On 24 August 2026, the Ministry of Finance announced that the Reserve Bank of India's special USD-INR forex swap facility — open to FCNR(B) deposits, Overseas Foreign Currency Borrowings (OFCB) and External Commercial Borrowings (ECB) — had mobilised $73 billion as on 21 August 2026. FCNR(B) deposits alone accounted for $65.40 billion. The facility was launched on 8 June 2026, so the money arrived in under eleven weeks. The response was strong enough that the RBI advanced the closure of the FCNR(B) window from 30 September to 31 August 2026 — it shut the tap early because the objective was met.
That is the news. The reason it is worth a full study session is that a single scheme touches almost every external-sector concept the CDS/OTA General Knowledge paper draws on: what a swap actually is, why forward premia exist, how reserves are built, what counts as debt, and the constraint economists call the impossible trinity.
What a currency swap auction actually is
Strip away the jargon. A foreign exchange swap is two transactions agreed at the same moment: a spot leg now and a reverse forward leg at a fixed future date, at a rate agreed today. No one is speculating on the exchange rate; the price of the deal is the difference between the two legs, which is the swap premium.
Direction is everything, and it is the single most examinable distinction here.
- In a buy/sell swap, the RBI buys dollars spot from banks and pays out rupees, then sells those dollars back forward at maturity. Rupees leave the RBI and enter the banking system: rupee liquidity is injected now and withdrawn later. Reserves rise in the interim.
- In a sell/buy swap, the RBI does the reverse — sells dollars spot and takes rupees in, then buys the dollars back forward. Rupee liquidity is absorbed.
When the RBI runs these as auctions, banks bid the forward premium in paisa terms, and the auction is settled on a multiple-price basis — each successful bidder pays its own quoted premium. That is the ordinary liquidity-management version of the tool.
The June 2026 facility is a different animal, and the difference matters. It is not a liquidity operation dressed up as a forex one; it is a subsidy on hedging cost, designed to make dollar deposits in Indian banks unnaturally attractive.
The mechanism: who pays for the hedge
Here is the problem the scheme solves. A bank that takes a dollar deposit and lends in rupees carries currency risk. To neutralise it, the bank must hedge in the forward market — and hedging a 3-to-5-year dollar-rupee exposure was costing roughly 2.8% to 3.3% a year, that being the prevailing swap rate for the tenor. That cost eats the entire spread. So the bank cannot offer an attractive dollar rate to a non-resident depositor.
The RBI's answer was to absorb the hedging cost itself — the announcement of 5 June 2026, operationalised by circular on 8 June 2026, offered a concessional swap worth roughly 280 to 300 basis points, effectively the whole prevailing cost. With the hedge free, banks could pass the economics through and quote dollar deposit rates in the 5.5% to 7.1% range, which is extraordinary for a hard-currency deposit. Non-resident money moved accordingly.
The other two windows work on the same principle but for institutions rather than individuals: the OFCB window for AD Category-I banks and the ECB window for public sector undertakings, both requiring a minimum maturity of three years, with drawdowns and swap access running into January 2027. The design intent throughout is long-dated money — nothing that can leave in a hurry.
The swap window did not travel alone. The same June package widened the Fully Accessible Route to all new 15-, 30- and 40-year government securities and removed capital-gains and interest-income tax for foreign portfolio investors with retrospective effect from 1 April 2026. This is a coordinated capital-account push, not a single lever, and the CDS/OTA economy notes treat it that way.
Why now: the rupee's 2026
Context makes the policy legible. The rupee crossed 96 to the dollar for the first time in its history in May 2026 — reports of the exact record low differ, clustering somewhere between roughly 96.4 and 96.8, so quote "past 96" rather than a decimal you cannot defend. The pressure came from a stronger dollar, elevated crude, geopolitical risk and portfolio outflows.
Two points of proportion are worth carrying, because they defend against panic-framing. First, the 2026 depreciation of about 7% was smaller in percentage terms than the roughly 20% fall of the 2013 taper tantrum or the 10% decline of 2022. Second, the standard rule of thumb — that every $10 rise in crude widens India's current account deficit by roughly $12 to $15 billion a year — explains why an oil-importing economy feels dollar strength twice over. The current-account arithmetic behind that rule is worked out step by step in the CDS/OTA study material.
