Research · 22 October 2026
Mostly not the tax. A safe 5% in the US raised the return they need from India, while Indian profits grew under 5% and the rupee fell.
11 exhibits · About ten minutes · Data to 21 October 2026 · Kavish Raj Singh
The scale
Foreign portfolio investors sold $33 billion of Indian shares between January and 21 October 2026, after $18.9 billion in 2025. Two record years in a row, against a long run of buying.
The question is what changed, and what would reverse it.
Exhibit 1
Net FPI investment in Indian equities, US$ billion, calendar years
Source: NSDL FPI Monitor; India FPI flow model, Annual tab. *2026 is 1 January to 21 October.
The framework
Every foreign fund weighs the same things: the safe return at home, the price of Indian profits, the currency, the alternatives, and what would change the picture. Each of the next screens takes one.
On every measure, India looks less attractive to a dollar investor than at its 2024 peak.
Exhibit 2
Measures at the September 2024 market peak and in October 2026
| Question | Measure | Sep 2024 | Oct 2026 |
|---|---|---|---|
| What can I earn at home? | Return India must beat, in rupees | 13.1% | 14.5% |
| What am I paying for Indian profits? | Profit growth the price implies, a year | 11.7% | 12.8% |
| What do I get back in dollars? | Value of $100 in the Nifty 50 | $100 | $77 |
| Where else could I go? | India's weight in MSCI Emerging Markets | ~20% | 11.9% |
| What would bring me back? | Global signposts met | – | 0 of 5 |
Source: India FPI flow model (Valuation, Monthly and Mosaic tabs); MSCI via Business Standard.
First, the usual suspect
In 2008, with no tax on long-term gains, foreign investors still sold $12 billion. When the tax returned in 2018 and rose in 2024, flows turned positive within one to two months.
A cut would help sentiment. It cannot close the gaps on the next four screens.
Exhibit 3
Net FPI investment in Indian equities, US$ billion
| When | Change | That month | Next two months |
|---|---|---|---|
| Calendar 2008 | No tax on long-term gains | -12.0 (year) | – |
| Feb 2018 | 10% long-term tax reintroduced | -1.8 | +0.9 |
| Jul 2024 | Long-term rate raised to 12.5% | +3.9 | +7.8 |
“Merely cutting or even abolishing capital gains tax… will not be enough to bring foreign investors back to the Indian stock market, analysts believe.”Business Standard, 3 March 2025
Source: NSDL FPI Monitor; India FPI flow model, Monthly and Episodes tabs.
Question 1 · What can I earn at home?
A US 10-year Treasury pays 5.15%, almost risk-free. A dollar investor wants that, plus a premium for owning shares, plus compensation for the rupee's average 3.96% annual fall since 2007. The three compound.
The bar rose 1.4 points from 13.1% in September 2024, entirely because US yields rose.
Exhibit 4
Required annual return in rupees, % (components in percentage points)
(1 + 5.15% + 5.0%) × (1 + 3.96%) − 1 = 14.5%. *The 5.0% premium is my assumption; change it in the Excel.
Source: US Treasury via Yahoo Finance; USD/INR January 2007 to October 2026; India FPI flow model, Valuation tab.
Question 2 · What am I paying for Indian profits?
At 22 times forward earnings and a 13.1% required return, the Nifty priced in about 11.7% profit growth a year, indefinitely. Profits grew 3.4% in FY25 and 4.5% in FY26.
The P/E has fallen to 17.6, but because the bar rose, today's price still implies 12.8% a year.
Exhibit 5
Nifty 50 earnings growth a year, %
DeliveredImplied by the price (Gordon growth)
Source: JM Financial via Business Standard; NSE; India FPI flow model, Valuation tab.
Question 3 · What do I get back in dollars?
From the September 2024 peak the Nifty is down 12.9% in rupees. Over the same period the rupee weakened from 83.7 to 94.8 per dollar, so $100 invested at the peak is worth about $77.
Selling weakens the rupee, which lowers the next investor's return. The loop runs until the currency turns.
Exhibit 6
Nifty 50, month-end, indexed to 100 at end-September 2024
Source: NSE and USD/INR via Yahoo Finance; India FPI flow model, Monthly tab. Month-end closes; the text uses the 27 September 2024 peak.
Question 4 · Where else could I go?
Global money chased the hardware behind artificial intelligence, made largely in Korea and Taiwan. India has almost no listed chip makers. Its weight in the MSCI Emerging Markets index fell from about 20% in 2024 to 11.94% by April 2026.
Foreigners also sold Korea as prices rose. What moved was portfolio weight, not a direct transfer from India.
Exhibit 7
Price return in US dollars, calendar 2025, %
Source: Yahoo Finance month-end closes, December 2024 to December 2025, converted at month-end exchange rates.
