The September 2026 Market Rout: Why This Sell-Off Is Different, What It Could Cost, and How Policymakers Should Respond

Markets did not fall off a cliff on a single day this September. Instead, several pressures that had been building for months began to feed into one another. On 28th September, after Washington rejected Iran's proposal to reopen the Strait of Hormuz, Brent crude jumped close to 4% to about $108 a barrel. It had been below $100 only a week earlier. The same day, the yield on the 10-year US Treasury note rose to about 5.2%, its highest level since mid-2007, and the 30-year yield reached roughly 5.5%, a level last seen in 2004. Indian equities took the next blow. The Sensex fell more than 1,100 points on Monday and slipped again on Tuesday, touching an intraday low of 72,064. That is 16.4% below its December 2025 peak and the index's lowest level in 27 months. Foreign portfolio investors sold a net ₹5,353 crore of Indian shares in a single session, the rupee weakened past ₹96 to the dollar, and India's volatility index rose to about 14.4. Early in the month, a sell-off in AI-linked technology shares had already erased more than $1 trillion of market value worldwide, and South Korea's Kospi fell nearly 4% in one session.

Whether this counts as a "crash" depends on where you look. For a reader in Mumbai, Seoul or São Paulo it certainly feels like one. For an investor holding only US large-cap shares, it hardly registers yet: the S&P 500 was still within a few per cent of its 13th August record in mid-September. That split is the most useful starting point for understanding the episode. The epicentre is the global bond market and the price of oil. Equities, especially in emerging markets that rely on foreign capital and imported energy, are where the damage becomes visible.

How a sell-off turns into an economic problem

A falling market becomes an economic problem only when it changes what households, firms and governments do. There are five main channels, and in September 2026 several of them are active at once.

The first is the wealth channel. When share prices fall, households feel poorer and spend a little less. Estimates for the US have typically put the effect at a few cents of lost consumption for every dollar of lost stock-market wealth. That sounds small, but it adds up when trillions of dollars are wiped off. It matters more this time because US consumer sentiment fell to 48.1 in September, the lowest reading in the survey's history, before most of the equity damage had occurred.

The second is the cost-of-capital channel, and here the damage is already severe. A 10-year Treasury yield of 5.2% sets the floor for mortgage rates, corporate bond yields and emerging-market sovereign borrowing costs around the world. India's own 10-year government bond yield has risen to about 7.16%. When the risk-free rate climbs this far, every future cash flow is worth less today. This is why growth and technology stocks, whose value depends on distant earnings, were the first to fall.

The third is the confidence and investment channel. Firms postpone share sales, acquisitions and capital spending when prices swing sharply, and at least one planned US listing has already been postponed this month on market uncertainty. The IMF's July 2026 World Economic Outlook identified this as a key risk: if expectations of AI-driven profits are revised down, technology investment "could retrench abruptly" and "frothy equity valuations" could "correct sharply".

The fourth is the capital-flow and currency channel. This is the one that falls hardest on emerging markets. With US yields at 5.25% and the dollar index above 101, the extra return global investors earn by holding Indian, Indonesian or Brazilian assets has narrowed sharply. Foreign portfolio investors bought about ₹29,600 crore of Indian equities in August but have sold about ₹20,700 crore in September. India's foreign exchange reserves fell by $14.9 billion in the week to 18th September, the sharpest weekly drop in about two years, driven by a fall in foreign currency assets as the RBI stepped in to steady the rupee. Reserves stood at $765.9 billion, down roughly $19.8 billion from the record $785.7 billion reached earlier in the month.

The fifth is the commodity channel. It links the other four, and it is the reason the standard response to a crash does not work this time. Higher oil prices raise inflation, which pushes bond yields up, which weighs on equities, which tightens financial conditions further. At about $107, Brent is roughly 20% above the $89 average the IMF assumed for 2026 in July.

