Agentic AI: Reshaping Markets & Securities Services

Agentic AI is quickly becoming a central capability, transforming how markets and securities services are delivered. KPMG suggests these systems could unlock around $3 trillion annually in corporate productivity gains. However, McKinsey notes that if banks don't adapt, global banking profit pools could shrink by roughly $170 billion over the next decade. At HSBC, daily AI Markets interactions have risen significantly since 2019, from hundreds to tens of thousands. This shows agents are now part of the operating fabric. The main issue for global banks is whether their infrastructure can support AI agents to improve client outcomes, not just add complexity. Many firms plan to deploy AI agents, but few have done so at scale. Clients often still face multiple portals and inconsistent answers. Agentic AI, introduced into fragmented environments, can reinforce these issues. The constraint is architectural; banks have systems optimized for products, but clients experience banks as single entities. HSBC AI Markets, an NLP platform, gives institutional investors instant access to data, analytics, and market insights. It acts as a conversational tool for research, real-time pricing, and trade execution. The key question is what operating environment these agents run on. An AI-era operating system must treat proprietary data as institutional memory, connecting positions, risk, and historical decisions in secure, governed environments. This matters because it directly impacts how efficiently and effectively financial services are delivered to clients.