NextFin News - Citigroup and Axis Bank have teamed up to offer leveraged dollar deposits to non-resident Indian investors, a deal that marks a clear escalation in India's sprint for foreign-currency funding: the competition has moved from who offers the highest rate to who will lend the most against a deposit. The pairing arrives as outstanding FCNR(B) deposits across Indian banks nearly doubled to $60.55 billion between June 5 and July 30, 2026, an 86 percent surge that has reshaped the country's external funding picture in under eight weeks.
The significance of the arrangement is not that leveraged FCNR structures are new — several banks were already offering leverage of nine times an investor's capital, with select private lenders at 12 to 15 times and at least one foreign bank at 19 to 29 times, according to CareEdge Ratings. The significance is who is now in the trade: a global bank with deep offshore dollar funding capacity and one of India's largest private-sector lenders with one of the country's biggest NRI franchises. Together they are betting that the window the Reserve Bank of India opened in June will not close quietly at the end of September, and that the diaspora's dollars are worth borrowing against at scale.
The Situation: A Window That Created a Boom
The boom has a precise starting point. On June 5, 2026, the RBI governor announced a special US dollar-rupee forex swap facility for fresh FCNR(B) deposits, formalised in circular RBI/2026-27/99 dated June 8. Under the facility, banks can mobilise three-to-five-year foreign-currency non-resident deposits until September 30, 2026, and swap the principal with the central bank at a fixed rate, with the swap facility itself available to banks until October 16. The stated purpose, as the minister of state for finance Pankaj Chaudhary put it in a written reply to the Lok Sabha, was to "attract stable foreign currency inflows, strengthen India's balance of payments and help ease recent pressures on the Indian Rupee."
attract stable foreign currency inflows, strengthen India's balance of payments and help ease recent pressures on the Indian Rupee.
- Pankaj Chaudhary, minister of state for finance, in a written reply to the Lok Sabha
The response was immediate and large. FCNR(B) deposits outstanding with authorised dealer banks rose from $32.56 billion on June 5 to $60.55 billion on July 30, meaning banks mobilised $27.99 billion of fresh foreign currency in less than two months. State Bank of India led the pack, adding $4.13 billion to take its book to $13.82 billion, followed by ICICI Bank with $3.70 billion of fresh inflows, Axis Bank with $1.60 billion, and HDFC Bank with $1.40 billion. By July 30, the five largest books were SBI at $13.82 billion, HSBC at $6.26 billion, ICICI at $6.06 billion, HDFC at $5.42 billion, and Axis at $4.67 billion. The central bank later said total inflows under the scheme had reached $36.7 billion.
Axis Bank's own trajectory illustrates the competitive intensity. Its FCNR(B) book grew from $3.08 billion on June 5 to $4.67 billion on July 30, and on August 17 the bank raised its rate on FCNR(B) deposits above $1 million for three-to-five-year tenors to 6.40 percent, the highest among large private lenders. That rate war is the first layer of the story. The Citi-Axis tie-up is the second: rather than compete on price alone, banks are now competing on how much they will lend against a customer's own deposit.
The mechanism is straightforward. An NRI puts up a portion of the funds in an FCNR(B) deposit, then borrows additional foreign currency against that deposit — typically at a floating rate benchmarked to SOFR plus a spread — and reinvests the borrowed proceeds in the same FCNR(B) product. The investor earns the deposit rate on the full, leveraged amount while paying the borrowing rate only on the loaned portion. The difference, the spread, is the profit. Because the RBI swap absorbs the currency-hedging cost on the principal, banks can afford to pay deposit rates of 6 to 7 percent in dollars, versus the 3 to 4 percent range that prevailed before the window opened. What looks like a deposit product is, in substance, a subsidised carry trade wrapped in a fixed-income wrapper.
That raises the question this piece pursues: is the Citi-Axis deal a sign of durable structural change in how India funds itself, or a cyclical sprint that will reverse once the subsidy expires? The answer matters because $60 billion of hot money that arrived in eight weeks can leave just as fast — unless the structure underneath it is stickier than it looks.
The Swap, Not the Deposit, Is the Real Product
To understand why leverage suddenly makes sense, start with the swap. In a normal FCNR(B) deposit, the bank takes in dollars, converts them to rupees to lend in India, and must hedge the currency risk so it can return dollars at maturity. That hedge historically cost roughly 3 to 3.5 percent a year, which is why pre-window dollar deposit rates languished in the 3 to 4 percent range even as domestic rupee rates sat far higher. The RBI's 2026 facility removes that cost for the principal: the bank hands the dollars to the central bank, receives rupees at today's rate, and gets the same number of dollars back at maturity at the same rate, regardless of where the rupee trades. The bank keeps the rupee lending margin; the depositor gets a dollar rate that finally competes with overseas alternatives.
This turns the FCNR(B) deposit from a marginal product into a policy instrument. The deposit is the wrapper; the swap is the subsidy. And the subsidy is doing exactly what it was designed to do. In the financial year ended March 2026, FCNR(B) inflows had plunged 86 percent to $946 million from $7.1 billion a year earlier, as earlier incentives expired and rates lost their appeal. Within eight weeks of the new facility, inflows had not just recovered but surpassed the entire prior year's stock. The before-and-after is the cleanest evidence that the flow is incentive-driven rather than conviction-driven.
