BMLL, the market data and analytics provider, has added historical prediction market data from Kalshi to its global market data platform, giving quantitative and systematic investment firms a standardised dataset for incorporating event probabilities into macro research, backtesting and trading models.
Under the partnership announced today, Kalshi’s historical order book will be normalised into the same schema BMLL uses for CME Event Contracts, allowing researchers to analyse the two datasets within a common framework. The data will be available through Snowflake, SFTP and BMLL Data Lab.
The agreement addresses a practical problem that has accompanied growing institutional interest in prediction markets. Although event contracts potentially offer a direct measure of expectations around economic, political and other real-world outcomes, the underlying data has typically been spread across different venues and APIs, leaving quantitative teams to build their own processes for sourcing, cleaning and standardising it.For BMLL, whose core business is harmonising historical Level 3, 2 and 1 order book data across traditional markets, prediction markets therefore present a familiar data engineering problem in a relatively new asset class. Its recent partnerships have increasingly focused on making large historical datasets easier to consume within quantitative and AI-driven workflows, including collaborations with SIGMA AI for benchmarking live market flow against historical order book data and Tradefeedr for equities and futures analytics.
“Our systematic hedge fund and quantitative clients have shown urgent and active demand for high-fidelity, historical prediction market data to support macro-level research,” comments Paul Humphrey, CEO of BMLL. “By adding Kalshi to our coverage and normalising its historical dataset to match the CME Event Contracts schema, we are removing the burden of data engineering.”
From event probabilities to market signals
The institutional data case for prediction markets is developing alongside the market for the contracts themselves. Kalshi contracts trade between one cent and 99 cents, providing market-implied probabilities for outcomes ranging from Federal Reserve decisions and economic releases to policy and other events. BMLL says researchers can use the historical data to backtest models around events including interest-rate decisions, CPI releases and GDP figures, and investigate relationships between event probabilities and movements across conventional asset classes.
That fits with a broader effort by Kalshi to move its data and trading infrastructure into the systems already used by institutional market participants. Earlier this year, Tradeweb partnered with Kalshi to distribute its event probabilities and market data through institutional workflows, while Trading Technologies subsequently announced connectivity to the exchange for execution and post-trade clearing.
The infrastructure around Kalshi has continued to expand. Earlier this month, the exchange partnered with FCA-regulated Otala.Markets to make exposure to selected event contracts available to European investors through conventional listed securities, initially focusing on Federal Reserve rate decisions. Kalshi is a CFTC-designated contract market and has also expanded beyond event contracts: in May, the CFTC approved its Bitcoin perpetual futures contract.
Andy Ross, Head of Institutional at Kalshi, says the BMLL agreement should make it easier for firms to examine the relationship between prediction markets and traditional financial instruments. “Institutional participants increasingly need better ways to price and manage event-driven risk directly, rather than relying solely on proxy assets,” he notes. “By bringing Kalshi’s historical market data into BMLL’s normalized research environment, firms can compare those signals, test strategies and incorporate event probabilities directly into their macro research and risk-management workflows.”
Building the data layer
The BMLL partnership adds another piece to an institutional ecosystem that is beginning to resemble that surrounding established asset classes. Execution connectivity, clearing, market making and distribution are increasingly being joined by the historical datasets needed for quantitative research and model development.
There are still questions around liquidity, pricing and how extensively institutions will use event contracts either as tradable instruments or data signals. Previous TTI reporting has found considerable institutional interest but a market that remains relatively early in its development. However, researchers don’t need to trade prediction contracts to investigate whether their prices contain useful information. Putting Kalshi alongside conventional order book datasets allows firms to test that proposition systematically, including whether event probabilities can improve macro models, identify cross-asset relationships or provide signals around specific economic releases.
For prediction markets, that creates a second route into institutional finance alongside direct trading. As their data becomes available through the same research environments and standardised schemas as established markets, event probabilities can increasingly be treated as another dataset for quantitative researchers to test rather than a separate and specialised information source.
Subscribe to our newsletter