The evidence
In every year since 2008 when the dollar fell and emerging markets rose, foreign money flowed into India. In 2025 the dollar fell 9.4% and emerging markets rose 31%. A model of global conditions predicted roughly zero; foreigners sold $18.9 billion.
The unexplained part, which I call the India factor, was more negative in 2025 and 2026 than in any year since 2008.
Exhibit 8
Actual minus predicted FPI equity flow, US$ billion a year
Source: India FPI flow model, Regression and Annual tabs. Monthly regression on EM returns and prior 12-month flow, R² 0.18. *2026 is January to September.
Question 5 · What would bring me back?
When the dollar fell more than 5% over a year, foreign investors were net buyers of India in 81% of cases. Every genuine turn since 2008 began within four months of the dollar's 12-month high (8 of 8).
2025 showed a weaker dollar is not enough while India's own case is weak. India's side has since improved; the dollar is still at its 12-month high.
Exhibit 9
Share of months since 2008 followed by 12 months of net FPI buying
| Condition over the next 12 months | Months | Net buyers |
|---|---|---|
| All months | 213 | 72% |
| Dollar falls more than 5% | 37 | 81% |
| Dollar rises more than 5% | 56 | 59% |
| EM shares fall more than 15% | 40 | 32% |
| FPIs had sold more than $10 bn in the prior 12 months | 23 | 65% |
Source: India FPI flow model, Odds tab. Twelve-month windows overlap, so the months come from a handful of separate episodes.
What I am watching
Twelve signposts, written down in advance with trigger levels. 4 of 5 on India's side are met: profit growth, valuation, a cooling AI trade and light positioning. 0 of 5 global triggers are.
The first to watch: the US 10-year below 4.5%, Brent below $85 and the rupee below 92 per dollar.
Exhibit 10
Levels are my judgement, set before the outcome is known
| Signpost | Now | Needs | Status | |
|---|---|---|---|---|
| Global trigger | US 10-year yield | 5.15 | below 4.50 | – Not yet |
| Global trigger | US 3-month bill (Fed cuts priced) | 4.06 | below 3.75 | – Not yet |
| Global trigger | Dollar index | 101.6 | below 97.0 | – Not yet |
| Global trigger | Brent crude ($) | 112 | below 85 | – Not yet |
| Global trigger | Rupees per dollar | 94.79 | below 92.00 | – Not yet |
| India | Nifty profit growth, latest quarter | 12.9% | at least 12.0% | ✓ Met |
| India | Nifty forward P/E | 17.6 | below 18.0 | ✓ Met |
| India | India's P/E premium to EM | 77% | below 60% | – Not yet |
| India | KOSPI below 2026 high | −34.7% | below −20.0% | ✓ Met |
| India | Fund managers' net position | −32% | below −25% | ✓ Met |
| Confirmation | FPI flow, last 3 months ($ bn) | −5.9 | at least 0 | – Not yet |
| Confirmation | LTCG or STT cut announced | 0 | at least 1 | – Not yet |
Source: India FPI flow model, Mosaic tab, linked to live data where available.
My view · 22 October 2026
Across four scenarios, the model gives a 11% chance of net foreign buying over the next 12 months, rising to 49% if US rates fall after a squeeze. That is below the 81% historical frequency because that figure assumes the dollar has already fallen; this one starts from today.
What would change my view: sustained foreign buying while US yields stay above 5% and oil above $100.
Exhibit 11
Probabilities are my judgement; flows and odds from 1,000 simulated years
| Scenario | Probability | Avg flow ($ bn) | Net buying |
|---|---|---|---|
| 1. AI boom holds, rates high | 35% | -30.2 | 1% |
| 2. US squeeze and AI bust | 20% | -12.9 | 12% |
| 3. After the squeeze: Fed cuts | 15% | -1.1 | 49% |
| 4. Soft landing, weak dollar | 30% | -29.0 | 2% |
| All four, weighted | 100% | -21.7 | 11% |
Source: India FPI flow model, Scenarios and Simulation tabs. A scenario model, not a forecast to trade on.
Every exhibit is a formula in the Excel: a monthly regression of flows on global conditions, conditional frequencies, a valuation build-up and a 1,000-run scenario simulation. Blue cells are inputs; yellow cells are my judgement.
NSDL monthly flows from January 2007, Yahoo Finance month-end prices, and 62 Business Standard articles on every foreign selling wave since 2008.
Research is the quantitative read: models run on the data, showing what the numbers say is happening and what comes next. Before the Headline is the qualitative read: one industry or sector at a time, how the pieces connect and how the story could play out.
The model
Every exhibit above is a formula in this workbook, and every sheet explains itself in plain English. Blue numbers are inputs; yellow cells are my judgement calls. Change them and the answer moves.
