FCNRB deposits: Why bank investors should watch RoE, not just NIM

Foreign-currency non-resident (FCNR(B)) deposits could prove beneficial for banks’ return on equity (RoE) despite being mildly dilutive to net interest margins (NIM), according to a report by Anand Rathi Share and Stock Brokers.

The report, titled “The dollar-deposit opportunity: Why bank investors should watch RoE, not just NIM”, said the market’s focus on potential NIM dilution overlooks the balance-sheet, capital-efficiency and duration benefits of the scheme.

The FCNR(B) deposit scheme was operationalised on June 8, 2026, with the forex-hedging cost borne by the Reserve Bank of India (RBI). It has received an impressive response, with $127.2 billion mobilised through FCNR(B). Including overseas foreign currency borrowings (OFCBs) of $5.3 billion and external commercial borrowings (ECBs) of $3.9 billion, the total stands at $136.4 billion. FCNR(B) flows account for around 4.5% of system deposits.


“While the street reads incremental FCNR(B) as NIM-dilutive, we view it as RoE-accretive despite being mildly NIM-dilutive,” said Yuvraj Choudhary and Subhanshi Rathi of Anand Rathi Share and Stock Brokers Limited, the authors of the report.

The brokerage analysed four scenarios and found the NIM impact to be negligible. “As FCNR(B)’s cost premium applies to a small slice of deposit base, the blended drag barely moves,” it said.

According to the analysts, exemption from CRR and SLR, along with the treatment of eligible advances under the PSL framework, allows banks to deploy the funds into their highest-yielding non-PSL lending book, while leverage against pledged deposits can create a near-zero-RWA book. The locked 3-5-year tenor also supports asset-liability management when assets reprice faster than funding in a rising-rate cycle. RBI has separately allowed advances against eligible fresh FCNR(B) deposits to be excluded from the adjusted net bank credit calculation used for PSL targets.

“Reserve exemption, capital efficiency of leverage and duration drive RoE accretion even when the headline NIM softens marginally,” said the analysts.

Why the NIM concern may be overstated

FCNR(B) carries an all-in cost slightly above the domestic funding mix, which forms the basis of the market’s NIM-dilutive view. However, the higher cost applies to only a small portion of the overall deposit base, limiting the impact on blended spreads.

FCNR(B) deposits under the specified scheme are exempt from CRR and SLR requirements, allowing the entire amount to be deployed into the bank’s highest-yielding non-PSL book.

The small headline cost premium, analysts said, is more than offset by the larger deployable book.

Why RoE is the key focus

The economics become more attractive when FCNR(B) is offered as a leveraged deposit.

Under this structure, NRIs pledge their deposits and banks lend against them, creating a larger foreign-currency loan book. Based on discussions with banks, Anand Rathi said these loans against pledged or overdraft deposits can scale with negligible incremental risk-weighted assets (RWA).

The lending spread is thin, but it is earned on near-zero-RWA assets that consume almost no capital. Several banks also charge fees on these loans.

“The result is strong RoE accretion with minimal CET-1 draw, which is why NIM is the wrong lens and RoE is the right one,” the report said.

ALSO READ: Small investors, big money! 15 retail-heavy subscribed IPOs of 2026 deliver up to 76% returns

Scenario 1: What if all incremental deposits are raised through FCNR(B)?

In the first scenario, Anand Rathi assumes that the entire incremental deposit accretion is raised through FCNR(B), rather than through CASA or domestic term deposits.

FCNR(B), once fully hedged, has a higher all-in rupee cost than the comparable domestic funding mix of CASA and domestic term deposits.

“At the extreme, if the entire deposit base were replaced with FCNR(B), the full ~110bps spread dilution would flow through to the bank's spreads. This is clearly unrealistic,” the report said.

Cumulative FCNR(B) deposits mobilised under the RBI window stood at around 4.5% of system deposits. At this level, the approximately 110bps spread dilution would translate into only around a 5bps impact on blended spreads/NIM, calculated as 4.5% × ~110bps, assuming the remaining 95.5% of deposits continue to be funded through the existing domestic mix.

Scenario 2: What if all incremental TD growth is raised through FCNR(B)?

The second scenario assumes that incremental domestic term-deposit growth is replaced by FCNR(B), while the CASA mix remains unchanged.

Because FCNR(B) deposits are exempt from CRR and SLR requirements, while eligible advances against such deposits receive specified relief under the PSL framework, the deployable book on every Rs100 of deposits increases from Rs79 to Rs91.6, lifting the effective asset yield from 8.12% to 8.53%.