Why the usual crash response is not available

Most investors remember market crashes as events that central banks resolved. In 1987, 2008 and 2020, the US Federal Reserve cut interest rates, supplied liquidity and, from 2008 onward, bought bonds on a large scale. Markets recovered because the shock was deflationary: demand collapsed, inflation fell, and central banks had room to ease.

September 2026 is different. The Federal Reserve raised its policy rate to 3.75%–4.00% on 16th September, its first increase since 2023, because inflation is still too high. Markets now put the probability of a further increase at the 28th October meeting at around 70%. Euro area inflation rose to 3.3% in August. India's retail inflation increased to 4.82% in August from 4.45% in July. With the RBI's repo rate at 5.25%, the real policy rate is only about 0.4 percentage points. Cutting rates into an oil-driven price shock would risk undoing the progress made on inflation and would probably push long-term bond yields higher still, since bond investors would demand more compensation for inflation risk. The IMF's July advice points the same way. Where price pressure is temporary but expectations are at risk, central banks should keep real rates broadly constant, which can mean raising nominal rates, not cutting them.

EconoLens Analysis

In the 2008 and 2020 crashes, trouble started in US equities and credit and then spread outward. This time the pressure is coming from the US bond market and oil, and it is spreading to the edges of the global system, with emerging-market equities and currencies taking the largest losses. Two numbers show why India is exposed even though its economy grew 7.8% in April–June.

The first is the yield cushion. The gap between India's 10-year government bond yield (7.16%) and the US 10-year yield (5.25%) is now about 1.9 percentage points. In early 2021 it was close to 5 points, and in late 2023 it was around 2.6 points (approximate market levels). Foreign investors used to be paid well for taking on rupee risk. Today they receive less than two points of extra yield to hold an emerging-market currency that has weakened steadily. That, more than any change in India's growth outlook, is why foreign money is leaving. It also means the rupee, not the RBI's interest rate, is the more natural adjustment valve.

The second is the oil bill. India imports roughly 240 million tonnes of crude a year, about 1.8 billion barrels. By EconoLens's estimate, every sustained $10 rise in the price of a barrel adds about $18 billion to India's annual import bill, roughly 0.4–0.45% of GDP. Brent at $107 instead of the IMF's $89 assumption implies an extra $32 billion a year, about 0.8% of GDP, if prices stay there. Much of that would come through in a wider current account deficit and a weaker rupee.

Taken together, these figures point to a clear conclusion for the RBI's 5th–7th October policy meeting. Raising the repo rate to protect the yield gap would be costly and would probably not work, because the gap is being driven by US Treasury yields that India cannot influence. Spending reserves to hold the rupee at ₹96 is a similar mistake. At the pace seen in the week to 18th September, even a $766 billion reserve pile would be depleted steadily while the underlying pressure stayed the same. The more durable approach is to let the rupee adjust gradually and use reserves to smooth sharp moves, not to hold a line. Rate policy should stay focused on domestic inflation, which, at 4.82%, is above target but still within the RBI's band.

How bad could it get? Three scenarios

The table below sets out three illustrative paths. These are EconoLens scenarios, not forecasts. They combine the IMF's July baseline (global growth of 3.0% in 2026 and 3.4% in 2027; global inflation of 4.7% in 2026) with a commonly used IMF staff rule of thumb: a 10% oil-price increase that lasts most of a year lowers global growth by about 0.15 percentage points and raises global inflation by about 0.4 points. On that rule alone, Brent staying 20% above the IMF's assumption would lower global growth by about 0.3 points and raise inflation by about 0.8 points before any market or confidence effects are counted.

History helps put today's losses in context. Even after a 16% fall, India's benchmark index is well short of the drawdowns in genuine crashes.

The economic consequences, by scenario

In Scenario A, the damage stays mainly in financial markets. Higher borrowing costs slow housing and investment for a few quarters, and emerging-market currencies partly recover. India's cost is mostly a slower pace of foreign inflows and a rupee that settles somewhat weaker, not a growth shock.