Download the model (.xlsx)12 tabs · 0.6 MB · opens in Excel, Numbers or Google Sheets
33 terms in plain English, for anyone new to finance. Start here.
| 1 | Glossary: every term in this file, in plain English | |||
|---|---|---|---|---|
| 2 | Written for someone who has never studied finance. Read this first if a word on another sheet is new to you. | |||
| 4 | Term | What it means | Where you'll see it | |
| 5 | FPI (also called FII) | Foreign Portfolio Investor: a fund based outside India that buys Indian shares or bonds, like a US pension fund. 'Net flow' is what they bought minus what they sold. | Monthly, column B | |
| 6 | Net selling / outflow | Foreign investors sold more than they bought. In this file it shows as a negative number in brackets: (33.1) means $33.1 billion sold. | Cover, Annual | |
| 7 | NSDL | National Securities Depository Ltd, which holds India's shares electronically. It publishes how much foreign investors bought and sold, every day and month. | Monthly | |
| 8 | Nifty 50 | An index of India's 50 biggest listed companies. When people say 'the market went up', they usually mean this number. | Monthly, column H | |
| 9 | Earnings, EPS | Earnings means profit. EPS is earnings per share. 'Nifty EPS growth' is how fast the combined profits of the Nifty 50 companies grew. | Valuation | |
| 10 | P/E (price to earnings) | How many rupees you pay for each rupee of yearly profit. A P/E of 22 means you pay ₹22 for ₹1 of profit a year. Forward P/E uses next year's expected profit. Lower means cheaper. | Valuation, row 5 | |
| 11 | Valuation premium | How much more expensive India is than other emerging markets. A 77% premium means India's P/E is 1.77 times the average. | Valuation, row 6 | |
| 12 | Emerging markets (EM), MSCI EM | Stock markets of developing economies: China, India, Taiwan, Korea, Brazil and others. MSCI EM is the index most global funds measure themselves against. | Annual, Odds | |
| 13 | Index weight | A country's share of an index. India was about 20% of MSCI EM in 2024 and is 11.9% now, so index-tracking funds hold less India automatically. | Mosaic | |
| 14 | EEM | A fund that tracks MSCI Emerging Markets. We use its price as a stand-in for 'how emerging-market shares did'. | Monthly, column J | |
| 15 | US 10-year yield | The interest the US government pays to borrow for 10 years. It is treated as the safest return in the world for someone holding dollars. | Valuation, Mosaic | |
| 16 | US 3-month bill | The same, for 3 months. It moves with what markets expect the US central bank to do next. | Monthly, column F | |
| 17 | The Fed | The US central bank. When it raises interest rates, dollars become more rewarding to hold and money flows back to the US. | Episodes | |
| 18 | Dollar index (DXY) | The value of the US dollar against other major currencies. Up means a stronger dollar. | Monthly, column D | |
| 19 | USD/INR | How many rupees one dollar buys. Going from 83.7 to 94.8 means the rupee got weaker. | Monthly, column G | |
| 20 | Rupee depreciation | The rupee losing value. A foreign investor's gains shrink when converted back into dollars. | Valuation, row 14 | |
| 21 | Equity risk premium | The extra return investors want for owning shares instead of safe government bonds. This file assumes 5% a year for India. | Valuation, row 13 | |
| 22 | Hurdle (required return) | The yearly return India must offer, in rupees, to beat the safe US option once risk and the rupee's fall are added. 14.5% today. | Valuation, row 15 | |
| 23 | Brent crude | The world's main oil price. India imports most of its oil, so expensive oil weakens the rupee and pushes up inflation. | Mosaic | |
| 24 | KOSPI, Taiex | The main stock indexes of South Korea and Taiwan. They are full of chip makers like Samsung, SK Hynix and TSMC. | Monthly, column I | |
| 25 | AI trade | Investors buying the companies that make chips and hardware for artificial intelligence. | Mosaic | |
| 26 | DII, SIP | Domestic institutional investors (Indian mutual funds and insurers). SIPs are monthly investment plans that bring a steady flow of Indian savings into funds. | Mosaic | |
| 27 | LTCG, STT | Long-term capital gains tax, on profits from shares held over a year, and the Securities Transaction Tax charged on each trade. | Cover, Episodes | |
| 28 | Regression | A way to measure, from past data, how much one thing moves when another moves. Here: how much foreign flows move when emerging-market shares move. | Regression | |
| 29 | Coefficient | The size of that link. 0.188 means: for every 1% rise in emerging-market shares in a month, foreign investors put in about $0.19 billion more. | Regression, row 14 | |
| 30 | t-stat | How confident we can be that a link is real and not chance. Above 2 (or below −2) is usually taken as real. | Regression | |
| 31 | R squared | How much of the ups and downs the model explains. 0.18 means 18%. The other 82% is India's own story and random noise. | Regression, row 12 | |
| 32 | India factor (residual) | What foreign investors actually did, minus what global conditions predicted. It is the part of the flow that is about India itself. | Annual, column D | |
| 33 | Probability, odds | How often something happened before. 81% means it happened in about 8 of every 10 similar cases. | Odds | |
| 34 | Scenario | One possible future, with its own assumptions. This file has four. | Scenarios | |
| 35 | Monte Carlo simulation | Running the model 1,000 times with random but realistic surprises, to see the whole range of what could happen rather than one guess. | Simulation | |
| 36 | Median, percentile | The median is the middle outcome. The 10th percentile is a bad case: only 1 outcome in 10 was worse. | Simulation | |
| 37 | Signpost | A level written down in advance. If it is reached, the view should change. It keeps the analysis honest. | Mosaic, section 3 | |
| 38 | Percentage point | The simple difference between two percentages. A yield going from 3.8% to 5.1% rose 1.3 percentage points. | Valuation | |
The short answer, how to use the file, and what each sheet does.