The second-order effect is what makes the Citi-Axis deal noteworthy. India is not merely attracting deposits; it is importing dollars through its diaspora at a subsidised hedge, in direct competition with two other funding channels: external commercial borrowings by companies and overseas foreign-currency borrowing by banks. A note from IDFC First Bank captured the scale of the shift:
The pace of capital inflows under FCNR(B), external commercial borrowings and overseas foreign currency borrowings has been much stronger than expected. Total inflows across instruments are likely to be $90 billion, if not higher.
- IDFC First Bank, in a research note
If that estimate holds, the FCNR(B) window has effectively crowded out more expensive forms of dollar borrowing, lowering India's aggregate cost of external funding. That is a balance-of-payments win that does not show up in any single bank's deposit report.
But the subsidy has an expiry date printed on it. Deposits must be mobilised by September 30, 2026, and the swap facility runs only until October 16. After that, banks either renew deposits at market hedging costs or let them mature. The Citi-Axis arrangement is therefore best read as a race to lock in cheap funding before the clock runs out — a sprint, not a marathon.
Leverage Turns a Six-Percent Deposit Into a Twenty-Percent Bet
The leverage layer is where the advertised returns turn eye-catching, and where investors most often misread the math. Consider an illustrative structure at a 6.45 percent deposit rate and a 5.75 percent borrowing cost, figures used in industry analysis of these products. An investor putting up $100,000 of their own capital with no leverage earns $6,450 a year, a 6.45 percent return. With nine times leverage, that $100,000 supports a $1 million deposit. Interest income rises to $64,500, borrowing costs are $51,750, and the net is $12,750 — a 12.75 percent return on the investor's own capital. At 19 times leverage the return on own capital reaches 19.75 percent; at 29 times, 26.75 percent.
Those headline numbers are real arithmetic, but they are not the whole story. Two distortions are embedded in most marketing material. First, published yields typically divide total interest over the tenure by the original principal and by the number of years, ignoring compounding and the fact that there are no interim payouts. A wealth-management analysis of circulated offer documents found that an Axis Bank leveraged structure advertised at 17.30 percent a year for a three-year tenor at 19 times leverage works out to a compounded annual return of 15.17 percent. The same analysis showed a nine-times, five-year SBI structure advertised at 13.83 percent actually yields 11.09 percent on a compounded basis. The gap between the advertised figure and the compounded figure widens with both leverage and tenor.
Second, and more important, the borrowing leg is often floating. The spread that makes the trade profitable — deposit rate minus borrowing cost — is fixed only on the deposit side. If SOFR rises during the life of the loan, or if the bank's spread widens, the net margin compresses and can flip negative. At nine times leverage, a one-percentage-point adverse move in the borrowing rate wipes out $9,000 a year on a $1 million leveraged position; at 19 times, nearly $19,000. Leverage does not amplify returns in isolation; it amplifies exposure to the very rate variable the structure is supposed to insulate the investor from.
There is also a liquidity cost that the headline return does not price. These deposits carry a one-year lock-in during which premature withdrawal is not permitted, and the loan against the deposit remains outstanding throughout. An investor who needs cash before maturity cannot simply redeem; they must service the loan or find alternate collateral. The structure converts a liquid dollar balance into a five-year commitment with a leveraged liability attached. That is a fundamentally different risk profile from the plain FCNR(B) deposit the RBI intended to promote.
The practical lesson: the advertised return is a function of the leverage multiple, not of credit quality or bank safety. Two banks offering the same deposit rate can produce materially different outcomes depending on how much they are willing to lend and on what terms. An investor comparing 6.40 percent at one bank with 7.10 percent at another is not comparing the same product if the first offers no leverage and the second offers 19 times.
Who Wins, and Who Is Left Holding the Rate Risk
Follow the spread and the winners become clear. Indian banks are the primary beneficiaries: they have swapped expensive, flighty dollar funding for cheap, policy-backed funding with a three-to-five-year maturity, improving both their foreign-currency liquidity and the duration profile of their liabilities. The RBI wins too: $28 billion of fresh dollars in eight weeks adds a buffer to reserves and eases pressure on the rupee without the central bank having to sell its own holdings. NRIs who understand the structure win the spread — provided they hold to maturity and rates do not move against them.
The risk, however, has not disappeared; it has been relocated. The currency risk on the principal sits with the RBI, but the interest-rate risk on the leverage sits squarely with the depositor, and the credit risk of the lending bank sits there as well. A wealth manager advising on these structures has cautioned that investors should verify in writing whether the borrowing rate is fixed, and should not allocate more than 25 to 30 percent of financial assets to such illiquid, five-year instruments. That guidance is conservative for a reason: the structure's profitability depends on three variables — the deposit rate, the borrowing rate, and the exchange rate at which the bank ultimately settles — and the investor controls none of them.
