A working model of what happens to gold, silver and the Indian economy across the rupee's whole plausible range — built on actual FY26 data, and on the point that the same exchange rate means three completely different things depending on why it got there.
Most rupee commentary asks where it is going. The right question is why it is moving — the same exchange rate produces completely different outcomes depending on the answer.
The rupee is not an independent variable. It sits at the end of a chain that starts with the US Treasury's funding decisions, and every link in that chain is measurable today.
If the US were monetising its debt, breakevens would be at 3.5–4% and the dollar would be falling against everything. They are at 2.30%, and the dollar is up 1.7% over twelve months with futures pricing a 56% chance of a September hike. What the US is actually exporting is a high real interest rate.
For India that is the worse of the two shocks. Dollar debasement would lift Indian gold and leave the rupee roughly stable against a falling dollar. A real-rate shock does the opposite: it pulls capital out of India, pushes the rupee down, and pushes gold down in dollars, so the gold hedge fails at exactly the moment the rupee needs it. That is Regime A, and it is the live one.
This is the single most counterintuitive result in the study, and the historical record is unambiguous about it.
The intuitive chain runs: America's finances deteriorate → the dollar collapses → the rupee strengthens → imports get cheaper. Every step of that is wrong in the acute phase of a crisis, because of what the dollar is. Roughly 88% of all foreign-exchange transactions have a dollar on one side, and the world's debts are written in it. When funding stress hits, everyone needs dollars more, not less — and they sell whatever they own to get them.
In 2008 the rupee fell 19.22% against the dollar. That was a crisis whose epicentre was a few blocks of Manhattan, and the currency of the country at fault appreciated. Foreign investors pulled $9.3bn from Indian assets that year. In March 2020 the same thing happened, and gold fell in the first fortnight before it rose.
The two-phase path
A US funding crisis is not one trade. It is two, in sequence, and they point opposite ways.
Holding gold in rupees as insurance against a US debt crisis does not pay out when the crisis arrives. It pays out afterwards. In January 2026 gold fell 21% from its record and silver fell about 30% in a single day — on leveraged-ETF liquidation and CME margin hikes — while every word of the debasement thesis remained true.
Indian households hold roughly 25,000 tonnes. A repeat of that January move is a paper loss of around ₹70 lakh crore in a fortnight. The position has to be sized to survive Phase 1 in order to collect in Phase 2. That is the whole discipline.
There is no mystery in the domestic gold price. It is an arbitrage, and it ties out to within two-thirds of a percent of the actual MCX quote.
Gold, rupees per 10 grams — wholesale (MCX) basis
International spot ÷ 31.1035 × 10 × USD/INR × (1 + 15% customs duty) × basis
Gold in rupees, across the full range
The three regimes diverge violently at the same exchange rate
Read the gaps, not the lines. At an identical USD/INR of 140, gold is ₹3.89 lakh under debasement, ₹1.41 lakh under real-rate dollar strength, and ₹1.01 lakh in an acute US funding crisis — a 3.9× spread. Note that regime D is the only line that slopes down: a rupee at 140 driven by a dollar scramble leaves Indian gold below today's ₹1.56 lakh. That is the Phase 1 problem, drawn.
Under Regime A the gold line is almost flat. If the rupee falls purely because the dollar is strong everywhere, gold falls in dollars by roughly as much as the rupee falls against the dollar, and the two effects cancel. Indian gold barely moves.
This is why “the rupee is going to 130, buy gold” is an incomplete thought. Gold only protects you in rupee terms if the rupee is falling for India-specific reasons, or if the dollar is being debased against real assets at the same time. Against a genuinely strong dollar, gold is not a rupee hedge at all.
Household gold: the balance sheet nobody puts on the balance sheet
Indian households and temples hold about 25,000 tonnes — the largest private stock on earth
India imports 88.7% of its crude and prices it in dollars. That is the main channel, and it is why the RBI cares about the rupee far more than its inflation-targeting mandate alone would suggest.
Inflation
Pass-through elasticity 0.12 — a 10% fall in the rupee adds ~1.2pp to CPI
Crude oil
Brent $87.30 · run-rate from Q1 FY27 actuals
—
India runs a $333bn goods deficit and a $214bn services surplus, and receives $135bn of remittances. A weaker rupee does not simply hurt — more than half the external account is on the earning side of the dollar.
Balance of payments, $ billion — FY26 actuals repriced
Volume elasticities applied: imports −0.30, exports +0.40, services +0.15, remittances +0.20
Reserves and external debt
$691bn reserves · $746bn external debt, ~54% dollar-denominated
The economy in dollar terms
Rupee GDP held constant — this is pure translation
Decompose the index by sector and something unexpected falls out: at the earnings level, the Nifty is almost exactly currency-neutral. IT's gain cancels the importers' loss. All the damage happens in the price-to-earnings ratio.
