Fixed income analytics firm Overbond has launched COBI-Pricing Live, a new real-time AI bond pricing product, which the firm says has the capacity to process “real-time historical pricing data” on coverage of over 30,000 securities with a refresh rate of under three seconds.
Overbond says the new product fills a gap in the current market by aggregating multiple data sources on the client side across trading venues, data aggregators themselves and fundamental and settlement layer data. It notes that bond markets suffer from illiquidity and non-uniform data sources used to price bonds and build credit curves, and fixed income traders require accurate bond pricing that can measure the liquidity of individual securities and enable automatic execution.
COBI-Pricing Live provides this requirement, the company says, by leveraging the power of AI models that optimise prices for bonds with various liquidity profiles and performing historical benchmarking and curve fitting. In phase one it prices curves for a specified list of companies with the highest liquidity profile and in phase two it uses a “nearest neighbours” algorithm to find similar bonds and companies to generate pricing curves for illiquid issuers. The third phase is the creation of individual ISINs pricing and curves for all illiquid issuers using support vector regression models.
“Both sell side dealers and buy side asset managers are increasingly relying on AI applications to price fixed income securities in live trading and automate part of their daily workflows,” says Vuk Magdelinic, CEO of Overbond. “However, most of the existing fixed income capital market data sources do not have enough coverage to provide traders with a view of true liquidity and price precision that can be executed automatically.”
The bond market, notes Magdelinic, heavily relies on segregated data disseminated between counterparties, which creates fragmented data sets that do not cover the bulk of over-the-counter traded flows. COBI-Pricing Live collates and organises large volumes of disparate data, including non-traditional data sets such as fundamental and settlement layer data. Using novel AI liquidity scoring, it tiers all trades and determines if these qualify for full-automation, trader supervision, or should not be traded at the current time.