There is also a system-level asymmetry worth naming. The deposits are denominated in dollars and fully repatriable. If the rupee strengthens or overseas rates rise enough to make the spread unattractive, the same diaspora dollars that rushed in can rush out after the lock-in expires. India's experience with earlier FCNR windows is instructive: the 2013 scheme, launched during the taper-tantrum currency crisis, brought in $34 billion but also created a large maturity wall that the central bank had to manage carefully. The current window is smaller in ambition but similar in shape — a time-bound incentive that solves a near-term funding need while creating a future redemption profile.
This is where the Citi-Axis deal cuts both ways. On one reading, the entry of a global bank with sophisticated risk management and offshore funding desks professionalises the market and broadens the investor base. On another, it signals that the easiest money has already been raised from domestic banks, and the marginal dollar now requires cross-border structuring and higher leverage to attract. Both readings can be true: the market is maturing even as the low-hanging fruit is gone.
Cyclical Surge Riding a Structural Base
So is this cyclical or structural? The honest answer is both, and keeping them separate is the only way to reach a defensible conclusion.
The surge itself is cyclical. It is policy-created, time-bound, and incentive-sensitive. The 86 percent collapse in FCNR(B) inflows in the financial year ended March 2026, when the previous incentives expired, is direct proof: remove the subsidy and the flow evaporates. The 86 percent rebound in eight weeks is the mirror image. Neither the surge nor the collapse reflects a change in the underlying attractiveness of Indian assets; both reflect the price the regulator was willing to pay. When the swap facility closes in October and deposits can no longer be booked under the window after September 30, the flow will normalise. A cyclical wave, by definition, mean-reverts.
But the base it rides is structural. India's non-resident population is a permanent source of foreign-currency savings, and the diaspora's preference for rupee-linked, dollar-denominated assets does not disappear when a window closes. Total NRI deposits — including NRE and NRO accounts — stood at $165.65 billion at the end of the financial year ended March 2026, and the NRI share of bank funding has been a stable pillar through multiple currency cycles. What changes across cycles is the price and the instrument, not the existence of the pool. If the rupee remains under pressure and overseas dollar yields stay attractive, banks have a lasting incentive to keep tapping this pool, even at market pricing.
The correct framing, then, is a cyclical wave riding a structural base. The 86 percent surge will not repeat; it was a one-off policy event. But the channel it reopened — leveraged, dollar-denominated NRI funding as a core component of India's external financing — is likely to persist, because the structural driver (a large overseas Indian savings pool seeking rupee exposure without currency risk) outlives the subsidy. Confusing the two leads to the two most common errors: assuming the boom is permanent, or assuming nothing has changed at all.
What to Watch: Scenarios and the Signal That Would Prove This Wrong
Three time horizons matter here, and they point in different directions.
In the short term, into the September 30 booking deadline, expect a mobilisation sprint. Banks will push leverage multiples and rate offers to the maximum the economics allow, and total FCNR(B) inflows under the scheme are widely expected by brokerages to land in the $50 billion to $55 billion range. The Citi-Axis tie-up is part of that sprint, not a standalone strategy shift.
In the medium term, through 2027, the key variable is what happens at renewal. Once the swap subsidy ends, the deposit rate must compete with unhedged alternatives. If SOFR stays elevated and the rupee is stable, leveraged structures compress and many deposits will not renew. If the rupee weakens, the currency-hedge value of the product persists even without the subsidy, and renewals hold up better.
In the long term, the structural question is whether India can retain NRI dollar funding at market pricing. The evidence from the financial year ended March 2026 suggests the answer is partial: the pool is sticky at the right price, but price-sensitive enough that a 300-basis-point move in offered rates can shift the flow by tens of billions. That makes NRI funding a reliable pillar, but a costly one in stressed markets.
Three scenarios frame the path ahead. The base case is that FCNR(B) inflows reach the $50 billion to $55 billion range by the September deadline, the window closes as scheduled, and spreads normalise toward 4 to 5 percent as the hedge subsidy fades. The upside case is that persistent rupee weakness keeps NRI demand strong, inflows exceed $60 billion, and regulators extend or replace the facility in some form. The downside case is that a rise in SOFR or a sharp rupee appreciation makes the leveraged spread unattractive, triggering heavier-than-expected redemptions once the one-year lock-in expires.
The falsifying signal is specific: if FCNR(B) outstanding falls below $50 billion by the end of 2026 — meaning redemptions and non-renewals have erased a meaningful share of the $28 billion raised — then the thesis that NRI dollar funding has become a sticky structural pillar is wrong, and the 2026 boom should be read as a purely cyclical, subsidy-driven event that has fully reversed.
The Citi-Axis deal, in the end, is less about two banks and more about what India is willing to pay for dollars. The country has not solved its external funding need; it has rented $60 billion of it from its diaspora for a quarter, at a subsidised rate, and the rent comes due in September. Whether that rental becomes a lease depends on what the rupee does next — and on whether the spread, not the subsidy, can carry the trade.
Explore more exclusive insights at nextfin.ai.
