India External Sector · September 2026 Analysis

The $127 Billion
Illusion

India raised more from a three-month leveraged NRI deposit scheme than it attracted in net foreign direct investment over four years. That is not a triumph of capital attraction. It is a confession of its failure.

Genuine foreign capital vs engineered inflows — two periods compared

Net FDI · FY2013–16
Foreign capital on its own steam
$97B Genuine 4-year net inflow · UPA era

Net FDI · FY2023–26
Foreign capital on its own steam
$46B Genuine 4-year net inflow · current era
FCNR(B) scheme · 2026
Engineered inflow over 3 months
$127B Leveraged · subsidised · forex risk nationalised

The ratio that matters: In 2013 the FCNR raise was 27% of the period's net FDI. In 2026 it is 276% — nearly three times the entire four-year net FDI. The scheme is no longer a supplement to genuine capital attraction. It is a substitute for it.

In the era India calls its worst, foreign capital came willingly. In the era India calls its best, it had to be engineered.

The FDI Story Nobody Is Telling

Gross FDI looks three times better. Net FDI collapsed.

Every government press release on foreign investment quotes gross FDI. India recorded $81 billion in gross FDI inflows in FY25 — a 14% rise. A triumph, by that measure.

Net FDI — what actually stays after repatriation, disinvestment, and outward investment by Indian companies — tells a different story entirely. Net FDI decreased to $0.96 billion in FY25. Not $96 billion. Not $9.6 billion. Less than one billion dollars in a year when the gross number was $81 billion.

The gap between gross and net is the money leaving. Foreign investors taking profits out. Earlier investors exiting. Indian companies investing abroad at a faster rate than foreign capital is coming in. In FY26, total dollars flowing out of India exceeded inflows by $30.8 billion — a more than six-fold increase over FY25.

Year Gross FDI Net FDI Retention Rate What it signals
FY13 ~$22B ~$19B 86% Foreign capital staying
FY14 ~$24B ~$21B 88% Stable, high retention
FY15 ~$31B ~$27B 87% Genuine growth
FY16 ~$36B ~$30B 83% Peak genuine inflow
FY13–16 total ~$113B ~$97B 86% UPA's "worst period"
FY23 ~$71B $28B 39% Gap beginning to widen
FY24 ~$71B $10B 14% Repatriation accelerating
FY25 ~$81B $0.96B 1.2% Effectively zero net
FY26 ~$84B $6.95B 8.3% Net outflow year overall
FY23–26 total ~$307B ~$46B 15% Current "best period"

Source: RBI Balance of Payments data, Ministry of Commerce & Industry, Rajya Sabha written reply July 2026. Net FDI = gross inflows minus repatriation/disinvestment minus outward FDI by Indian companies.

"Gross FDI tripled. Net FDI collapsed to near zero. The retention rate fell from 86% to 15%. That is not an investment attraction story. That is an investment exit story."

The Leverage Structure

What $127 billion actually is

When the RBI announced the FCNR(B) swap scheme in June 2026 and it mobilised $127.2 billion by August 31, it was described as proof of global confidence in India. The number deserves closer examination.

The scheme enabled a leverage structure that banks actively marketed. An NRI puts in $100,000. The Indian bank issues a Standby Letter of Credit to an overseas lender. The overseas lender lends the NRI $900,000 against that guarantee. The NRI deposits the full $1,000,000 as FCNR(B). The deposit earns 6.5%. The loan costs 5.7%. The NRI pockets a leveraged return of 15–16% on their original $100,000.

Some foreign banks — competing aggressively for business — pushed the leverage ratio to 19x, not 9x. At 19x, the real fresh dollar from the NRI's own pocket is just 5% of the deposited amount.

How the leverage structure works

01 NRI contributes $100,000 of own money
02 Indian bank issues Standby Letter of Credit (SBLC) to overseas lender Bank risk
03 Overseas lender provides $900,000 loan to NRI at ~5.7%
04 Full $1,000,000 deposited as FCNR(B) at 6.5% — this enters RBI swap
05 RBI swaps dollars for rupees at today's rate — full forex risk absorbed by RBI Sovereign risk
06 NRI earns spread on 9x leveraged capital — effective return 15–27% on own money
07 At maturity 2029: NRI repays loan, takes deposit back, exits. RBI must return full $1M at original swap rate regardless of where rupee is. Sovereign cost
5–10% Genuine fresh NRI dollar at maximum 9x leverage. Rest is borrowed from overseas banks against Indian bank guarantees.
19x Maximum leverage some foreign banks offered to win FCNR business. At 19x, real NRI money is just 5% of the deposit.
$127B RBI's swap obligation — against the full leveraged deposit, not just the genuine NRI contribution.

There is no public disclosure of what proportion of the $127.2 billion was leveraged. The RBI has not released a breakdown. The headline number — celebrated as proof of NRI confidence in India — conflates genuine diaspora savings with an arbitrage trade structured by Indian banks and overseas lenders.

