Banks May Take 3-4 Months to Deploy FCNR(B) Inflows, Says SBI Chairman C S Setty

Indian banks could take three to four months to deploy the large pool of funds raised through foreign currency non-resident (bank), or FCNR(B), deposits, according to State Bank of India Chairman C S Setty. 

Invest in Indian Markets and Unlock Future Potential With 5paisa!

Open Account Now

Speaking at the Global Fintech Fest 2026 in Mumbai on Thursday, Setty indicated that the additional liquidity is likely to be deployed gradually and is unlikely to result in “abnormal lending”. 

His comments come after banks raised substantially more through FCNR(B) deposits than the market had initially anticipated, pushing liquidity in the banking system sharply higher. 

FCNR(B) Inflows Reach $127.2 Billion 

Inflows through the Reserve Bank of India’s concessional swap facility reached $136.4 billion by August 31. 

Of this, banks mobilised $127.2 billion through FCNR(B) deposits, considerably higher than market estimates of $90-100 billion. 

The RBI had introduced the concessional swap facility on June 8 for fresh FCNR(B) deposits, overseas foreign currency bonds and external commercial borrowings. 

The FCNR(B) window was initially scheduled to remain open until September 30, 2026. However, the central bank brought the deadline forward by a month following the strong response, and the window closed on August 31. 

Setty did not disclose how much SBI eventually mobilised through the scheme. The country’s largest state-owned lender had set a target of nearly $10 billion. 

Banking System Liquidity Near ₹10.5 Trillion 

The large inflows have added substantially to liquidity in the banking system, which stood at nearly ₹10.5 trillion as of September 9. 

Banks have several avenues available to deploy these funds. Industry experts cited in the report identified three broad options: investing in five-year government bonds, replacing higher-cost deposits, or lending to corporates. 

The third option has raised some concern because credit growth is already strong, particularly in the corporate segment. 

Setty, however, expects deployment to take place over three to four months rather than immediately. 

Agentic AI Could Have a Wider Role in Banking 

Beyond liquidity, Setty also addressed the growing role of agentic artificial intelligence in banking. 

Unlike AI systems designed primarily to assist with individual tasks, agentic AI can act on behalf of customers and adapt its actions as circumstances change. 

Within banks, such systems could potentially be deployed across different stages of the financial lifecycle. The areas identified by Setty included fraud and new-account detection, know-your-customer processes, anti-money laundering checks, loan appraisal, underwriting and reconciliation. 

For customers, the technology could support more proactive and contextual interactions. 

Setty described the next stage as a transition from digital banking towards more intelligent banking, with AI models capable of understanding India’s diversity and specific requirements. 

High Initial Costs Remain a Challenge 

Cost remains an important consideration before agentic AI can be deployed widely across Indian banking. 

Setty said the initial fixed cost can be high even though the incremental cost of operating such systems can subsequently be lower, creating scope for economies of scale. 

He also identified a broader opportunity beyond automation. 

Sophisticated financial advice and wealth-management capabilities are currently concentrated among affluent and financially knowledgeable customers. Agentic AI could potentially extend personalised financial assistance to hundreds of millions of people at different stages of their financial lives. 

That expansion, however, does not mean removing people from the process. 

Human Oversight to Remain Important 

Setty said AI should augment human capabilities rather than replace them. 

Routine, high-volume and data-intensive activities can increasingly be handled by AI, while complex and high-risk financial decisions would continue to require appropriate human oversight. 

Greater automation could also allow bank employees to spend more time on customer relationships, problem-solving and other value-added activities. 

But giving AI systems greater autonomy introduces a different set of risks. An error made by an autonomous system could set off actions across multiple connected systems, potentially affecting customers, counterparties and financial institutions at machine speed. 

Banks May Need to Move From ‘Know Your Customer’ to ‘Know Your Agent’ 

Setty outlined accuracy, accountability and access without asymmetry as three principles required to build trust as AI adoption increases. 

Accuracy would need to remain consistent across different customer groups rather than being judged only by average performance. 

Accountability would require monitoring, traceability and the ability to determine why an important action was taken, particularly as AI systems move from recommending decisions to executing them. 

The rise of AI agents could also add a new dimension to risk management. Banks have traditionally built safeguards around the principle of knowing their customer. As autonomous agents begin participating in financial transactions, institutions may increasingly need mechanisms to identify and understand those agents as well. 

Such systems could involve agent identity and authentication, customer consent, transaction limits, audit trails and location. Banks may also have to operate in an environment where AI systems interact with other AI systems, increasing the need for real-time risk assessment. 

For the banking industry, the immediate challenge is therefore unfolding on two fronts: deploying the unusually large FCNR(B)-linked liquidity in an orderly manner, while preparing for a more autonomous generation of AI systems without weakening human oversight, accountability or financial safeguards.