Where a 1% rupee depreciation lands, sector by sector
Weight × earnings sensitivity = contribution to Nifty EPS
| Sector | Weight | EPS beta | Contribution |
|---|---|---|---|
| Information technology | 13.8% | +1.75% | +24.2 bp |
| Pharma | 4.0% | +1.00% | +4.0 bp |
| Metals | 4.0% | +0.50% | +2.0 bp |
| Financial services | 33.5% | −0.20% | −6.7 bp |
| Oil, gas & fuels | 12.1% | −0.50% | −6.1 bp |
| FMCG | 8.0% | −0.40% | −3.2 bp |
| Autos | 6.7% | −0.30% | −2.0 bp |
| Telecom | 5.2% | −0.30% | −1.6 bp |
| Everything else | 12.7% | −0.10% | −1.3 bp |
| Weighted Nifty EPS | 100% | — | +9.4 bp |
A 10% rupee fall moves Nifty earnings by less than 1%. The index is a currency hedge that accidentally nets to zero.
Nifty 50 — and the gap between rupee and dollar returns
Base 24,334 at PE 20.42 · 10-year median PE 23.36
—
Why the rupee-terms fall is survivable: who owns the market now
The buyer of last resort changed, and it changed recently
Domestic money outbought foreign selling nearly five to one in FY26. This is the single biggest structural change in the Indian market in a decade, and it is why the model compresses the multiple far less than a 2013-style episode would have.
The Nifty trades at a PE of 20.42 against a ten-year median of 23.36 — about 13% below its own history. The S&P 500 is at a CAPE of 42.5, a level exceeded only in 2000, with the top ten names at ~40% of the index. India's Buffett indicator is 132%, off its 141% high and below the US, Japan, Korea and Taiwan.
So the asymmetry runs opposite to the currency story: the rupee is the vulnerable variable, but Indian equities are the cheaper asset. A global de-rating starts from a much higher place in New York than in Mumbai.
Indian residential prices are famously sticky in nominal rupees. What actually moves is the mortgage payment, the transaction volume, and the real return — and the channel runs through the RBI's reaction to imported inflation, not through the exchange rate directly.
₹1 crore property · ₹50 lakh loan · 20 years
Base home-loan rate 7.10% · RBI reaction assumed at 0.6× the CPI move, capped at ±3pp
The NRI arbitrage
What a ₹1 crore flat costs a dollar earner
Nominal versus real
Where the adjustment actually happens
—
Developers and listing portals quote Indian house-price growth at 9–12% for 2026, and 8–24% across the top seven cities. The RBI's own House Price Index has prices up 4.2% in early 2026. That is not a rounding difference — it is the gap between asking prices on unsold inventory and transactions that actually cleared.
Every real-return number in this section is computed off the base rate you believe. At 4.2% nominal growth against 4.45% inflation, Indian residential property is already a slightly negative real asset before the rupee moves at all. That single fact matters more than anything the slider does.
A currency move is a transfer, not a loss. At the level currently on the slider, here is roughly where it goes.
Computed under the regime currently selected. Switch regimes above and this table recomputes.
Gold and silver are wholesale. Retail adds 3% GST, and jewellery adds making charges on top.
If you arrived here worried about US debt, it is worth seeing the two side by side, because the comparison does not run the way most people assume — in either direction.
| Measure | United States | India | Reads better for |
|---|---|---|---|
| Government debt / GDP | 101% held by public | 84% general govt | India |
| Fiscal deficit / GDP | 5.8% | 4.4% | India |
| Interest / tax revenue | ~19% | ~43% | US, by a wide margin |
| External debt / GDP | ~30%+ | 19.2% | India |
| Current account | Persistent deficit | −0.6% of GDP | India |
| Reserves | n/a — issues the reserve asset | $691bn · 10.7 months | Different games |
| Debt currency | Own — and the world's | Own, and domestically held | Both protected |
| Core vulnerability | Rollover at a rising real rate | 88.7% oil import dependence | — |
India spends roughly 43 paise of every rupee of central tax revenue on interest — ₹11.6 lakh crore against ₹26.7 lakh crore of net tax receipts, and the single largest line in the Union budget. The equivalent US figure is about 19%. On the metric that actually measures fiscal strain, India is more than twice as stretched as the country most people worry about.
The reason it is not a crisis is the part that matters: India's debt is in rupees and held at home, largely by domestic banks, insurers and provident funds. A country cannot have a currency crisis over debt denominated in its own currency and owned by its own savers. What it can have is a fiscal space problem — and India's shows up as a low tax take, not as default risk. The external account, which is where currency crises actually come from, is in good order: a 0.6% deficit and nearly eleven months of import cover.
