When a central bank volunteers that something is cheap, the honest response is to ask who is paying.
On a routine briefing out of Mumbai, the Reserve Bank of India did something macro traders rarely get: it preemptively rejected the cost concerns surrounding its foreign-currency deposit drive, insisting the program would add to reserves, put a floor under the rupee, and — in the same breath — offset its own hedging expense through investment returns. Two of those claims can coexist. All three cannot. If the hedging cost were genuinely self-liquidating through reserve portfolio yield, no central bank would ever need to run a drive at all; banks would already be arbitraging it to death. Structural skepticism active.
That isn't cynicism. It's arithmetic. The phrasing itself is a tell. Central banks deny cost problems in public only when the cost problem has already been raised in private — by the institutions expected to execute the program.
India's foreign-currency deposit drive belongs to a specific shelf of the intervention toolbox: market-based inflow attraction, as distinct from outright reserve liquidation. The template is old and well documented. During the 2013 taper tantrum, with the rupee in free fall and reserves bleeding, the RBI opened a special FCNR(B) swap window, effectively paying banks to bring non-resident dollar deposits onshore while absorbing the currency risk on the sovereign's own balance sheet. It worked. It also left behind a precise and portable lesson about who eventually pays.
Here is the mechanism, compressed. A bank takes a dollar deposit and swaps it into rupees for domestic placement. To avoid carrying an open FX position, it hedges — selling rupees forward for dollars. The price of that hedge is the forward premium, which under covered interest parity approximates the India-US interest differential. The forward premium is not a fee; it is the market's quoted price of the state's monetary credibility gap. When that premium is wide, hedged dollar funding is expensive and the drive stalls. When the RBI "rejects cost concerns," it is not disputing the premium. It is signaling that the premium will be absorbed somewhere else — most plausibly a targeted swap window, a rate concession, or an explicit sovereign backstop. Quasi-fiscal cost, wearing a monetary policy costume. Macro lens focused.
Consider the bank's economics, illustratively. If the dollar deposit costs an offshore benchmark plus a spread, and the forward premium runs roughly in the neighborhood of the policy differential, the hedged all-in rupee cost lands above domestic deposit rates by the width of a spread the bank cannot hedge away. That gap is a tax with three candidate payers: the bank's margin, the depositor's rate, or the sovereign's balance sheet. In 2013 the answer was the third. The denial is therefore best read not as a factual claim but as an announcement that the third payer is coming.
Liquidity check engaged: the second-order effects matter more than the plumbing. Dollar deposits arriving onshore create rupee liquidity when converted. If the RBI sterilizes — mopping up with central bank paper, cash reserve ratio adjustments, open market operations — it pays interest on the mop, adding a second cost layer on top of the first. If it doesn't sterilize, the liquidity feeds domestic credit and asset prices, which is stimulative in precisely the way a central bank defending a currency usually does not want to be. This tool gets reached for when the interest rate instrument is already pinned by domestic inflation — which tells you the policy corridor has stopped being a fully available degree of freedom. India is choosing between a bill it can see and a risk it cannot. Either branch confirms the program carries a real cost. They differ only in where it books.
Now the dimension entirely absent from the coverage: reserve accretion quality. A deposit-driven reserve increase is a liability-side expansion. It is not earned through exports, not the residue of a current-account surplus, not the mark-to-market uplift from revaluing gold or Treasuries. It is borrowed dollars wearing a "reserves up" headline — and the durability of borrowed dollars is a function of their stickiness. Deposits that migrate in during a period of rupee weakness can migrate out the moment that weakness inverts or the differential compresses. The headline number rises on print. The margin of safety may not.