| 2 | Why foreign investors are selling India, and what would bring them back | |||
|---|---|---|---|---|
| 3 | A flow model built on every month of NSDL data since 2007. Built by Kavish Raj Singh, co-founder of Investace Capital. 22 October 2026. | |||
| 4 | New to finance? Open the Glossary tab first: 33 terms explained in plain English. Then read Cover, Annual and Mosaic. | |||
| 5 | The short answer | |||
| 6 | FPI equity flow, 2026 so far ($ bn) | (33.1) | Jan to 21 Oct. The biggest year of selling on record, and 2025 was the previous record. | |
| 7 | India factor, 2025 ($ bn) | (18.6) | What global conditions don't explain. 2025 had a falling dollar and booming EM, and India still lost money. | |
| 8 | Earnings growth the Sep 2024 price assumed | 11.7% | Against 3–5% delivered in FY25 and FY26. | |
| 9 | Return a foreign investor now needs in rupees | 14.5% | US 10-year above 5% plus a risk premium plus the rupee's average fall. The bar rose even as India got cheaper. | |
| 10 | Odds of buying in the 12 months after $10 bn+ of selling | 65% | History says rebounds are normal, but today's episode has already broken that once. | |
| 11 | Odds of buying when the dollar then falls more than 5% | 81% | The single best friend of FPI inflows. | |
| 12 | Probability-weighted 12-month flow, Nov 2026–Oct 2027 ($ bn) | (21.6) | Across the four worlds on the Scenarios sheet, at our probabilities. | |
| 13 | Probability FPIs are net buyers over that year | 11% | From 1,000 simulated years. | |
| 14 | India factor needed to turn buyers if the Fed cuts after a squeeze ($ bn) | 6.1 | World 3. Compare: 2023 was +15. | |
| 15 | Where the mosaic says we are today | Loaded, waiting for the trigger | ||
| 17 | How to use this file | |||
| 18 | 1. Start on Mosaic: 16 pieces of news, scored, then 12 signposts and where they leave us today. | |||
| 19 | 2. Change the blue and yellow cells on Scenarios (probabilities, EM returns, the India factor). The Cover, Mosaic and Simulation update on their own. | |||
| 20 | 3. Test history yourself: change the dates on Episodes or the levels on Mosaic and watch the answers move. | |||
| 21 | 4. To refresh, paste new month-end numbers into the blue cells at the bottom of Monthly. Every other number is a formula. | |||
| 23 | How to read it | |||
| 24 | The thesis, tested: a US liquidity squeeze or AI bust does not send money to India at first. In 2008 and 2022 FPIs sold India along with everything else. The money came back the year after (2009, 2023), once US rates stopped rising and the dollar softened. | |||
| 25 | What is different now: in 2025 the dollar fell and EM rallied, and the money went to Korea and Taiwan's chip makers instead of India. A weak dollar is necessary, not sufficient. India also needs earnings growth that beats a higher bar. | |||
| 26 | Where India could benefit: in an AI bust, India is the large EM with little AI in it. Funds are 32% underweight India (BofA Asia survey, Aug 2026). If chip markets fall and earnings recover to the 15% expected for FY27, the India factor can flip positive, as it did in 2009 and 2023. | |||
| 27 | The number to watch is the India factor. Global conditions explain under a fifth of monthly flows. The rest is whether India's earnings, valuation and rupee give a foreign investor a reason to come back. | |||
| 29 | Sheets | |||
| 30 | Glossary | 33 terms in plain English. Start here if you are new to finance | ||
| 31 | Mosaic | Every dot that matters, twelve signposts, and our read | ||
| 32 | Monthly | Raw data, 2007–2026, and calculated columns | ||
| 33 | Regression | What moves monthly flows (LINEST) | ||
| 34 | Annual | Each year: global push vs India factor, with chart | ||
| 35 | Odds | History as conditional probabilities | ||
| 36 | Episodes | Ten selling waves since 2008, calculated from the data | ||
| 37 | Valuation | What the price assumed, and the foreign investor's hurdle | ||
| 38 | Scenarios | Four worlds for the next 12 months. Edit the blue cells | ||
| 39 | Simulation | 1,000 simulated years | ||
| 40 | Draws | Fixed random numbers behind the simulation | ||
| 41 | Conventions | |||
| 42 | Blue: raw data or an assumption you can change | |||
| 43 | Black: calculated | |||
| 44 | Green: pulled from another sheet | |||
| 45 | Yellow shading: judgement calls that drive the answer | |||
| 46 | $ bn throughout unless marked. Negative numbers in brackets = FPI selling. | |||
| 48 | Research and teaching material. Views are personal. Not investment advice, not a recommendation, and not a forecast to trade on. Scenario probabilities are judgement. Investace Capital is a brand of SM Research Services Pvt. Ltd., SEBI RA INH000017736. | |||
16 dated pieces of news, scored, then 12 signposts and today's stage.
