FCNR(B) Inflows Won’t Trigger Excess Lending, SBI Chairman Says Amid ₹10.5 Trillion Liquidity Surge

State Bank of India Chairman CS Setty revealed that the significant foreign currency non-resident (bank), or FCNR(B), deposits collected by banks will be absorbed within the next few months and are not expected to trigger excessive lending. These concerns emerged after banks raised more than anticipated funds through the Reserve Bank of India's concessional swap facility, which led to worries about increased liquidity accelerating lending.

Setty stated that it would take about three to four months for the liquidity to be deployed. The bank had initially targeted FCNR(B) deposits of around $10 billion. FCNR(B) inflows have been pushing banking liquidity higher, with banks mobilizing $127.2 billion through these deposits by August 31, bringing total inflows to $136.4 billion.

The FCNR(B) window closed on August 31 after receiving strong demand. The Reserve Bank of India had launched the facility on June 8, accepting deposits, foreign currency bonds, and external commercial borrowings. The higher-than-expected inflows pushed the banking system's liquidity to nearly ₹10.5 trillion. Banks now have various options to utilize these funds, such as investing in government securities, replacing costly deposits, or increasing loans.

However, experts believe that robust corporate credit growth might limit the extent of fresh lending. Setty also discussed the potential of agentic artificial intelligence (AI) in the banking sector, stating that adoption would depend on lowering initial costs and developing scalable technology. He mentioned that while the initial fixed cost of agentic AI can be high, the incremental cost can be low, leading to the economics of scale.

Agentic AI could assist banks in various areas like fraud detection, KYC, anti-money laundering checks, loan assessment, and reconciliation. Setty believes that India's financial sector's next stage would be a transition from digital banking to "intelligent banking." The next frontier for Indian AI, he said, is not just about building larger models but creating models that understand India's diversity and uniqueness.