BNP Paribas Sees Post Trade at 97-99% STP, With AI Fixing the Rest

BNP Paribas' Securities Services business runs at roughly 97 to 99 percent straight-through processing (STP, trades settling without manual touch) in routine flows — yet the residual still amounts to thousands of transactions daily, and that gap is where artificial intelligence now aims. The figure came from Gary O'Brien, Sydney-based Head of Financial Intermediaries & Corporates APAC, speaking on Traders Magazine's Open Order Podcast alongside London-based Financial Intermediaries Segment and Regional Strategist Sonal Meghani.

The two surveyed the post trade landscape of 2026: clearing, settlement, custody and reconciliation, long the laggards of the trade life cycle, now facing compressed timetables, surging volumes, expanding regulation and AI-driven change.

What does settlement compression change for desks?

Meghani characterised the new era as "faster, more connected, and more demanding." Settlement cycles keep shortening: India moved to T+1 in 2023, the United States followed in 2024, and the United Kingdom, European Union and Switzerland are preparing the same, with Brazil after that, Meghani said. Australia pushes T+1 toward 2029 or 2030 because re-platforming the CHESS clearing system comes first.

O'Brien framed the operational consequence: T+1 mandates require more automation and transmission of key data points, while leaving less time to fix problems. He drew a hard line between cycles: a market can move to T+1 on a T+2 framework, but T+0 would require "radical and innovative change," including deeper alignment of trading hours, funding, FX and liquidity management.

Asia offers partial counterexamples. China and Hong Kong's Stock Connect settle partly on T+0, as does Australia for some discount securities. China is a no-fails market, O'Brien said, but one needing more manual intervention — it does not work at scale. Digital asset frameworks, built for near-instant settlement from the outset, may find T+0 easier than legacy markets.

Is data the bottleneck or the differentiator?

Both. Meghani put the stakes in settlement terms: under T+1 there is no slack to fix a data problem after the fact. "You could automate your solutions as much as you like," she said, but bad data still means failed trades. The differentiator is not owning more data but having "reliable, governed and well-structured data that can also move across the ecosystem."

O'Brien cited "shareable technology" — clients accessing a provider's information directly — cutting reconciliation between two data stores and the overhead before an investor can act.

Where does AI fit in the exceptions queue?

Meghani distinguished generations of AI: the older kind helped firms work faster; the emerging agentic era helps them work smarter, with specialised agents orchestrating workflows under human oversight. Adoption, she said, must be "responsible and explainable," backed by strong governance, because trust is critical in post trade.

O'Brien grounded the use case in the STP residual: AI can work out why a trade fell out of automation and propose, or even automate, the fix, turning exceptions into "happy path" trades that settle on time. He also flagged the downside: "It's not just the nice people that get to use these technologies," so cybersecurity and risk frameworks must keep pace.

Regulation: constraint or catalyst?

For years, O'Brien noted, rules were seen as stifling innovation by absorbing budgets that might have funded it. That reading has flipped — mandates such as T+1 now force automation and data discipline. Meghani likened the moment to maintaining a home: "Over a certain time you have to re-evaluate the pipes and the plumbing," work done in collaboration with regulators, infrastructures and clients. O'Brien's strategic advice was blunt: "The days where you can be a unicorn that is providing all services to all people are really gone." He described post trade as a period of "complex collaborative competition," where a firm may compete with a peer in one market and partner with it in another.

Do people still decide the winners?

Both executives said yes. Meghani called post trade a people business, requiring data literacy and commercial awareness on top of deep operational knowledge, in a collaborative, cross-functional culture. The function spans operations, product, technology, risk and legal, and that breadth draws talent. O'Brien added that showcasing what an organisation can become helps attract and retain top people — and that back-office advancement ultimately lets the front office do more.

In closing, O'Brien urged firms to set mid- and long-term visions and build short-term priorities toward them. Meghani said the mega trends are an opportunity for firms to focus on what they uniquely bring to the marketplace.