| 1 | The mosaic: every dot, what it says, and where it leaves us | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 2 | Each dot is one fact from the news or the data. Score how it pushes foreign money (−2 out, +2 in) and how much it matters (1–3). The signposts below say what has to change. | ||||||||
| 3 | How to read this sheet. Part 1 lists 16 facts from the news. 'Push' says which way each fact moves foreign money: −2 strongly out, +2 strongly in. 'Weight' says how much it matters, from 1 to 3. Score = push × weight. Part 2 averages the scores by theme. Part 3 lists twelve signposts: levels that would change my view, with today's numbers pulled live from the data. Part 4 counts how many signposts are met and names the stage we are in. Every score and level is my judgement: change any blue cell and the answer moves. | ||||||||
| 4 | 1. The dots | ||||||||
| 5 | # | Date | What happened | Driver | Push on FPI flows (−2 to +2) | Weight (1–3) | Score | Source | |
| 6 | 1 | 13 Oct 26 | FPIs sold a record ₹3.08 trn of Indian shares in 2026 to 12 Oct, nearly double 2025's record. | Flows | -2 | 3 | -6 | Business Standard, 13 Oct 2026 (NSDL) | |
| 7 | 2 | 11 Jun 26 | Nifty earnings grew 3.4% in FY25 and 4.5% in FY26, against 12–15% expected. | Earnings | -2 | 3 | -6 | JM Financial via Business Standard, 11 Jun 2026 | |
| 8 | 3 | 20 Aug 26 | FY27 Nifty earnings seen +15%; Q1 FY27 came in at +12.9%. | Earnings | +2 | 3 | +6 | JM Financial via Business Standard, Aug 2026 | |
| 9 | 4 | 12 Sep 26 | Nifty forward P/E down from 22x (Sep 2024) to 17.6x; premium to EM from 110% to 77%. | Valuation | +1 | 2 | +2 | NSE Market Pulse, Jul 2026; Business Standard, 12 Sep 2026 | |
| 10 | 5 | 22 Oct 26 | US 10-year yield above 5%: dollars earn more at home, so India's hurdle is about 14.5% in rupees. | US rates | -2 | 3 | -6 | Yahoo Finance; Valuation sheet | |
| 11 | 6 | 7 Oct 26 | RBI raises the repo rate 25 bp to 5.5%, its first hike since Feb 2023; rupee at 96.78. | Rupee | -1 | 2 | -2 | Business Standard, 7 Oct 2026 | |
| 12 | 7 | 21 Oct 26 | Rupee gains for an eighth straight session, the longest run in eight years, to 94.47. | Rupee | +1 | 1 | +1 | Business Standard, 21 Oct 2026 | |
| 13 | 8 | 22 Oct 26 | Brent crude at about $112. India imports most of its oil: worse inflation, a weaker rupee. | Oil | -2 | 3 | -6 | Yahoo Finance (BZ=F) | |
| 14 | 9 | 11 Oct 26 | Over 12 months KOSPI +87%, Taiwan +81%, Nifty −10.6%. Money followed AI chip profits. | AI rotation | -2 | 3 | -6 | Business Standard, 11 Oct 2026 (Bloomberg data) | |
| 15 | 10 | 15 Oct 26 | Chip stocks crash: KOSPI loses 7–8% in single days in July and October as AI trades unwind. | AI rotation | +1 | 2 | +2 | AP and Reuters reports, Jul and Oct 2026 | |
| 16 | 11 | 20 Aug 26 | BofA Asia fund managers net 32% underweight India, the lowest in Asia. Room to buy. | Positioning | +1 | 2 | +2 | BofA Asia FMS via Business Standard, 20 Aug 2026 | |
| 17 | 12 | 13 Aug 26 | India's weight in MSCI EM down to 11.9% from about 20% in 2024. | Positioning | +1 | 1 | +1 | Business Standard, 13 Aug 2026 | |
| 18 | 13 | 31 Dec 25 | 2025: dollar −9%, EM +31%, and FPIs still sold $18.9 bn of India. The weak-dollar rule broke. | India factor | -2 | 3 | -6 | Annual sheet (NSDL, Yahoo) | |
| 19 | 14 | 13 Oct 26 | DIIs bought ₹6.83 trn in 2026; SIPs at a record ₹32,490 cr in Sep. Holds the market up, and makes exits easy. | Domestic money | 0 | 1 | 0 | Business Standard, 13 Oct 2026; AMFI | |
| 20 | 15 | 13 Oct 26 | Finance Minister open to views on LTCG and STT; Sensex +1,109 points that day. | Tax | +1 | 1 | +1 | Business Standard, 13 Oct 2026 | |
| 21 | 16 | 2 Feb 26 | US–India trade deal cuts the US tariff on Indian goods from 50% to 18%. | Trade | +1 | 1 | +1 | Business Standard, 11 Oct 2026 | |
| 23 | 2. What the dots add up to | ||||||||
| 24 | Driver | Average push | Weight | Reading | |||||
| 25 | US rates | -2.0 | 3 | Pushing money out | |||||
| 26 | Oil | -2.0 | 3 | Pushing money out | |||||
| 27 | Rupee | -0.3 | 3 | Leaning out | |||||
