Original Lens and publication context
A 19.4% MSCI weight became 12% in eighteen months. Six broker downgrades arrived in sixty days. Korea is up 210% in dollar terms. India is down 10%. The downgrade cascade is not a verdict on India — it is a verdict on price, profit pools, and the AI gravity well pulling EM capital north.
- Six of the largest sell-side desks turned cautious on India inside sixty days — two in March, three in April, the last today — with one posture: rich on valuation, modest on earnings momentum, exposed on relative attractiveness. A verdict on price, not on the story.
- India's weight in the MSCI EM index has fallen from 19.4% in Q4 2024 to 12.0% in May 2026; Korea's has more than doubled from 9.0% to 19.0% and Taiwan's has risen 5.3 points to 25.0%. The mirror is exact and the driver is price.
- The profit pool has moved north: three AI-leveraged companies are expected to earn a combined $500 bn this year — two-and-a-half times corporate India's $200 bn — and make up roughly a quarter of the EM benchmark. India's index sensitivity to the AI cycle is approximately zero.
- In dollar terms Korea is up 210% over twelve months, Taiwan 104%, India down 10% — a 220-point dispersion no allocation framework survives without rotating. Domestic SIP, insurance and EPFO flows absorb the foreign selling, keeping the drawdown shallow and valuations elevated.
- Position: India stays a core long-term EM allocation but not a tactical overweight. The trade is composition, not country — underweight expensive staples and index proxies, own earnings-visibility pockets in private banks, manufacturing capex and infrastructure, avoid narrative small and mid-caps. Base case: a range trade.
Indian markets are not being downgraded because India's long-term story is broken. They are being downgraded because global capital is becoming more selective, and India is no longer the obvious answer to the only question every emerging market allocator asks every quarter — where do I get the most earnings growth per unit of valuation risk.
That distinction sits at the centre of everything that has happened in EM allocation flows over the last six months. The downgrade cascade, the MSCI weight collapse, the dollar return dispersion, and the profit pool concentration all tell the same story when read together. India has not been abandoned. India has been repriced relative to a North Asian earnings shock that has no Indian analogue.
I. TRIGGER — THE DOWNGRADE CASCADE
The sequence has been clean and continuous. UBS and Morgan Stanley turned cautious in March. JPMorgan, HSBC, and Nomura followed in April. Citi closed the cycle today. Six of the largest sell-side desks now carry the same near-term posture on Indian equities: rich on valuation, modest on earnings momentum, exposed on relative attractiveness.
Each desk arrived at this view independently. That matters. When six independent risk committees converge on the same posture inside a sixty-day window, the signal is structural, not idiosyncratic. The instinct is to read this as a sell call. It is not. None of these desks have changed their long-run thesis on India's domestic consumption base, financialisation, manufacturing capex, or formalisation. What has changed is the answer to that single quarterly allocator question.
For most of the last three years, India answered that question better than any other emerging market. Today, it does not. The downgrades are the lagging confirmation of a flow rotation that has already happened in the price tape.
Six desks. Sixty days. One verdict — not on India's story, but on India's price.
II. FLOW — THE MSCI WEIGHT COLLAPSE
The clearest evidence of the rotation is in the index itself. India's weight in the MSCI Emerging Markets Index has fallen from 19.4% in Q4 2024 to 12.0% in May 2026. Korea's has more than doubled, from 9.0% to 19.0%. Taiwan has gained roughly 5.3 percentage points to 25.0%.
PERIOD | INDIA | CHINA | TAIWAN | KOREA |
|---|---|---|---|---|
Q4 2024 | 19.4% | 27.8% | 19.7% | 9.0% |
Q1 2025 | 18.5% | 31.3% | 16.9% | 9.0% |
Q2 2025 | 18.1% | 28.4% | 18.9% | 10.7% |
Q4 2025 | 15.3% | 27.6% | 20.6% | 13.3% |
Q1 2026 | 12.6% | 25.5% | 22.5% | 15.5% |
May '26 SF | 12.0% | 19.0% | 25.0% | 19.0% |
MSCI Emerging Markets Index — Country Weights, Q4 2024 to May 2026.
The mirror is exact. India down 7.4 points. Korea up 10. Taiwan up 5.3. The benchmark-weight shift is mechanical in part; it does not by itself measure the direction or size of active cross-border flows. Weight changes in MSCI indices are driven by relative price performance, free-float adjustments, and rebalancing methodology. In this cycle, the dominant driver is price. When Korean and Taiwanese equities outperform India by triple-digit margins in dollar terms, weights compound mechanically and passive flows reinforce the active rotation.
This is not a one-quarter shift. This is a regime change in EM allocation.
III. RISK — THE PROFIT POOL CONCENTRATION
Index weights eventually follow profit pools. The numbers from the broadcast feed currently circulating are the most important data point in EM equities today, and they need to be stated cleanly:
- Samsung Electronics: $250 bn expected profit, current year
- SK Hynix: $150 bn
- TSMC: $100 bn
- Combined: $500 bn — and roughly 25% of the MSCI EM Index by weight
- India Inc, in aggregate: $200 bn
Three companies — all leveraged to the AI infrastructure cycle — produce two-and-a-half times the earnings of corporate India combined, and now account for a quarter of the EM benchmark. This is the most concentrated profit pool the EM index has ever carried, and the concentration is rising, not falling.
