FCNR Deposits and India’s Reserves: When Insurance Becomes Expensive

A foreign-currency mobilisation programme buys insurance and creates obligations at the same time. India’s latest public numbers make that balance more important than a headline reserve target. The 2 September source model’s $90–120 billion stopping region is now a dated scenario, not a policy recommendation to carry unchanged into October.

What changed by 8 October 2026

RBI’s 21 September release reports $132.980 billion of FCNR(B) inflows mobilised through 31 August. Including $5.320 billion of overseas foreign-currency borrowing and $5.296 billion of ECB inflows reported to 18 September, the facility total was $143.596 billion. These are different borrowing categories, not all FCNR deposits. The FCNR window ended on 31 August; the other windows were stated to run through 31 December.

The 7 October Governor’s statement puts reserves at $734.6 billion on 2 October, around 11 months of import cover, and reports September average daily LAF absorption of ₹7.8 lakh crore. The source’s September liquidity forecast has therefore been replaced by an observed official figure. The policy question has shifted from how much more to attract through the closed deposit window to how to manage liquidity, maturity concentration and renewed external stress.

RBI’s June-quarter debt release reports total external debt of $778.2 billion and $337.5 billion falling due within a year on a residual-maturity basis. These are June stocks. Dividing $337.5 billion by October reserves gives 45.94%, a mixed-date monitoring ratio—not an official same-date adequacy statistic. For the same reason, the Governor’s total external-debt cover and this short-term-debt ratio must not be confused.

Gross borrowing is not an estimate of the programme’s causal reserve addition. Valuation, intervention and unrelated flows move the reserve stock. A new optimal stopping point cannot be estimated honestly from these totals alone: the marginal hedge subsidy, sterilisation cost, forward book and maturity ladder are still required.

The decision is about marginal insurance

Let $Q$ denote gross foreign-currency deposit mobilisation, $B(Q)$ the resulting economic benefit and $C(Q)$ the associated cost. The policy objective is to maximise net welfare:

$$W(Q)=B(Q)-C(Q).$$

If the functions are sufficiently smooth and the optimum is interior, the stopping condition is

$$B'(Q^*)=C'(Q^*).$$

Here $Q^*$ is the chosen mobilisation level and the prime denotes the change generated by one additional unit. This is an organising principle, not evidence that either curve has actually been estimated. The archived model uses reserve-adequacy proxies to locate a plausible region rather than fitting a complete policy loss function.

The first dollars can protect against disorderly currency adjustment, reassure external creditors and expand intervention flexibility. Later dollars still provide protection, but the improvement becomes smaller when reserves already cover plausible financing needs. Meanwhile, the authorities may continue paying for hedging support, absorbing domestic liquidity and accumulating concentrated future maturities.

Total benefit may therefore remain positive beyond the point at which further expansion becomes unattractive. Comparing the total reserve gain with the total programme cost would miss that distinction.

Figure 01 · Conceptual framework

From foreign-currency inflow to reserve protection

How the parts connect

1. AFCNR mobilisationinflowForeign-currency use
2. BForeign-currency usepossible accumulationReserve buffer
3. COther reserve movementschanges stockReserve buffer
4. DForeign-currency userupee counterpartDomestic liquidity
5. EFCNR mobilisationliability createdFuture repayment

The numerical starting point

The archived analysis used the following inputs. They are retained at their original stated precision so that the calculation can be reproduced. Their inclusion does not independently verify the underlying economic releases.

Historical inputs retained at their stated precision

The debt measure concerns remaining maturity, which is economically different from debt originally issued with a maturity below one year. A long-dated liability nearing repayment can matter just as much as a short-dated instrument when assessing the next year’s financing pressure.

Merchandise imports are also a deliberately limited denominator. They do not describe all foreign-currency receipts or all external payments. Services earnings, remittances, investment flows and intervention obligations can materially change the actual liquidity picture. The useful question is what each indicator reveals, not whether one ratio can stand in for the entire balance of payments.

Reconstructing the import-cover test

If $R$ is reserves and $M$ monthly merchandise imports, import cover in months is

$$I=\frac{R}{M}.$$

At the archived opening inputs, cover was approximately $681.6/64.6=10.55$ months. Selecting a target $I^*$ implies required reserves $R^*=I^*M$. A target is a policy assumption: it is not a universal rule that every economy should hold that number of months.

Import-cover objectives and required reserve additions

The original note rounded these to roughly $743 billion and $775 billion, requiring about $61 billion and $93 billion respectively. The small differences above arise from using its rounded monthly denominator. They are not changes in the underlying policy claim.

The calculation establishes scale. It does not show how much of the import bill would need financing from reserves during a crisis, whether payment compression would occur, or how quickly other receipts would fall. A more complete stress exercise would model those cash flows jointly.

A second test: near-term external debt

Let $D_{12}$ be debt due within twelve months and $d=D_{12}/R$ the debt-to-reserves ratio. Holding the debt stock fixed, a target ratio $d^*$ requires

$$R^*=\frac{D_{12}}{d^*}.$$

Using $326.9 billion of debt produces the following approximate requirements:

Debt-to-reserves objectives and required reserve additions

This overlaps with the import-cover calculation. Both suggest that net additions of roughly $60–90 billion could move the opening position towards the selected protection bands. The convergence is useful because the indicators describe different financing exposures.