| 28 | Earnings | 0.0 | 6 | Neutral | |||||
| 29 | Valuation | +1.0 | 2 | Pulling money in | |||||
| 30 | AI rotation | -0.8 | 5 | Leaning out | |||||
| 31 | Positioning | +1.0 | 3 | Pulling money in | |||||
| 32 | India factor | -2.0 | 3 | Pushing money out | |||||
| 33 | Flows | -2.0 | 3 | Pushing money out | |||||
| 34 | Domestic money | 0.0 | 1 | Neutral | |||||
| 35 | Tax | +1.0 | 1 | Pulling money in | |||||
| 36 | Trade | +1.0 | 1 | Pulling money in | |||||
| 37 | All dots, weighted | -0.65 | Net push out | ||||||
| 40 | 3. Twelve signposts: what has to change before foreign money comes back | ||||||||
| 41 | # | Group | Signpost | Now | Test | Level | Status | Where the number comes from | |
| 42 | 1 | Trigger | US 10-year yield (%) | 5.15 | below | 4.50 | Not yet | Monthly sheet, latest close | |
| 43 | 2 | Trigger | US 3-month bill (%): markets expect Fed cuts | 4.06 | below | 3.75 | Not yet | Monthly sheet, latest close | |
| 44 | 3 | Trigger | Dollar index | 101.6 | below | 97.0 | Not yet | Monthly sheet, latest close | |
| 45 | 4 | Trigger | Brent crude ($) | 112 | below | 85 | Not yet | Monthly sheet, latest close | |
| 46 | 5 | Trigger | USD/INR | 94.79 | below | 92.00 | Not yet | Monthly sheet, latest close | |
| 47 | 6 | India pull | Nifty earnings growth, latest quarter | 12.9% | at least | 12.0% | Met | Q1 FY27, JM Financial via Business Standard | |
| 48 | 7 | India pull | Nifty forward P/E | 17.6 | below | 18.0 | Met | Bloomberg via Business Standard, 12 Sep 2026 | |
| 49 | 8 | India pull | India's forward P/E premium to MSCI EM | 77% | below | 60% | Not yet | Same; long-run average about 57% | |
| 50 | 9 | Fuel | KOSPI below its 2026 high (AI trade cooling) | (34.7%) | below | (20.0%) | Met | Monthly sheet, month-end closes | |
| 51 | 10 | Fuel | Fund managers' net position in India | -32% | below | -25% | Met | BofA Asia FMS, Aug 2026: very underweight = room to buy | |
| 52 | 11 | Confirmation | FPI flow, last 3 months ($ bn) | (5.9) | at least | – | Not yet | Monthly sheet (Oct 2026 to the 21st) | |
| 53 | 12 | Confirmation | LTCG or STT cut announced (1 = yes) | 0 | at least | 1 | Not yet | Budget, Feb 2027; FM 'open to views', 13 Oct 2026 | |
| 56 | 4. Where that leaves us | ||||||||
| 57 | Global triggers met (US rates, Fed, dollar, oil, rupee) | 0 of 5 | |||||||
| 58 | India pull and fuel met (earnings, price, AI cooling, positioning) | 4 of 5 | |||||||
| 59 | Confirmation met (FPIs actually buying, tax) | 0 of 2 | |||||||
| 61 | Stage | Loaded, waiting for the trigger | |||||||
| 62 | Stages run: Selling continues → Loaded, waiting for the trigger → Turning → Buying. The formula moves on its own as signposts are met. | ||||||||
| 63 | The stage is worked out by a formula. Three or more global triggers met moves us to 'Turning'; actual buying confirms 'Buying'. Until the global side moves, India's own improvements only load the gun. | ||||||||
| 64 | Model numbers behind it | ||||||||
| 65 | Probability FPIs are net buyers over the next 12 months | 11% | |||||||
| 66 | Same, if the Fed cuts after a squeeze (world 3) | 49% | |||||||
| 67 | Probability-weighted 12-month FPI flow ($ bn) | (21.7) | |||||||
| 69 | Our read, 22 Oct 2026 | ||||||||
| 70 | Not yet. Most of the India-side pieces are in place: earnings are recovering, the price is lower, chip markets are cracking and funds are very underweight. | ||||||||
| 71 | None of the global triggers has fired: US yields above 5%, Brent above $110, a firm dollar and a rupee near 95. Until two or three of them turn, selling is likely to continue. | ||||||||
| 72 | When they turn, the return can be quick, as in 2009 and 2023, because so many funds are underweight. Watch signposts 1 to 5 first, then 11 for proof. | ||||||||
| 74 | Scores, weights and levels are our judgement and are meant to be argued with: change any blue cell. Research and teaching material, not investment advice. SEBI RA INH000017736. | ||||||||
Every year since 2008: what foreign investors did vs what global conditions predicted.