The 2013 precedent, and why the comparison is only half fair
The government's own framing invited the comparison, so make it properly.
In September 2013, with the rupee under taper-tantrum pressure, newly appointed Governor Raghuram Rajan opened two windows. The first let banks swap dollars raised through FCNR(B) deposits of three years and above into rupees at a fixed 3.5% a year — roughly 3 percentage points below the market rate of the day. The second let banks swap overseas foreign currency borrowings at about 1% below market. By the close on 30 November 2013, the two together had mopped up about $34 billion, of which the FCNR(B) window alone accounted for $26 billion.
So the 2026 haul of $73 billion in eleven weeks is genuinely larger and faster. But note two things the headline elides. The 2013 scheme ran during a balance-of-payments emergency with reserves under strain; the 2026 version ran with reserves already above $690 billion. And the correct like-for-like comparison of the whole exercise is $73 billion against $34 billion, not against $26 billion — the $26 billion figure is the FCNR(B) sub-total. An honest answer states both numbers and says which is which.
The longer lineage runs back further. Resurgent India Bonds (1998), floated after the post-Pokhran sanctions, raised a little over $4.2 billion at coupons of 7.75% in dollars, 8% in sterling and 6.25% in deutschmarks. India Millennium Deposits (2000) raised about $5.5 billion at 8.5%. Set against those, the 2026 mobilisation is an order of magnitude larger.
FCNR(B): the instrument, and the thing it fixed
FCNR(B) stands for Foreign Currency Non-Resident (Bank). It was introduced on 15 May 1993. Its predecessor, FCNR(A), was closed to fresh deposits in August 1994 and phased out, and the reason is the exam-worthy part: under FCNR(A) the RBI carried the exchange-rate risk, so a depreciating rupee turned depositor gains into a direct central-bank loss. Under FCNR(B), the bank carries that risk. The 2026 facility is therefore best understood as the RBI temporarily and explicitly taking back the hedging cost it deliberately handed to banks in 1993 — a targeted, time-boxed exception, not a reversal of principle.
Distinguish the three non-resident accounts cleanly, because two-statement questions live here. NRE is rupee-denominated and repatriable. NRO is rupee-denominated for income earned in India, with repatriation limits. FCNR(B) is held in foreign currency, so the depositor takes no rupee risk at all — which is exactly why it was the instrument chosen when the objective was to attract dollars rather than rupee savings.
What it did to the reserves
The inflow is visible in the weekly data. For the week ended 14 August 2026, India's foreign exchange reserves stood at $716.907 billion, up $9.905 billion in a single week. The composition breaks down as:
| Component | Amount |
|---|---|
| Foreign Currency Assets (FCA) | $581.851 billion |
| Gold | $111.417 billion |
| Special Drawing Rights (SDRs) | $18.74 billion |
| Reserve Tranche Position in the IMF | $4.899 billion |
| Total | $716.907 billion |
Those four components are the standard answer to "what are forex reserves composed of," and FCA is always the largest. For comparison, reserves at end-March 2026 were about $691.1 billion, providing roughly eleven months of import cover and covering about 90.3% of total external debt. Reserve adequacy is conventionally judged on exactly those two ratios plus short-term debt cover — a point the CDS/OTA current affairs material returns to every quarter.
The impossible trinity, and where India sits
The organising theory is the Mundell-Fleming trilemma, usually called the impossible trinity. An economy can have at most two of these three at once:
- A fixed or tightly managed exchange rate
- Free movement of capital across borders
- An independent domestic monetary policy
The logic is simple. If capital moves freely and you insist on holding the exchange rate, then every attempt to set your own interest rate is arbitraged away by inflows or outflows — you end up importing the anchor country's monetary stance.
