, “FVMP is a critical part of Trumid’s evolving automation story and a testament to our ongoing investment in data and data-driven solutions. With platform activity surging across Trumid RFQ, Attributed Trading, Portfolio Trading, and Swarms, we are producing more unique – and more valuable – platform intelligence and insights to fuel our differentiated data profile, further enhancing the quality, and predictive accuracy of FVMP.”

The model serves as a key component of Trumid’s integrated trading ecosystem. It supports pre-trade analysis, trade automation, market-making, and portfolio pricing for investment grade, high yield, and emerging market bonds. Its updates every 30 seconds are intended to provide traders with current estimates regardless of liquidity or market volatility.

Mutisya Ndunda, Head of Data Strategy and AI, commented, “FVMP’s core competency is its blend of sophisticated machine-learning algorithms and statistical quantitative methods to interpret vast datasets and provide real-time, predictive pricing in any market environment. And, as the model continually adapts to new data, that information is weighted dynamically so that FVMP is designed to maintain its accuracy without relying explicitly on historical predictions.”

Algorithms aim to reduce pricing outliers and adapt quickly to shifting market dynamics

The model’s inputs include both unique trade intelligence from Trumid’s proprietary protocols and publicly available or contributed market data. Its algorithms aim to reduce pricing outliers and adapt quickly to shifting market dynamics, offering predictive accuracy that supports both liquid and illiquid bonds.

The FVMP upgrade builds on other recent developments in Trumid’s data-led automation suite. The company’s AutoPilot tool for RFQ has seen adoption growth, with 81% of eligible line items executing “no touch” in February 2025. Trumid’s PT Pricer has also seen traction by helping users evaluate whether portfolio trades offer better value compared to single-name execution.

Over the past four years, Trumid has expanded its Data and Intelligence and Automation teams, hiring across artificial intelligence, analytics, software engineering, and trade automation. Most recently, the firm named Ryan Gwin to oversee the delivery of custom data solutions for clients.

Trumid’s core platform focuses on U.S. dollar-denominated corporate credit, including investment grade, high yield, distressed, and emerging market bonds. Its technology-first approach integrates market expertise with design-oriented product development, delivering trading protocols through a single unified interface.