Macro · Research note
The Rupee, the Outflow, and What the Reserve Bank Actually Controls
Foreign investors pulled more from Indian equities in seven months than in all of 2025, and the rupee hit a record low. Almost none of the cause was decided in India.
The rupee reached 96.84 to the dollar on 20 May 2026, a record low, after starting the year near 89.86. Foreign portfolio investors withdrew roughly ₹2.29 lakh crore from Indian equities in the first seven months, more than the ₹1.66 lakh crore they took out across the whole of 2025, and foreign ownership of Indian equities fell to a fourteen year low.
The natural reading of that sequence is that something went wrong in India. This piece argues that almost nothing did, and that the useful exercise is separating the variables the Reserve Bank can act on from the ones it can only absorb. The distinction matters because the policy response to an imported shock and the response to a domestic one are different, and confusing them wastes reserves.
The move
Figure 1 traces the year: the currency weakened by roughly 7% from January to the May low, then recovered part of that move to around 94.35 by the end of the first half. A rising line means a weaker rupee.
Figure 2 shows the outflow, which is the larger number in proportional terms. Seven months of 2026 have exceeded twelve months of 2025, and the fourteen year low in foreign ownership indicates this is a level shift rather than a rotation between sectors.
Three causes, none of them Indian
The proximate drivers are identifiable and they are all external.
The first is oil: India imports the overwhelming majority of the crude it consumes, so the Hormuz disruption arrived directly in the import bill. Brent went from roughly $72 a barrel on 27 February to about $120 at its March peak, fell back after the June memorandum, and rose again when the United States reimposed its naval blockade on 12 July. A country importing that volume experiences an oil price spike as a current account shock and a currency shock simultaneously.
The second is trade policy: a 50% US tariff overhang dating from mid-2025 has weighed on Indian export competitiveness, which affects the same external balance from the other side.
The third is the global price of duration: Japanese long yields repriced in January, the 40-year JGB broke 4% for the first time, and Japanese investors sold $29.6 billion of US debt in the first quarter. A carry unwind of that kind withdraws capital from high-yielding emerging markets first, and it does so without reference to their individual merits.
| Driver | Origin | Effect on India | Can the RBI address it |
|---|---|---|---|
| Crude import bill | Strait of Hormuz disruption | Current account, imported inflation | No, only cushion the currency effect |
| US tariff overhang | Washington, mid-2025 | Export competitiveness | No |
| Global carry unwind | Tokyo, January 2026 | Portfolio outflows | No |
| Domestic liquidity and rates | Mumbai | Bond yields, credit conditions | Yes |
| Currency volatility | Market | Speed of adjustment | Partly, through intervention |
| Capital account rules | Mumbai and New Delhi | Composition of flows | Yes, but slowly |
Attribution of the 2026 rupee move follows contemporaneous reporting. Sources: Univest, Finnovate, Business Standard, Under the Market Lens.
Figure 3 sorts these by what policy can reach, and the bottom-right cell holds most of what actually moved the currency this year, none of which is decided in Mumbai.
What intervention buys, and what it does not
Figure 4 gives the two numbers that matter. The Reserve Bank is estimated to have sold about $55 billion defending the rupee, against reserves of roughly $682 billion.
Note what that ratio does and does not tell you: India’s reserve position remains among the largest in the world, and $55 billion is well within what it can absorb, so the constraint here is not adequacy.
The constraint is purpose: intervention against a shock originating in the oil price and in Tokyo cannot restore the level, because the underlying flows continue regardless of the central bank’s participation. What it can do is control the speed of the adjustment, which is a real and defensible objective: a disorderly move triggers stop-losses, unhedged corporate borrowers face sudden distress, and the adjustment overshoots. Slowing it prevents that.
Therefore the correct standard for judging the intervention is not whether the rupee reached a record low, because it plainly did. The standard is whether the path there was orderly enough to avoid forced selling, and the recovery from 96.84 to around 94.35 suggests the market found a level rather than a floor imposed by the central bank.
The case for resilience
Three features distinguish India from the emerging markets that usually break in this environment.
The domestic institutional base is the most important: systematic investment plan flows into mutual funds have created a large and price-insensitive domestic bid for Indian equities, which is why foreign ownership could fall to a fourteen year low without the market collapsing. A foreign seller needs a buyer, and in India there was one.
Real yields are positive, which means the central bank is not choosing between defending the currency and supporting growth in the way it would be with inflation running ahead of the policy rate. The reserve stock provides room to smooth without approaching any threshold that would invite speculation.
And the rupee is not expensive in real effective terms, which matters for what comes next. A currency that weakens from an overvalued starting point often keeps weakening; one that weakens from fair value tends to find buyers.
Where this argument is weak
The resilience case has an uncomfortable feature: it is the same case that was made about every emerging market that later had a problem. Large reserves, a growing domestic investor base and sound fundamentals were all true of several countries shortly before they were not. The argument is a statement about the current balance of pressures, and it does not carry information about what happens if oil stays above $100 for a further year.
Furthermore, the attribution of the move to external causes is cleaner in this piece than the evidence supports. Domestic factors including the growth outlook, the fiscal path and equity valuations after a long run all plausibly contributed to the foreign selling, and separating them from the carry unwind is not possible with the flow data available. The claim that “almost nothing went wrong in India” is a directional judgement, not a decomposition.
Finally, the domestic bid that absorbed foreign selling has not been tested through a drawdown, since systematic investment plan flows have grown throughout a period of rising markets. Whether they persist through a sustained fall is unknown, and if they do not, the sequence described here looks very different.
Conclusion
India in 2026 experienced a currency at a record low and the largest foreign equity outflow on its recent record, driven by an oil shock originating in the Gulf, a tariff regime set in Washington, and a duration repricing that started in Tokyo.
The Reserve Bank spent about $55 billion of a $682 billion reserve stock slowing the adjustment, which is what that stock is for. It did not defend a level, and it was right not to; none of the three drivers responds to Indian policy.
The instructive part is the asymmetry: a central bank in this position controls the speed of an adjustment and the composition of the flows, and it controls neither the direction nor the destination. Reading a record low as a policy failure misidentifies who made the decisions that produced it, and the practical consequence of that misreading is spending reserves on a level that was never available to buy.