It should not be treated as two statistically independent confirmations of an optimal number. Both depend on the same reserve numerator and on thresholds chosen by the analyst. The exercise is a transparent consistency check, not a confidence interval derived from an estimated probability model.

Figure 02 · Qualitative comparison

Two reserve tests, different questions

Two reserve tests, different questions

Test	What it asks	What can mislead
Import cover	What it asksHow many months of the chosen import measure could reserves cover?	What can misleadA partial import denominator omits other external cash flows.
Residual-maturity debt	What it asksHow large are repayments due within the next year relative to reserves?	What can misleadOriginal maturity excludes long-term debt nearing repayment.
Joint stress review	What it asksWhat happens if imports rise while financing weakens?	What can misleadMixing observation dates creates a monitoring ratio, not an official same-date statistic.

Gross deposits are not net reserve additions

Gross mobilisation and persistent reserve accumulation can differ. Deposit inflows can coincide with currency intervention, valuation changes, other capital flows and changes in gold or other reserve assets. A central bank can attract deposits while using some foreign currency to stabilise markets. Headline reserves may then rise by less than gross programme receipts without indicating that the remaining funds were lost.

The archived calculation compared a $47.72 billion reserve increase with $72.85 billion of reported mobilisation:

$$\lambda_{\mathrm{proxy}}=\frac{47.72}{72.85}\approx 65.5\%.$$

The original note used approximately 65%. This number is a crude sizing proxy. It does not isolate the causal effect of the deposit programme because the reserve movement includes unrelated changes.

If $\Delta R^*$ is the desired net reserve addition and $\lambda$ an assumed persistent pass-through, gross mobilisation is

$$Q^*=\frac{\Delta R^*}{\lambda}.$$

Gross mobilisation under alternative pass-through assumptions

This table explains why a single mobilisation target would overstate precision. The same reserve objective can imply materially different gross receipts depending on how much remains in reserves. The wider mechanical range is approximately $75–138 billion; the archived $90–120 billion central band is an analytical narrowing within that range, not a separate algebraic identity.

Four costs that accumulate with scale

The four costs of continued mobilisation

These costs cannot be inferred from gross mobilisation alone. They require a maturity ladder, the actual incentive terms, intervention information and the cost of liquidity absorption. The stopping range should remain conditional until those elements are available.

Figure 03 · Conceptual framework

A disciplined mobilisation decision

How the parts connect

1. ADefine the shockstress scopeMeasure protection
2. BMeasure protectionmarginal benefitCompare at the margin
3. CPrice full costsmarginal costCompare at the margin
4. DCompare at the marginbenefit exceeds costContinue conditionally
5. ECompare at the margincost is not justifiedStop and monitor

How an oil or currency shock changes the answer

A deterioration in the external environment can increase the marginal value of reserves. Higher oil costs raise the foreign-currency import bill; weaker capital inflows can reduce financing availability; a sharper exchange-rate adjustment can increase demand for protection. These channels can operate together.

The archived stress case referred to oil above $90 per barrel, a weaker rupee or a severe geopolitical disturbance. It argued that these conditions could move the preferred gross range towards $120–140 billion. This was a conditional extension of the model, not a prediction that such mobilisation would occur or that the specified oil price was itself an optimal trigger.

A useful stress test changes both sides of the balance sheet. Higher imports worsen the cover ratio even if reserves remain constant. Larger near-term repayments can weaken the debt test. Increased intervention needs may also lower pass-through. Simply adding an arbitrary reserve cushion while holding every other input unchanged would understate the interactions.

A disciplined policy review

A review of the programme should ask five connected questions. First, what external shock is the insurance meant to cover? Second, how much persistent reserve protection has actually been obtained? Third, what is the additional incentive and liquidity cost of continuing? Fourth, what repayment concentration is being created? Fifth, what observable change would justify reopening or extending the programme?

The review should preserve a baseline and a stress case. It should not continually redefine the target so that every actual outcome appears optimal. In particular, a later shock does not automatically prove an earlier closure was wrong; the relevant comparison is the information and alternatives available at the decision date.

The archived conclusion favoured closure near $100 billion under its assumptions. Actual FCNR mobilisation subsequently recorded at $132.980 billion exceeded its central band. That does not establish policy error: the model did not estimate a full welfare function, and the external environment changed. Its severe-stress band was conditional, not a retrospective claim of forecasting the actual outcome. The lasting contribution is the framework: temporary reserve insurance should end when its marginal protection no longer compensates for its full monetary and refinancing cost.

Source record and update scope

• Reserve Bank of India — FCNR $132.980 bn, OFCB $5.320 bn, ECB $5.296 bn; total $143.596 bn; facility window dates. Observation/publication: 2026-09-21. Verified primary source; observed date stated.

• Reserve Bank of India — Reserves $734.6 bn on Oct 2; around 11 months import cover; September average LAF absorption ₹7.8 lakh crore. Observation/publication: 2026-10-07. Verified primary source; observed date stated.

• Reserve Bank of India — June total debt $778.2 bn; residual maturity under one year $337.5 bn. Observation/publication: 2026-06-30. Verified primary source; observed date stated.