The implication is straightforward. EM equity returns are increasingly a function of AI capex, hyperscaler order books, and HBM memory demand cycles. India's index sensitivity to those drivers is approximately zero. Indian IT services participates in the spend, but at the bottom of the value stack and on heavily compressed margins. India does not own the foundry. It does not own the memory. It does not own the equipment. The earnings flywheel is being built somewhere else.
Capital flows where the marginal earnings growth lives. For the last twelve months, that has been Hsinchu and Suwon. Not Mumbai.
Three companies. Five hundred billion dollars. A quarter of the EM index. India does not own a single name in the stack.
IV. PRICE — RELATIVE PERFORMANCE IS DOING THE TALKING
The dollar return dispersion is what makes this cycle uncomfortable for any allocator still anchored on India's structural narrative. Korea is up 210% over twelve months. Taiwan is up 104%. India is down 10%.
REGION | 2026 YTD | 1-YEAR | 3-YEAR |
|---|---|---|---|
Taiwan | +45% | +104% | +44% |
Korea | +72% | +210% | +42% |
China | −5% | +9% | +7% |
India | −11% | −10% | +7% |
EM Aggregate | +5% | +29% | +16% |
MSCI EM Returns in USD Terms — 2026 YTD, 1-Year, 3-Year.
On a one-year basis, the gap between Korea and India in dollar terms is over 220 percentage points. There is no allocation framework — top-down, bottom-up, factor, or thematic — that survives a 220-point dispersion without rotating capital out of the laggard. The downgrades did not cause the rotation. The rotation forced the downgrades.
The three-year column gives the source’s longer-horizon comparison, but its annualised-versus-cumulative convention requires confirmation before those percentages can be interpreted. The premium India enjoyed for most of the post-COVID cycle has compressed to nothing on a relative basis. What changed is not India's economy. India's growth, fiscal trajectory, and current account profile remain among the strongest in EM. What changed is that the AI cycle delivered an earnings shock to North Asia of a magnitude India cannot replicate from its current sector composition.
Markets do not pay for stories. Markets pay for the earnings the story produces. India's story remains intact. The earnings, in this cycle, are landing elsewhere.
V. THE DOMESTIC OFFSET — WHY THE DRAWDOWN IS SHALLOW
The next question for any allocator is the obvious one. If FIIs have been net sellers and the index has been derated relative to peers, why has India not seen a deeper drawdown in absolute terms?
The answer is the domestic flow architecture. Monthly SIP run-rate, retail direct participation, insurance balance-sheet allocation, and EPFO contributions have created an internal liquidity engine that absorbs foreign selling pressure before it reaches price. This is structurally bullish for the long-run market. It is also tactically problematic, because it has kept Nifty and broader-market valuations elevated even as the cross-border flow has reversed sharply.
The FII view, expressed cleanly, is this: India is a great country that is not currently priced for the next twelve months of earnings. The domestic view, expressed equally cleanly, is this: every monthly SIP tranche is allocated regardless of valuation. Both views are internally consistent. They are simply running on different time horizons and different return objectives.
The risk is the convergence event — the moment domestic flow growth slows and the valuation cushion erodes simultaneously. That is not the base case for the rest of 2026. It is the tail case for 2027 if earnings disappoint into expensive multiples while monthly SIP growth flattens.
VI. POSITION — THE BUY SIDE LENS VIEW
Position. India remains a core long-term EM allocation. It is not a tactical overweight at current valuations. The downgrade cascade is a discipline signal, not a panic signal, and the right interpretation is selectivity rather than retreat.
The trade is no longer country. It is composition. What the next phase of Indian equity returns rewards is not blind beta to the index, but disciplined positioning across pockets where earnings visibility, valuation entry, and capital intensity actually align.
- Underweight expensive consumption staples and broad-market index proxies where multiple compression risk is highest.
- Selectively own earnings-visibility names with reasonable valuation entry points and clean cash-flow profiles.
- Use private-bank, manufacturing capex, and infrastructure exposure where the multiples have already compressed and operating leverage is real.
- Avoid concept-stock and narrative-driven small/mid-cap exposure where the domestic flow has been most distortive and where any FII normalisation hits hardest.
Three scenarios into year-end:
- Base case (~55%): India range-trades. The AI cycle continues to pull marginal flows to North Asia. Indian equities outperform only on stock selection, not at index level. Year closes flat to mid-single-digit USD.
- Bull case (~25%): Earnings season delivers a positive surprise on margin expansion in financials and capex names. FII positioning rebuilds tactically. Index closes with high single-digit USD gains as the valuation compression pauses.
- Bear case (~20%): Domestic flow growth slows, earnings underperform, and the valuation cushion compresses. A 10–15% drawdown is the risk boundary, concentrated in the most overvalued cohorts.
The signal to monitor. Watch the Korea-India correlation in EM allocation surveys and the trajectory of Korean and Taiwanese MSCI weight. The day Korean weight starts to flatten and Indian weight stabilises is the day the marginal allocator has finished rebalancing. That has not happened yet. Until it does, every tactical bounce in Indian equities should be treated as a flow event, not a regime change.
India is not being abandoned. India is being repriced. That is healthy for a market wanting to mature from a liquidity-led story into an earnings-led story.
The mistake would be to read every downgrade as bearish. The bigger mistake would be to ignore valuation completely just because the India story is strong. Both errors arrive at the same destination: a portfolio positioned for the last cycle, not the next one.
The next phase of Indian equity returns will not be a beta phase. It will be a discipline phase. Quality matters. Earnings visibility matters. Reasonable valuation matters. Cash-flow discipline matters. Sectors with actual profit momentum will matter more than broad narratives. Trigger. Flow. Risk. Price. Position.