Scenario B looks like the 1970s more than 2008. Persistently expensive oil keeps headline inflation above 4.5% worldwide, central banks hold or raise rates, and the economic cost appears as a slow squeeze on real incomes rather than a sudden collapse. Energy-importing economies such as India, Japan, the euro area, Pakistan and much of Southeast Asia take the largest hit through their trade balances. For India, the extra oil bill of about 0.8% of GDP, weaker rupee-denominated returns for foreign investors and firmer inflation would together push the RBI towards at least one precautionary rate increase, even though inflation remains inside its target band.

Scenario C is the true crash case, and its mechanics differ from the other two. A sharp downgrade of AI earnings expectations would hit the most concentrated part of global equity markets. If this coincided with a disorderly bond market, where Treasury supply outpaces buyers and leveraged holders are forced to sell, the resulting liquidity squeeze could spread well beyond the original problem. A flight to safety would then pull yields down and eventually weaken oil demand, so the episode would become deflationary in its later stages. Only at that point would the traditional crash response, rate cuts and central-bank asset purchases, become appropriate again.

How to respond: a playbook for an inflationary crash

1. Keep liquidity support separate from monetary-policy stance. The Bank of England showed how to do this in 2022, when pension-fund selling threatened the gilt market. It bought long-dated bonds for a short, clearly limited period and continued raising interest rates. Central banks facing a disorderly Treasury or government-bond market should do the same: offer time-limited, targeted purchases or repo facilities to restore market functioning, and say explicitly that these are not monetary easing. The Fed's standing repo facility and its repo facility for foreign central banks exist for this purpose and should be used early.

2. Hold the inflation line, but communicate the path clearly. Central banks should not signal cuts that the inflation data do not support. They should, however, make clear that they are reacting to data and not committed to a series of increases. That reduces the risk that markets price an extended tightening cycle and push long-term yields higher on their own.

3. Target fiscal support, and avoid broad subsidies. The IMF's July advice to use "temporary, tightly targeted" help for vulnerable households and to avoid broad subsidies, tax cuts and price controls applies directly here. Governments that cut fuel taxes across the board would add to already heavy bond issuance at a time when bond markets are the weak point.

4. Release buffers where they exist. Bank regulators should be ready to release countercyclical capital buffers, and they should test banks now for losses on bond holdings. With yields rising this quickly, those losses are the kind of risk that caught regional US lenders in 2023.

5. For emerging markets such as India, allow the exchange rate to adjust and protect market functioning. This means using reserves to smooth rather than defend a level. It also means using tools such as FX swap windows and the forward book to meet dollar demand from oil importers without draining spot reserves. If liquidity tightens as the RBI sells dollars, open market purchases of government bonds can keep the yield curve orderly. On the fiscal side, a temporary and partial reduction in fuel excise, reversed once Brent falls back, is less damaging than open-ended subsidies. Coordinated drawdowns of strategic petroleum reserves with other large importers through the International Energy Agency framework would address the actual cause of the shock.

6. Coordinate internationally before stress peaks. Central-bank dollar swap lines, precautionary IMF credit lines for vulnerable economies, and clear signals from the G20 finance track all work better when put in place before a crisis than after one.

What to watch in the next two weeks

• Iran–US talks and Hormuz shipping volumes. Only 132 vessels passed through the Strait in the week to 27th September, compared with about 130 a day before the conflict. Any reopening would weaken the main source of pressure.
• The US 10-year yield around 5.25%. A sustained move above that level would make Scenario B more likely.
• The RBI's decision on 7th October, and whether its statement treats the rupee as the adjustment mechanism.
• The Fed on 28th October. A hike that markets have already priced would do less damage than signals of a longer tightening cycle.
• India's September CPI data on 12th October, which will show the first full month of higher oil prices feeding into inflation.

For live data on these series, see the EconoLens indicators dashboard. Background on the oil shock is in The Strait of Hormuz Shock, and background on US policy is in the Fed's September rate hike. For the wider emerging-market exposure, see Emerging Markets in the Dollar Trap.