Why Is Foreign Money Leaving India?

The problem: money is leaving, and leaving faster

Nearly $58 billion of foreign money has walked out of Indian stocks since October 2024 — and the pace is increasing. In just the first five months of 2026, foreign investors withdrew about ₹2.25 lakh crore (~$24 billion) — already more than they sold in all of 2025.

Read quickly, that looks like a verdict on India — a judgement that the story has broken. It is not. But to see why, we have to ask the right question. The outflows get blamed on a familiar list of culprits: the rise in the capital-gains tax, weak market performance, a falling rupee, and a strong dollar pulling money back to the United States. Each holds a little truth. But they cannot all be the main cause — and as the data will show, the real driver is none of them. So what is actually pulling the money out? To answer that, we first have to understand something most people outside markets never have reason to learn: how the world’s money actually moves.

How the world’s money actually moves

A large segment of money that flows into “emerging markets” does not study India, weigh its prospects, and decide how much to invest. It tracks one list — the MSCI Emerging Markets Index — a single ranking of 1,205 companies across 24 developing countries, each given a weight according to its market value. It is the gravitational centre of the asset class: passive and benchmark-bound funds estimated at over $750 billion simply hold what the index holds, in the proportions the index sets. The larger pool of active money treats it as the benchmark — a scorecard against which every fund manager’s performance is judged.

The crucial point: a country’s weight on that list is not a judgement about the country. It is the automatic output of how its companies’ market values compare with everyone else’s. And because hundreds of billions of dollars follow the list mechanically, that weight actively directs where the money goes. Change the weights, and you change the flows — with no human deciding anything. Keep that in mind, because it is the key to the whole story.

The trigger: real AI demand, then the formula

Over the past two years, one force reshaped that list — the artificial-intelligence build-out. AI runs on advanced semiconductors, and the most advanced chips come from a handful of companies concentrated in two countries: Taiwan (above all TSMC) and South Korea (Samsung and SK Hynix). As the world raced to build AI, the real-world demand for their chips surged — and so did their earnings and their share prices. This was not hype or fund fashion; it was genuine business performance meeting genuine demand.

Here is where the formula takes over. Because index weight follows market value, those soaring share prices automatically lifted Taiwan’s and Korea’s weight in the index — and the $750 billion of passive money then mechanically bought more of them, simply because the index now said so. That buying lifted prices further, which lifted the weights again. The performance was earned; the amplification was automatic. And that same performance, and the promise of AI, also began drawing active fund managers toward these markets — reinforcing a move the formula was already making.

Now the mirror, and the part that matters for India. Index weights must sum to 100%. So every point Taiwan and Korea gained was, by definition, a point taken from everyone else — automatically, regardless of their own merits. As the chip giants swelled, India’s slice of that fixed pie shrank as a matter of arithmetic, and the passive money trimmed India to match. No one decided India was worth less. The formula did it.

You don’t have to do anything wrong to lose the world’s money. You just have to not own the thing it’s racing toward.

What happened to the index

The scale of the shift is best seen in the index itself. India’s weight in the MSCI Emerging Markets Index climbed from around 8% in early 2020 to a peak of about 20.8% in September 2024 — when India briefly stood among the very largest weights, having overtaken China. Less than two years later it has fallen to 10.87% — roughly half, and a six-year low. To see why, look at what now sits at the top of the index:

The table tells the whole story. Nine of the top ten companies are technology businesses; the first non-tech name appears only at number ten. The top three — TSMC, Samsung, SK Hynix — are 28.84% of the entire index between them. And for the first time in 26 years, not a single Indian company sits among the ten largest — HDFC Bank and Reliance, both top-10 names as recently as this spring, have slipped to 11th and 12th. India’s largest listed companies are banks, energy and consumer businesses — fine franchises, but with nothing to do with the AI hardware boom. So when the world’s money rotated toward AI, there was simply nothing in the Indian market for it to land on. The capital moved on the basis of what each market gives exposure to. The money chased the AI layer; performance followed the money, not the other way round. Performance is an output of this story — not its cause.

