BMLL puts Kalshi order-book history into institutional research systems

BMLL Technologies and Kalshi announced a partnership on Sept. 17 that brings Kalshi’s historical prediction-market order-book data onto BMLL’s institutional data platform. The arrangement is aimed at quantitative researchers, macro investors and systematic hedge funds seeking to use event-contract prices in research and event-driven trading workflows.

BMLL will normalise Kalshi’s historical order-book records into the same unified schema it uses for CME Event Contracts. That common format is intended to let a single model work across the two datasets, rather than requiring firms to build separate parsers, field definitions and data-quality processes for each venue.

The standardised historical data will be available through Snowflake, SFTP and the BMLL Data Lab. BMLL said the delivery channels allow researchers to bypass complex API parsing and use the data in historical testing, market-microstructure research and model-development workflows.

The company said quant teams have often had to gather fragmented prediction-market data from disparate APIs, a process that can consume substantial engineering time. Paul Humphrey, BMLL’s chief executive, said demand for high-fidelity historical prediction-market data had been met slowly and expensively because sourcing and normalising the datasets was resource-heavy.

The new feed is designed for researchers to backtest and calibrate models around Federal Reserve interest-rate decisions, CPI releases and GDP prints. BMLL said users could compare event-market signals with rates, currencies and equities, develop prediction indices and forward curves, and assess event-driven regulatory risks across active portfolios.

Kalshi is a CFTC-regulated designated contract market, according to BMLL. Its contracts trade from 1 cent to 99 cents, with prices presented as market-implied probabilities. BMLL argues that these prices provide signals based on financially committed capital, including for policy announcements and macroeconomic releases.

There is a practical limitation to that claim. FinanceFeeds noted that standardisation resolves a data-engineering problem, not liquidity constraints: thin order books can have wide bid-offer spreads, and a midpoint does not necessarily represent a probability that an institution could execute at scale.

The deal focuses on historical research data rather than live trading or execution. Finance Magnates reported that Kalshi’s earlier institutional integrations this year included Tradeweb’s June addition of real-time event probabilities for US institutional clients and Talos’s subsequent trading access for market makers and hedge funds using its digital-assets system.

Kalshi has separately widened the distribution of its live data. As covered in August, it began streaming live order books through DoubleZero Edge, initially covering its most actively traded sports event contracts and crypto perpetual futures.

BMLL said the shared schema will also help firms evaluate newer prediction-market products, including multivariate events and perpetual futures, as they incorporate event probabilities into macro research and risk-management processes.