India's chosen position is a deliberate middle: a managed float rather than a peg, partial capital account convertibility rather than full, and a genuinely independent monetary policy under the flexible inflation-targeting framework operated by the Monetary Policy Committee. The MPC has six members — three from the RBI, including the Governor as ex-officio chairperson with a casting vote, and three nominated by the Government — and it became operational on 27 June 2016 under the RBI Act, 1934 as amended by the Finance Act, 2016. The inflation target is set by the Centre in consultation with the RBI for a five-year block; the block running to 31 March 2026 fixed 4% CPI inflation with a tolerance band of 2% to 6%.
Read the swap facility through that lens and it stops looking like an ad-hoc rescue. It is a capital-account instrument used precisely so that the monetary-policy instrument does not have to be. Defending a currency with interest rates means raising rates into a slowdown. Defending it by importing long-dated dollars leaves the policy rate free. That is trilemma management, executed deliberately.
The cost side, honestly stated
Three real objections, and they belong in any full-mark answer.
It is debt, not equity. FCNR(B) balances are liabilities owed to non-residents, so they enter India's external debt, which stood at $762.8 billion at end-March 2026, up $26.3 billion year on year, with the external debt-to-GDP ratio rising to 20.8% from 19.8%. The short-term share rose to 19.6% of the total from 18.3%, and short-term debt as a proportion of reserves rose to 21.6% from 20.1%. By currency, the US dollar is 55.5% of the stock, the rupee 29.4%, the yen 6.4%, SDRs 4.3% and the euro 3.7%. The 2013 window had exactly this effect on the debt stock, and the 2026 one will too. Reserves and external debt both rise — the net external position improves by far less than the reserves headline suggests.
It has to be repaid. Three-year-plus money means a redemption cliff, and one large enough to require planning. Managing the 2016-17 maturity of the 2013 FCNR(B) cohort was itself a significant operation. A $65.40 billion cohort creates a correspondingly larger one.
Someone pays for the subsidy. Absorbing 280 to 300 basis points of hedging cost on tens of billions of dollars is a real charge on the central bank's forward book, settled over the life of each contract. It is cheaper than the alternatives, and the RBI has said as much implicitly by choosing it — but "cost-efficient" is a comparison, not the absence of a cost.
Against all three: the money is long-dated and on tap, it arrived without a rate hike, it strengthened the buffer before it was needed rather than during a run, and it was closed early once the target was met. Buying insurance while it is cheap is a defensible use of a central bank's balance sheet. That two-sided reading is the answer a good examiner is looking for.
🔑 Revision block
The announcement. Ministry of Finance, 24 August 2026 → RBI's special USD-INR swap facility for FCNR(B) + OFCB + ECB raised $73 billion as on 21 August 2026 · FCNR(B) alone $65.40 billion · launched 8 June 2026 (announced 5 June) · under 11 weeks · FCNR(B) window closure advanced from 30 September to 31 August 2026.
The mechanism. RBI absorbs the hedging cost of roughly 280–300 bps on 3–5 year money → banks quote 5.5%–7.1% on dollar deposits → non-resident dollars flow in. Buy/sell = RBI buys USD spot, sells forward → injects rupee liquidity. Sell/buy = the mirror → absorbs it.
The precedent. September 2013, Governor Raghuram Rajan → FCNR(B) swap at a fixed 3.5%, about 3 points below market, plus a bank-borrowing window at 1% below market → $34 billion by 30 November 2013, of which FCNR(B) was $26 billion. Earlier: Resurgent India Bonds 1998, $4.2 bn · India Millennium Deposits 2000, $5.5 bn.
The instrument. FCNR(B) introduced 15 May 1993; FCNR(A) closed to fresh deposits August 1994 because the RBI bore the exchange risk under it — under FCNR(B) the bank does.
Figures to carry. Reserves $716.907 billion, week ended 14 August 2026 → FCA $581.851 bn · gold $111.417 bn · SDRs $18.74 bn · IMF reserve tranche $4.899 bn · ~11 months import cover · external debt $762.8 billion, 20.8% of GDP, short-term share 19.6%.
The theory. Impossible trinity → fixed rate + free capital + independent monetary policy: pick two. India runs a managed float + partial convertibility + independent MPC — a capital-account tool used so the policy rate stays free.