The 2013 Comparison

Last time was emergency medicine. This time is elective surgery at five times the dose.

The 2013 FCNR scheme is the template being invoked to justify the 2026 one. The comparison deserves to be made precisely, not selectively.

Parameter 2013 Scheme 2026 Scheme Direction of risk
Amount raised ~$26B $127.2B 5x larger
Trigger Rupee crisis — Fragile Five Preventive — reserves adequate Weaker justification
Hedging cost borne by banks 3.5% p.a. 0% 100% shifted to RBI
US borrowing rate for NRIs ~0.25% ~5.7% Leverage trade much thinner
FCNR-to-NetFDI ratio 27% 276% Supplement → substitute
Max leverage available 10–20x up to 19x Similar — but spread far thinner
RBI pre-existing forward book Modest $64B+ before scheme Added to a stressed book
Even then, it was called "FCNR Scam" — Business Standard letters, September 2013. "Least bad option" — Rajan himself, 2016.

In 2013, economists including Ajit Ranade of Aditya Birla Group explicitly warned that leveraged NRI deposit flows could reverse as abruptly as they arrive. The 2016 unwind of just $20 billion required Rajan to publicly reassure markets, caused visible weekly declines in reserves, and put pressure on the rupee. Rajan called the scheme the "least bad option" — not a triumph. A necessary gamble under crisis conditions.

In 2026, the same mechanism has been deployed at five times the scale, at zero hedging cost to banks, into a forward book that was already stretched, with reserves that were described as "comfortable" before the scheme launched. The crisis justification that made 2013 defensible does not exist in 2026.

The Forward Liability

What ₹1.27 lakh crore of exposure looks like in 2029

At the swap inception rate of approximately ₹94 per dollar, the RBI's obligation is to return $127.2 billion at that rate in 2029 — regardless of where the rupee actually is.

The rupee fell 9.6% against the dollar in FY26 alone. It touched a record low of ₹96.97 in May 2026. The structural depreciation trend over the last decade has averaged 3–5% annually. A conservative move from ₹94 to ₹104 by 2029 — less than the FY26 depreciation in a single year — produces the following:

₹1.27L cr RBI's swap loss if rupee moves from ₹94 to ₹104. In dollar terms: ~$12.2 billion.
~$25B Expected loss at ₹114 — one standard deviation above the conservative scenario.
$38–50B Midpoint expected loss range at historical 6% annual depreciation over 3–5 year tenor.

None of this appears on any balance sheet today. It sits in the RBI's swap book as a contingent liability, accruing quietly until 2029 when the deposits mature in a concentrated window. The forex reserves — celebrated at a record $729 billion — include the dollars that came in under FCNR. Those dollars are not free reserves. They are matched by a forward obligation to return them at the swap rate.

Gross reserves of $729 billion net of the forward book — now at a record $136.7 billion and rising — gives freely deployable reserves of approximately $592 billion. Still large. But 19% smaller than the number in every press release.

The question is not whether the scheme worked. It is what it reveals.

In FY13–16 — the period India describes as its worst economic stretch — net FDI was $97 billion over four years. Foreign capital came because the underlying investment case was strong enough, even through a crisis, to retain 86 cents of every dollar that came in.

In FY23–26 — the period India describes as its strongest — net FDI was $46 billion over four years, with a retention rate of 15%. The economy needed a three-month, heavily subsidised, RBI-guaranteed, leverage-amplified scheme to raise $127 billion to compensate for what genuine capital attraction is no longer delivering.

The FCNR scheme is not evidence of India's external strength. It is a measure of the gap between the external strength being claimed and the external strength that actually exists.

The $127 billion is real. The confidence it represents is not.

Sources & qualifications: Net FDI data from RBI Balance of Payments statistics and Ministry of Commerce & Industry (Rajya Sabha written reply, July 28, 2026). Gross FDI from DPIIT. FCNR(B) final position from RBI Press Release 2026-2027/1024, dated September 2, 2026 (signed Brij Raj, Chief General Manager). Leverage structure details from Canara Bank marketing materials, Vinod Kothari Consultants analysis (July 30, 2026), Jefferies India research, and JM Financial analysis. The 2013 comparison uses figures from RBI Annual Reports and Business Standard contemporaneous reporting. Forward book data from RBI Weekly Statistical Supplement. The $38–50B loss estimate uses a 6% annual rupee depreciation assumption compounded over 3–5 years on $127.2B — consistent with the rupee's historical average depreciation against the dollar since 2014. Actual cost will depend on the rupee/dollar rate at each deposit's maturity date between 2029–2031. No public disclosure exists of the proportion of the $127.2B that was leveraged vs genuine fresh NRI savings. This is a critical disclosure gap. The RBI's decision not to publish this breakdown is itself a data point.
Also in India & Economy
7.8% and Zero — India's Two Economies →