Finastra: The New Operating Model in Trade Finance

Matthieu Andrieu, Finastra, spoke with FF News at Sibos 2025 about what’s shifting in working capital and trade finance. With two decades in the field, he says regulation, and new comfort with fully digital documents, is the main catalyst.

Andrieu points to fresh regulations in France, Germany, the US, Abu Dhabi and parts of Southeast Asia that grant electronic trade documents the same legal status as paper. That’s moving the industry from manual, paper-based processes to digital workflows, forcing banks to rethink target operating models, controls and skills. Africa and South America are progressing more slowly but are expected to follow as legal frameworks mature.

Customer experience has lagged in trade compared with retail. For years, banks focused on back-office automation, not front-end journeys, and now Andrieu sees growing demand for corporate-friendly channels such as mobile, chatbots and especially generative AI, to simplify processes and make status, exceptions and documents clearer to clients.

From Finastra’s vantage point, over 200 banks use its Trade Innovation platform and two challenges dominate. Staying on top with regulations and best practice is the first challenge, including the shift to ISO 20022 and with the November 2025 SWIFT release, many must move from MT to MX messaging.  The second challenge is adopting an ecosystem mindset so instead of one vendor, banks need a partner network, making open APIs essential for speed and co-innovation.

Cloud is the other through-line. Most Western institutions now default to cloud as a step toward managed services, shifting more run/maintain/upgrade responsibility to providers and changing commercial relationships, freeing banks to focus on differentiation.

Sibos remains Finastra’s launch pad and this year, Andrieu highlights “Nexus,” which is designed to help banks collaborate with partners across AI, machine learning and blockchain on top of Finastra’s corporate banking stack, an orchestration layer for faster experimentation without core rip-and-replace.