Electronification is gathering pace across capital markets, and with it the proliferation of trading venues, from DLT-based platforms to crypto exchanges. But this progress carries a hidden cost: banks risk fines or losses when they miss a change to a venue's rulebook or capabilities.
Over the past six months, Corlytics has met with more than 15 global banks across New York and London. The largest of these institutions now trade across hundreds of venues, some exceeding a thousand, yet none have dedicated tooling to track how those rulebooks evolve.
RegTech firm Corlytics recently talked about the elephant in the trading room nobody is pricing.
In most cases, the responsibility falls to a single junior analyst armed with little more than a spreadsheet.
For business heads, this represents more than a control gap; it is a brake on growth. Firms cannot expand electronic trading, or safely adopt new venues and capabilities, while remaining blind to the rules governing the platforms being added. Ambition, as the firm puts it, meets a wall of invisible rules.
The exposure is tangible and growing. Missing a change to an order type, a market-making obligation or a surveillance requirement carries consequences that are far from hypothetical. Fines and PnL hits can be significant, as can the opportunity cost of electronification projects stalled by rules no one can trust.
Perhaps most concerning is that this remains an unknown unknown. Few institutions have quantified the risk because they have never measured it, and what cannot be seen cannot be priced. The root cause is the absence of any standard: no machine-readable rules, no common language for venue rules and no automated engine.
Corlytics has built a solution to address this gap. The RegTech firm has converted millions of pages of unstructured rulebooks into structured, proprietary data, using an ensemble of models in which only 3% of the work is data extraction, with the remainder devoted to quality assurance to ensure the output can be trusted.
The system rests on models refined over five years, €2.5m of R&D investment, 98% model accuracy and model governance embedded from the outset, all backed by SLAs guaranteeing data delivery.
The result is a network in which venues are normalised and benchmarked, each venue's core capabilities are mapped to a firm's obligations, and every change is surfaced in real time. Rebuilding such capability internally would be slow, costly and duplicative; joining the network mutualises both the work and the cost.
The banks that see every venue clearly will scale electronic trading safely. Those that stay blind are one missed change away from a fine or a missed market.