Since the earliest days of regulated bond markets, buy-side firms have operated with a significant information disadvantage. The problem is structural: every buy-side trading interaction signals their intent and is seized upon by trading counterparties as a valuable datapoint to inform the bids and offers they quote back.

The current market architecture essentially hands bond dealers a wealth of pre-trade intelligence from their clients before they even provide a price - an information asymmetry that the buy-side has long wished to correct. If buy-side entities were equipped with their own pre-trade analytics - the ability to estimate trading counterparty price commitment before showing their hand to the market - the structural information asymmetry that has long defined fixed income trading could be meaningfully reduced.

The buy-side is not lacking the market data needed to model trading counterparty behavior - on the contrary, large asset managers receive millions of price points from market makers each day. The challenge is converting those data points into high-quality, repeatable, decision-supportive metrics that can be used by traders in real time.

ICE Compass is a platform that brings these signals together to provide intraday pre-trade price intelligence by ranking prospective counterparties on their estimated trade competitiveness. The platform helps enable data-driven counterparty selection, providing institutional investors and asset managers the ability to estimate the bid or ask they can expect from specific counterparties, customized for them prior to indicating their trading intent, across global corporate and government bonds.

The challenge in estimating trading outcomes pre-trade is that the data informing transaction costs is unstable, noisy in signal, sparse in coverage, and shifts with the market. Trading counterparty behavior also changes over time, with factors like bond inventory, cost of funding, and risk appetite all influencing how aggressively counterparties compete to win trades.

This makes it difficult for trading analytics models to isolate consistent patterns across different market conditions and time periods. Extracting high-quality trading signals requires an approach that leverages machine learning and that can adapt to this complexity: this is precisely what ICE Compass does.

The depth of ICE's underlying data and data science expertise is what Compass is built upon. Every estimate the platform produces is anchored to ICE Continuous Evaluated Pricing (CEP), since without a high-quality point-in-time reference price, there is no way to measure the competitiveness of trading counterparty quotes.

Around that anchor, the platform layers iNAV and live exchange feeds to track ETF premiums and discounts intraday, a step change in how trading costs can be observed in real time. ICE Compass also layers in measures of volatility to capture shifts in rate volatility, reference data, TRACE prints, and our proprietary liquidity metrics to read market directionality.

The buy-side firm's own request for quote (RFQ) responses, indications of interest (IOIs), and trade history sit alongside all of this in a cloud-hosted data model purpose-built for ICE Compass. It is this architecture that allows the platform to evaluate and estimate counterparty behavior at the speed the trading desk requires, quantifying more than 70 distinct features in its counterparty profile analyses.

For example, each time a trading counterparty responds to an RFQ, the aggressiveness of that price is compared with their IOIs/axes and CEP. If a counterparty’s IOI says they will buy at 100, but the RFQ response comes back at 99.50, that data point is stored and added to their profile. Hit ratios are monitored to see if a counterparty wins a trade or comes in the top three, signaling that they are willing to trade that bond or comparable bonds. Large quantities of datapoints are collected each day to further refine trading counterparty profiles.

This vast quantity of market data feeds ICE's machine-learning-driven pre-trade analytics, producing a ranking of trading counterparties by estimated competitiveness for a specific order, with the estimate customized to the unique characteristics of the buy-side firm. ICE Compass users like T. Rowe Price, the platform’s pre-trade pricing analytics fill in a previously missing element in the counterparty selection equation.

“At T. Rowe Price, we are focused on leveraging data more effectively to support trading decisions and strengthen our investment process. Through our collaboration with ICE on Compass, including our role as an early thought partner, we helped identify real gaps in the pre-trade fixed income workflow,” says Matthew Murphy, Credit Trader at T. Rowe Price.

“That collaboration helped ICE translate the realities of how traders operate into a tool that turns large volumes of market data into actionable, pre-trade insight. By giving traders a more transparent view of counterparty behavior before disclosing trading intent, Compass can support better counterparty selection, faster and more disciplined execution decisions, and a more data-driven operating model across the desk,” Murphy adds.

Since ICE Compass continuously profiles data behavior across the day, the platform has already made a wealth of unexpected observations into fixed income market dynamics that shed new light on the mechanics of bond market pricing.

One recurring insight is that trading counterparty competitiveness can vary meaningfully over the course of a trading session. ICE Compass can incorporate observed pricing behavior into its intraday assessment, helping traders understand which counterparties may be more or less competitive for similar opportunities as market conditions evolve.

The platform has also highlighted that counterparty behavior is not always uniform across channels, coverage models, or market environments. By bringing these signals together, ICE Compass helps traders form a more complete view of where competitive liquidity may be available before engaging the market.

In an environment where fee compression is pressuring investment manager operating margins, ICE Compass’ ability to estimate which counterparty is likely to provide the most competitive price on a trade before going to market has the potential to meaningfully change how the buy-side approaches pre-trade counterparty selection.