| 1 | Every year since 2008: global push vs the India factor | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 2 | Actual FPI equity flow, what global conditions alone would have predicted, and the gap. 2026 is Jan to Sep. | ||||||||
| 3 | How to read this sheet. 'Actual' is what foreign investors really bought or sold that year (NSDL). 'Predicted by global model' is what the Regression sheet says global conditions alone would have produced, mostly from how emerging-market shares did. 'India factor' = actual minus predicted: the part that is about India itself, like earnings, valuations and the rupee. Before 2024 it bounced around zero. 2024, 2025 and 2026 are three big negative years in a row, which is the heart of this whole file. | ||||||||
| 4 | Year | Actual FPI flow ($ bn) | Predicted by global model ($ bn) | India factor ($ bn) | EM return | Dollar index change | US 10y change (pp) | Nifty return in $ | What happened |
| 5 | 2008 | (12.0) | (7.6) | (4.3) | (50.2%) | 6.0% | (1.79) | (60.6%) | Global financial crisis |
| 6 | 2009 | 17.5 | 11.3 | 6.2 | 66.2% | (4.2%) | 1.60 | 82.1% | Stimulus rebound, election |
| 7 | 2010 | 29.4 | 15.7 | 13.7 | 14.8% | 1.5% | (0.54) | 22.1% | Record inflows |
| 8 | 2011 | (0.4) | 7.8 | (8.1) | (20.4%) | 1.4% | (1.43) | (36.3%) | Euro crisis, Indian inflation at 9%+ |
| 9 | 2012 | 24.4 | 9.5 | 14.9 | 16.9% | (0.5%) | (0.12) | 23.6% | |
| 10 | 2013 | 20.1 | 13.4 | 6.7 | (5.7%) | 0.3% | 1.27 | (5.4%) | Taper tantrum (equity flows held up) |
| 11 | 2014 | 16.1 | 9.0 | 7.2 | (6.0%) | 12.8% | (0.86) | 28.1% | Election rally |
| 12 | 2015 | 3.2 | 4.4 | (1.2) | (18.1%) | 9.3% | 0.10 | (8.5%) | China devaluation, MAT notices |
| 13 | 2016 | 3.2 | 3.7 | (0.6) | 8.8% | 3.8% | 0.18 | 0.8% | Demonetisation, Trump |
| 14 | 2017 | 7.8 | 9.9 | (2.1) | 34.6% | (10.0%) | (0.04) | 36.9% | Weak dollar |
| 15 | 2018 | (4.4) | (1.1) | (3.3) | (17.1%) | 4.4% | 0.28 | (5.8%) | Fed hikes, crude, LTCG back, rupee 74 |
| 16 | 2019 | 14.4 | 5.9 | 8.5 | 14.9% | 0.2% | (0.77) | 9.8% | |
| 17 | 2020 | 23.0 | 8.6 | 14.5 | 15.2% | (6.7%) | (1.00) | 12.0% | Covid crash, then liquidity flood |
| 18 | 2021 | 3.8 | 14.8 | (11.1) | (5.5%) | 6.4% | 0.59 | 22.0% | Expensive, DIIs take over |
| 19 | 2022 | (16.5) | (13.2) | (3.3) | (22.4%) | 8.2% | 2.37 | (6.1%) | Fed hikes, Ukraine, rupee -10% |
| 20 | 2023 | 20.7 | 6.1 | 14.7 | 6.1% | (2.1%) | (0.01) | 20.7% | Fed pause, earnings strong |
| 21 | 2024 | 0.1 | 10.3 | (10.2) | 4.0% | 7.1% | 0.71 | 4.4% | IPO buying hid ₹1.21 trn of exchange selling |
| 22 | 2025 | (18.9) | (0.3) | (18.6) | 30.8% | (9.4%) | (0.41) | 5.6% | Weak dollar, EM +31%, and India still sold: AI rotation |
| 23 | 2026 | (27.8) | (4.4) | (23.4) | 22.1% | 3.2% | 1.13 | (19.1%) | War, oil, US 10y above 5%, AI trade, rupee 96 |
| 24 | Total | 103.6 | 103.6 | (0.0) | |||||
| 26 | The India factor in numbers | ||||||||
| 27 | Average, 2008–2023 ($ bn a year) | 3.3 | |||||||
| 28 | Spread (std dev), 2008–2023 | 8.6 | |||||||
| 29 | Best year | 14.9 | |||||||
| 30 | Worst year before 2024 | (11.1) | |||||||
| 31 | 2024 | (10.2) | |||||||
| 32 | 2025 | (18.6) | |||||||
| 33 | 2026, Jan–Sep | (23.4) | |||||||
History as probabilities: what happened in the 12 months after each starting point.