So the honest read is: India is far better insulated than the ₹140 end of this slider implies, and the route there runs through oil and capital flows, not through insolvency.
Ranked by how much each would change the answer. The first three are American; they reach the rupee before any Indian data does.
| Signal | Now | Trips at | What it would mean for India |
|---|---|---|---|
| US 10y breakeven | ~2.30% | >2.75% | Regime flips from A to C. Gold becomes a real rupee hedge; the dollar starts falling. Counterintuitively good for the rupee. |
| India–US 10y spread | 219 bp | <200 bp | Foreign demand for Indian debt historically dries up below this. The most direct, most immediate link on the board. |
| US 30y yield | 5.21% | >5.50% with DXY falling | The bad correlation. Yields up and dollar down together is a genuine loss of confidence, and pushes toward Phase 2. |
| Brent crude | $87.30 | >$110 | India's own trigger. A $20 move is roughly $45bn on the import bill and lands straight on CPI and the rupee. |
| RBI reserves | $691bn · 10.7 mo | <$600bn or <8 months | The defence budget running down. The RBI has been spending it near-daily around 95.60–95.80. |
| Bullion import duty | 15% | Any further rise | A tell, not a cause. It was raised from 6% in May 2026 to defend the rupee; another hike says the RBI is losing the argument. |
The live path is Regime A, and it is the one where gold does not protect you. US real yields at a 17-year high with anchored breakevens pull capital toward the dollar; the India–US spread is 19bp from the level where foreign bond demand historically stops; FPI equity outflows are already the heaviest since 1993. That combination pushes USD/INR up while pushing gold down in dollars, and the two cancel in rupee terms.
Rupee gold pays off in Regime C, and Regime C has not started. The trigger is a US breakeven above 2.75%, and it is at 2.30%. If it goes, the whole picture inverts: the dollar falls, gold rises in dollars, and rupee gold rises on both legs at once. That is the scenario worth owning gold for — and it is not today's.
And a US crisis gets you there via Phase 1, not around it. Anyone holding leveraged metal through the acute phase gets liquidated before the thesis pays. January 2026 was the rehearsal: every word of the debasement argument stayed true while gold fell 21% and silver fell 30% in a day.
The pass-through elasticity of 0.12 is the softest number here. I could not find a single current published RBI figure for it, so it is set from the general literature range of 0.08–0.18 for India over a twelve-month horizon. If the true value is 0.18, every inflation number on this page is 50% too low. Treat CPI outputs as an order of magnitude, not a forecast.
The trade elasticities are conventional textbook values, not estimated from Indian data. The regime betas — how gold in dollars responds — are stipulated, not fitted. They are there to show you the shape of the dependency, which is the real finding, not to predict a price.
What the model does well. The gold and silver arithmetic is an arbitrage identity, not a forecast, and it ties to the observed MCX quote within 0.62%. The balance-of-payments repricing uses actual FY26 line items. The translation effects — dollar GDP, external debt servicing, import cover, IT margins — are definitional and hold exactly.
What it ignores. Second-round inflation effects and wage responses. RBI intervention, which has been near-daily around 95.60–95.80 and would not disappear at 120. Capital-account behaviour: at 125+, the model assumes portfolio flows behave linearly when in practice they gap. Domestic fiscal response. Any change to the 15% bullion duty — which the government raised from 6% in May 2026 precisely to defend the rupee, and would likely move again in the scenarios at the far end of this slider.
The current-account line breaks down at the extremes, and you should not read it literally past about ₹110. The model applies fixed trade elasticities linearly across the whole range, so at ₹140 it produces a current-account surplus of 4.4% of GDP. India has essentially never run a surplus of that size. What the arithmetic is really saying is that import compression on that scale would require a recession, and that export elasticities do not hold over a 46% move. Read the direction — a weaker rupee narrows the deficit, a stronger one widens it — and treat the far columns as showing the force of the correction, not its landing point.
On the regime betas, now that they carry more weight. Regime A's −0.83 is built from a DXY–INR relationship whose measured correlation is only −0.44 on monthly averages — real, but far from deterministic, and the rupee routinely moves on domestic factors while DXY does nothing. Regime D's −1.20 is calibrated to the January 2026 episode rather than estimated across crises. Treat the four regimes as bracketing the space, not as fitted models.
On the ₹70 end. Treat it as a stress test in the other direction rather than a forecast. Getting there needs a sustained collapse in the dollar combined with a large Indian productivity or capital-inflow surge. It is not impossible — the rupee was at 70 as recently as 2018 — but nothing in the current data points that way, and the model is a repricing engine, not a probability statement.