This is where my own work inverts the usual reading. In 2013, the RBI's competition for NRI dollars was bank term deposits and a handful of sovereign schemes. The pitch was simple: a rate, a tenure, a government-adjacent guarantee. The alternatives were slow, jurisdictional, and heavily intermediated. In 2025-26, that is no longer true. The offshore dollar currently has a live, permissionless, always-on alternative: tokenized money market funds, on-chain Treasury wrappers, and stablecoin savings rails that clear without a correspondent bank, without a lock-up, and without asking anyone's permission. Exiting a tokenized dollar position at 2 a.m. is one signature, not three weeks of paperwork.
That changes the elasticity of India's accessible dollar pool in a way no 2013-era model captures. A sovereign can compel its banks. It cannot compel a wallet. The premium required to pull onshore dollars out of on-chain instruments is structurally higher than the premium required to pull them out of a term deposit — and, more importantly, the effective duration of money with an always-available exit is shorter. The deposit doesn't settle. It parks. Modular resilience observed — but the resilience belongs to the alternative rails, not to the reserve book.
I learned to read this shape in 2020, when I built a model tracing yield-farming incentive loops across lending protocols and found that reported capital efficiency was being manufactured rather than earned. The lesson generalized cleanly: incentives attract TVL, but they do not retain users. Remove the subsidy and the flow reverses within days. A dollar-deposit drive with an implicit premium subsidy has exactly that geometry. The relevant question is never whether it works while subsidized. It is what the flow looks like the quarter after the subsidy is quietly withdrawn — and whether anyone notices the withdrawal in time.
There is an investment-return offset embedded in the RBI's framing, and it deserves a fair hearing before dismissal. Reserve assets parked in short-dated US Treasuries yield a positive nominal return. The forward premium paid to hedge is a cost. If portfolio yield roughly matches the premium, the carry is a wash and the drive costs the sovereign little beyond operational friction. That arithmetic holds only in one macro state: a compressed differential. It requires US short rates to fall relative to India's, India's to rise, or both. The RBI's cost-denial is therefore an implicit rate-differential bet, expressed through the reserve book rather than the policy corridor. If that bet is wrong — if the differential stays wide, or global dollar funding tightens — the offset evaporates and the subsidy becomes legible to everyone.
And here is the contrarian reading. The dominant narrative around episodes like this is de-dollarization: BRICS settlement rails, reserve diversification, the slow retreat from dollar hegemony. This policy points the other way. It is a deliberate re-dollarization of India's domestic balance sheet, increasing the rupee system's dependence on dollar liquidity, and doing so through market instruments rather than coercion — precisely because the coercive version is expensive and visible. Reserve diversification is what you do when you want less dollar risk. This is what you do when you want more dollars, faster, on better terms. The operation is defensive and technically proficient, and it is not a step away from the dollar system. It is a step deeper into it, taken while preserving the rhetorical option to say otherwise.
The genuinely contrarian claim is subtler. The binding constraint on Indian external stability has never really been the level of reserves. It has been the credibility of their usability — whether they can be deployed without the deployment itself becoming the story that accelerates the panic. A deposit drive sits exactly on that fault line. It preserves headline reserves by importing a liability, which looks like strength in the data and reads as improvisation in the market. Structural skepticism active, again — not because the RBI is incompetent, which it plainly is not, but because this instrument's optics and its economics point in opposite directions, and the market prices whichever one it believes first.
My position, after tracking institutional dollar plumbing through the ETF complex in 2024 and testing it against the 2013 FCNR(B) template: watch three things and ignore the rest. First, whether a special swap window or rate concession appears within the next several months — that is the definitive confirmation that the cost was real and the denial was anticipation management. Second, the USD/INR forward premium at the one-to-three month tenors — if the premium refuses to compress, the drive is fighting the market's own pricing of risk rather than shaping it. Third, migration in NRI remittance corridors toward tokenized dollar instruments, because that is where the marginal offshore dollar now clears.
If the marginal price of a dollar deposit is increasingly discovered on-chain — outside capital controls, outside bank intermediation, at 2 a.m. on a Saturday — then what exactly is the RBI defending? The rupee, or the last mile of its capital account?
