BMLL and Exponential Technology have launched a data product that tries to answer a question equity traders have long chased: when a stock moves, is it institutions or retail investors doing the buying? The two firms said today (Thursday) that the tool, XTech US Equity Flow, breaks down US order book activity into net buying and selling by investor type.Sorting the Institutions From the Retail CrowdThe pitch centers on a familiar headache for discretionary money managers. When a stock jumps, they want to know whether a large player is simply pushing the price as it trades, or whether the market is repricing on fresh information.Exponential says its layer can tell the two apart by reading full-depth order book data across US venues, blending BMLL's historical records with delayed feeds and its own inference methods. The dataset runs from January 2020 to the present and can be delivered in one-minute, hourly, daily or weekly intervals, the firm said.This is not the first time BMLL data has been used to separate retail from the rest. Its own Trades Plus dataset, launched last September, already flags retail and block trades. XTech goes further by labeling continuous flow rather than individual prints, and it adds dashboards and what the companies call agentic research workflows, aimed at asset managers who are not quants.Big Accuracy Claims From a Two-Year-Old FirmExponential's sales case rests on some striking numbers, all of them self-reported. Morgan Slade, the firm's chief executive and head of research, said "institutional flow...explains roughly 80% of open-to-close price moves in US equities" and points to where prices head next.The company, founded in March 2024 and based in Chicago, also sells macro forecasting. It says its inflation forecasts publish about 20 days before the official reading and call the direction correctly 81.9% of the time, against a 75.5% market consensus. The premise that retail flow can be spotted and traded around has been building for years, a shift FinanceMagnates.com has tracked as brokers rethink the old assumption that retail traders always lose. Slade said the work draws on more than 25 years of running high-frequency and systematic strategies.Another Partner Product Stacked on BMLL's DataFor BMLL, the deal fits a pattern it has kept up since Nordic Capital bought the firm in October 2025. Rather than build every dataset in house, it rents its Level 3 order book records to partners who wrap products around them.The company teamed with Features Analytics in February on surveillance benchmarking tools built from its data. In March it opened a year-long pilot with Tradefeedr to extend transaction cost analysis from foreign exchange into equities and futures.The closest cousin to XTech came in April, when BMLL added SpiderRock's US options data and a joint paper on estimating dealer gamma positioning. Both efforts try to infer who is behind a move rather than just report the move itself.BMLL sits in a crowded niche for granular historical data, competing with specialists such as Databento, which sells nanosecond-precision feeds through cloud APIs and built its own Databricks integration in 2024, and Kaiko, which pushes similar distribution deals for digital-asset order book records. The Read on the LaunchXTech US Equity Flow is the first product from the partnership, and the companies said more will follow. BMLL's chief executive, Paul Humphrey, framed the tie-up around a move away from surface-level data, saying "the richest, most actionable insights are no longer found in top-of-book data."The launch lands during a busy stretch for BMLL, which has been hiring and stacking distribution deals since minority shareholder Optiver first took a stake in October 2024. For now, the product's central promise, that it can reliably tell institutional buying from retail, rests on Exponential's own testing, and the firm says its estimates will be revised as new data comes in.BMLL, the London-based provider of historical order book data, supplies the underlying feed. Chicago's Exponential applies its own methods to sort the flow into institutional, market-maker and retail activity, with advanced tiers splitting it further into high-frequency traders, funds and retail on a per-exchange and per-ticker basis, according to the companies. This article was written by Damian Chmiel at www.financemagnates.com.

Read full article on FINANCEMAGNETS