The trap. $73 billion vs $34 billion is the fair 2026-versus-2013 comparison; $26 billion is only the 2013 FCNR(B) sub-total.
The two-sided line, for essay and GD. Buffers built early, at long tenor, without a rate hike — but every dollar of it is external debt with a three-year redemption cliff and a subsidised hedge sitting on the central bank's forward book.
🎯 Practice MCQs
1. In a USD/INR buy-sell swap conducted by the Reserve Bank of India, the immediate effect on domestic liquidity is that: (a) Rupee liquidity is injected into the banking system (b) Rupee liquidity is absorbed from the banking system (c) Liquidity is unaffected, since both legs settle simultaneously (d) Dollar liquidity is absorbed and rupee liquidity is unchanged
→ (a) — the RBI buys dollars spot and pays out rupees, injecting liquidity; the reverse forward leg withdraws it at maturity.
2. Which of the following is not a component of India's foreign exchange reserves as reported by the RBI? (a) Gold (b) Special Drawing Rights (c) Reserve Tranche Position in the IMF (d) External Commercial Borrowings
→ (d) — ECBs are a liability, not a reserve asset; the four components are FCA, gold, SDRs and the IMF reserve tranche position.
3. The FCNR(A) scheme was replaced by FCNR(B) principally because under FCNR(A): (a) Deposits were denominated in rupees (b) The exchange-rate risk was borne by the Reserve Bank of India (c) Interest rates were linked to the repo rate (d) Only resident Indians could open such accounts
→ (b) — under FCNR(A) the RBI carried the currency risk; FCNR(B), introduced in May 1993, shifted it to the banks.
4. The "impossible trinity" in international economics holds that a country cannot simultaneously have: (a) Low inflation, high growth and full employment (b) A fixed exchange rate, free capital mobility and an independent monetary policy (c) Fiscal deficit control, low interest rates and currency stability (d) A current account surplus, a capital account surplus and rising reserves
→ (b) — the Mundell-Fleming trilemma; at most two of the three can be held at once.
5. Foreign Currency Non-Resident (Bank) deposits raised under the RBI's swap facility are classified in India's balance of payments and debt statistics as: (a) Foreign direct investment (b) Foreign portfolio investment (c) Part of India's external debt (d) A component of foreign exchange reserves only, with no liability effect
→ (c) — NRI deposits are liabilities to non-residents and form part of external debt, which was $762.8 billion at end-March 2026.
📋 How this gets asked (PYQ pattern)
External-sector economics is a dependable scorer in the CDS/OTA General Knowledge paper, and this story sits on four of its most recurrent angles.
The composition-of-reserves item is the most frequent of all. It appears as a straight "which of the following is not a component" question, or as a matching item, and the distractors are almost always liabilities dressed as assets — external commercial borrowings, NRI deposits, FDI stock. If you know the four components and that foreign currency assets are always the largest, the whole family collapses.
The instrument-identification item is the second. NRE, NRO and FCNR(B) get shuffled and the question turns on which one is denominated in foreign currency and who bears the exchange risk. That single discriminator resolves most versions.
The trilemma item recurs in both direct and applied form: name the three legs, or identify which leg a described policy sacrifices. The applied version is harder and increasingly common, and this scheme is a clean worked example of choosing capital-account action over monetary-policy action.
The institution-and-date item covers the RBI's own architecture — the MPC's six-member composition and its 3-plus-3 split, the Governor's casting vote, and the statutory route through the RBI Act, 1934 as amended in 2016. These are static facts that a current-affairs story merely gives you an excuse to revise.
The fresh hook this cycle is a comparison question — 2026 against 2013 — and the safest way to carry it is as a pair of matched numbers with their scopes attached: $73 billion across all three windows in 2026, against $34 billion across two windows in 2013, of which $26 billion was FCNR(B). As always, learn the recurring shape rather than trusting any specific numbered question from a past paper; the pattern is what repeats.
One story a day, worked through the same way — news, then the mechanism, then the questions it can generate — runs at /cds-ota-current-affairs. If a classroom, a fixed schedule and someone marking your written answers would suit you better than reading alone, the upcoming batches are listed at /upcoming-courses-cavalier-delhi.