Investortools integrates SOLVE's price predictor

Software solutions firm Investortools is integrating SOLVE's AI-powered price predictor into the former's dealer network and portfolio management system.

This integration adds to the Investortools platform, "combining AI-based predictive pricing with pre-trade quote data," while for SOLVE, "the collaboration extends the reach of its data and analytics by embedding them directly into a platform fixed income professionals use every day," according to a press release.

With SOLVE Px now integrated into the Investortools environment, "traders and portfolio managers can access indicative institutional prices tailored to a $1 million trade size as well as size- and side-specific quotes," the release said.

"Our clients and their clients are saying, 'We would like this predictive price integrated into the workflow to make my life easier,'" said Gregg Bienstock, group head for municipal markets at SOLVE. "That's the driving force behind this partnership."

The deal is the latest in a series of technical integrations for Investortools' Dealer Network platform, which the company describes as a "centralized, electronic trading cockpit" as market participants increasingly seek consolidation and integration of information and systems.

Investortools currently connects with over 50 outside firms, and the Dealer Network consolidates the secondary market across 15 dealers and four ATS platforms, according to Investortools Senior Vice President and Head of Sales James Morris.

SOLVE Px uses AI to create predictive pricing for municipal and, as of last week, corporate bonds. Among the inputs is data from SOLVE's quotes database, which is aggregated from chatter between market participants.

Bienstock said that data is what sets their system apart from other tools. 

After comparing their predictions to actual trades with and without the quote data, Bienstock said the model is 34% more accurate with that information. Overall, SOLVE says its system has a median margin of error of 5.7 basis points.

"Everyone has access to trade data, everyone has security master information, and everyone has smart people," he said. "But what others don't have access to is the quotes."

Along with the predicted prices, SOLVE Px provides a confidence score of 1-10, describing its faith in each prediction, Bienstock said.

"As more firms lean on algorithms for assessment and execution, having a reliable price they can trust is no longer optional," Morris said in a press release. "Partnering with SOLVE brings that capability to our platform in a meaningful, practical way."