China: the comparison that sharpens it

If the AI story is the real driver, China proves it from the other direction. China was once the dominant emerging market — 38.7% of the index in early 2021. Then a regulatory crackdown and a property crisis cut it down to roughly 24% by late 2024. India, rising, briefly overtook it. The obvious assumption was that China’s decline was structural and India’s rise permanent.

Then China came back — and the trigger was AI. In January 2025, the Chinese startup DeepSeek showed the world a frontier-class AI model built at a fraction of the expected cost, and global investors abruptly re-rated China’s technology giants as AI companies. Over 2025, Alibaba rose about 96%, Tencent about 55%, Baidu about 60%; the Hang Seng Tech index gained 41%, beating the Nasdaq. China clawed back index weight it had lost for years.

For India, this was the second blade of a scissors. It was being squeezed from both sides at once — the AI rally lifting Taiwan and Korea, and China’s AI-driven recovery clawing weight back. But the sharpest lesson is why China could come back and India could not: China had the companies to re-rate. It had Alibaba and Tencent with real cloud and AI-model businesses, and even domestic chip efforts in Baidu and SMIC. When the AI wave came, China had vessels for the money to flow into. India had none — not a chipmaker, and not even a large, re-ratable software or platform champion. (China’s rally carries its own risks — much of it rests on sentiment and stretched valuations as much as earnings — but the structural point stands: India had nothing comparable to re-rate.)

“But isn’t it the tax, the rupee, the dollar?” — the data says no

Return now to the three culprits everyone blames. Each dissolves on contact with the data — and tax dissolves most clearly of all. India’s 12.5% capital-gains tax on foreign investors is routinely named as the cause of the exodus. But compare across markets:

The pattern runs exactly opposite to the tax explanation. Brazil and Mexico tax foreign equity gains far more heavily than India — up to ~22.5% and ~25–35% respectively, against India’s 12.5% — yet they did not fall as hard as India did. If tax drove the flight, the higher-taxed markets should have bled most. They didn’t. India, the lower-taxed market, fell the hardest.

China makes the same case even more pointedly. China’s foreign investors are effectively exempt from capital-gains tax — but through an administrative circular, not law, one that Beijing can revoke at will, on top of a statutory 10% rate. Investors poured into China anyway, to chase the AI re-rating. If money will accept China’s discretionary, revocable zero-rating to buy AI exposure, then India’s transparent, legislated 12.5% is plainly not what is repelling it. The markets that won the world’s money are not the ones with the lowest taxes — they are the ones with the AI exposure.

The rupee and the dollar fall away the same way. A strong dollar and high US interest rates pulled money from every emerging market alike — yet money still flooded into Taiwan and Korea through that same strong dollar. And the rupee is as much an effect of the selling as a cause of it: when foreigners sell Indian shares they sell rupees too, weakening the currency — a gear inside the machine, not a separate engine. These three are the weather — they explain the timing and the size of the swings. They do not explain the climate: that the index has become an AI vehicle, and India is not in it. The cyclical pressures will ease; the structural drain will not, because it is driven by what India owns.

Why this matters: financial gravity

Step back and the episode reveals something larger than an index. Owning a layer of the technological frontier turns out to be more than strategic leverage — it is financial gravity. The countries that build the frontier attract the capital that compounds, and global indexing then concentrates that capital further, automatically. The countries that don’t build it watch their share drain away by the plain arithmetic of the benchmark. Taiwan’s rise and India’s slip are not two stories; they are one phenomenon seen from opposite ends.

So the real question is not “why is the money leaving?” It is: what must India build so the formula eventually works for it, not against it? The direction is clear, even if the work is long — genuine presence in the layers India can credibly come to own: chip design, where it already has a real claim; the manufacturing and materials base now beginning to form; and, in time, listed champions large enough to register on a global benchmark.

And this is where the capital story opens onto a larger one. The drain of index weight is one visible, measurable symptom of a deeper question: is India building what it takes to stay competitive in the world that is now arriving? Capital is one dimension of that competitiveness — concrete and unfolding in front of us. There are others, in technology, talent, and industrial capability, and these notes will take them up in turn.

They share a single thread: in a world reorganising around a handful of frontier capabilities, you do not have to do anything wrong to fall behind. You only have to not build the thing the world is racing toward. Nothing here is a verdict. It is a map of what must be built.