Using a deposit-cost assumption of 4.53% in this scenario, the net spread rises from 3.6% to 4.01%, implying a roughly 41bps benefit on the incremental FCNR(B) book.

At the overall-book level, applying the roughly 41bps incremental benefit to FCNR(B)’s approximately 4.5% share of system deposits implies around a 2bps improvement in blended spread/NIM, assuming the remaining domestic deposit base remains unchanged.

Domestic TDs versus FCNR(B)

Anand Rathi said domestic rupee deposits are an expensive way to fund loans because a meaningful portion cannot be deployed into the bank’s highest-yielding assets.

Out of every Rs100 raised, around Rs3 goes to CRR, which earns zero yield, while Rs18 goes to SLR, parked in government securities yielding around 6.5%.

This leaves Rs79 available for lending. Of this amount, around 40% must be allocated towards priority-sector lending (PSL), which earns a lower yield of around 8.5%.

FCNR(B), in contrast, is exempt from CRR and SLR requirements. In addition, eligible advances against specified fresh FCNR(B) deposits are excluded from the adjusted net bank credit calculation for PSL targets. Therefore, the entire Rs100 can be deployed towards the bank’s highest-yielding non-PSL loans, subject to the applicable regulatory conditions.

The report assumes a TD/FCNR(B) cost of 6.5%, a G-sec yield of 6.5%, a PSL yield of 8.5% and a non-PSL yield of 9%.

“The 6.5% headline cost of domestic TDs applies to full Rs100, while only Rs79 is available for lending, with a further ~40% of deployable funds constrained to lower-yielding PSL assets,” Anand Rathi said.

“FCNR(B) eliminates both ~Rs21 reserve drags and the PSL mix-down, allowing the full Rs100 to be deployed into the highest-yielding non-PSL book.”

The brokerage therefore said FCNR(B) is structurally spread-accretive versus domestic TDs even before considering the interest-rate cycle.

Scenario 3: The leverage overlay

The third scenario examines how leverage can unlock RoE through zero-RWA lending.

Banks can offer FCNR(B) as a leveraged deposit, where the NRI pledges the deposit and the bank lends against it. The report assumes leverage of around nine times the deposit.

This raises the customer’s effective return above the headline 6.5% FCNR(B) coupon and allows the bank to build a larger foreign-currency loan book.

According to Anand Rathi, these leveraged loans are essentially loans against overdraft or pledged deposits and carry zero risk weight, resulting in minimal incremental RWA.

The lending spread is relatively thin at around 50bps, but it is earned on a near-zero-RWA book. Multiple banks also charge fees on these leveraged loans.

The key attraction of the leveraged structure, analysts believe, is capital efficiency.

The overall structure generates a roughly 1.55% net blended spread for the bank, before factoring in fees from the leveraged loans.

“With zero/near-zero RWA and minimal incremental capital consumption, the spread and fee income translate into strong RoE accretion, allowing banks to scale the business without materially consuming CET-1 capital,” Anand Rathi said.

Scenario 4: FCNR(B) gains from rising rates

The fourth scenario focuses on the duration advantage offered by FCNR(B) funding.

FCNR(B) deposits are locked in for 3-5 years, while the asset book is largely floating-rate. Anand Rathi modelled 50bps and 100bps repo-rate hikes over the next year and traced the resulting spread over a one- and two-year horizon.

The analysis assumes that the asset book is predominantly floating, linked to EBLR/repo rates, with a roughly 1x repricing beta. FCNR(B) funding remains fixed over the period, with a beta of around zero, while domestic term deposits reprice with a lag, with around 50% pass-through in Year 1 and 100% by Year 2.

“The longer tenor of FCNR(B) creates a natural asset-liability duration mismatch in the bank's favour during a rising-rate cycle: loan yields reprice upwards, while FCNR(B) funding cost remains locked in,” the report said.

The spread therefore widens as rates rise, with the benefit increasing with the magnitude of the repo hike and persisting through the initial repricing period.

“This makes FCNR(B) particularly attractive when banks expect rates to move higher,” Anand Rathi said.

The report noted that the advantage reverses in a rate-cut cycle, making the benefit contingent on a bank’s interest-rate view. It said a rising or higher-for-longer rate path is the more likely near-term backdrop.

(Disclaimer: Recommendations, suggestions, views and opinions given by the experts are their own. These do not represent the views of Economic Times)