| 1 | History as probabilities | ||||
|---|---|---|---|---|---|
| 2 | Each month since 2008 is a starting point. What share of the next 12 months saw FPIs as net buyers? | ||||
| 3 | How to read this sheet. We take every month since 2008 as a starting point and look at the 12 months after it. 'Net buyers next 12m' is how often foreign investors bought more than they sold over that year. Each row filters history to one situation, for example only the times the dollar then fell more than 5%. Read it as: when this happened before, how often did the money come in? | ||||
| 4 | Condition at the start (or over the next 12 months) | Months | Net buyers next 12m | Average next-12m flow ($ bn) | Note |
| 5 | All months | 213 | 72% | 7.0 | |
| 6 | Dollar falls over the next 12 months | 87 | 76% | 8.8 | |
| 7 | Dollar falls more than 5% | 37 | 81% | 12.8 | The textbook case for EM inflows |
| 8 | Dollar rises more than 5% | 56 | 59% | 1.1 | |
| 9 | EM shares rise more than 15% over the next 12 months | 60 | 78% | 8.7 | |
| 10 | EM shares fall more than 15% | 40 | 32% | (6.3) | The squeeze or crash case |
| 11 | FPIs had sold more than $10 bn in the prior 12 months | 23 | 65% | 5.5 | Today: see C14 |
| 12 | Sold more than $10 bn, and the dollar then fell | 17 | 76% | 10.0 | |
| 13 | Sold more than $10 bn, and EM shares then rose | 16 | 69% | 4.5 | |
| 14 | FPI flow over the last 12 months to Oct 2026 ($ bn) | (36.0) | |||
| 16 | Windows overlap (Jan–Dec and Feb–Jan share 11 months), so the month counts overstate how many independent episodes there are. Heavy selling has happened about four times: 2008, 2022, 2024–25 and 2025–26. | ||||
| 17 | Rows that condition on the next 12 months tell you what happens if that world arrives, not whether it will. | ||||
Ten selling waves since 2008, computed from the data, with the reasons reported at the time.
| 1 | Every big FPI selling wave since 2008 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2 | Flows and market moves are calculated from the Monthly sheet; the reasons are from Business Standard's reporting at the time. | ||||||||||
| 3 | How to read this sheet. Each row is a period of heavy foreign selling. 'FPI flow' is how many dollars left in that window. 'Nifty in ₹' is how far the index fell. 'Rupee vs $' is how much the rupee lost (negative = weaker). 'Dollar index' and 'US 10y' show what was happening abroad at the same time. The last two columns are the reasons Business Standard reported then, and what brought the money back. Change the From and To dates and every number recalculates. | ||||||||||
| 4 | Episode | From | To | FPI flow ($ bn) | Nifty in ₹ | Rupee vs $ | Dollar index | US 10y (pp) | FPI flow, 12m after ($ bn) | What set it off | What brought them back |
| 5 | Global financial crisis | Jan 08 | Dec 08 | (12.0) | (51.8%) | (18.4%) | 6.0% | (1.79) | 17.5 | US and UK banks failed; funds sold EM to raise cash at home (BS, 29 Dec 2008). | Global stimulus and the May 2009 election; $17.5 bn came back in 2009. |
| 6 | Euro crisis, Indian inflation | Jul 11 | Dec 11 | (1.0) | (18.1%) | (15.5%) | 7.9% | (1.29) | 24.4 | Inflation above 9% for a year, 13 RBI hikes, policy paralysis; Sensex -24.6% in 2011 (BS, 31 Dec 2011). | Rate cuts and reforms; 2012 brought $24.4 bn. |
| 7 | Taper tantrum | Jun 13 | Aug 13 | (3.8) | (8.6%) | (14.2%) | (1.5%) | 0.58 | 21.5 | Fed hinted at tapering; India was one of the 'Fragile Five' with a wide current-account deficit. | RBI's FCNR-B deposit window under Rajan; equity flows were positive for 2013 as a whole. |
| 8 | China devaluation, MAT | Aug 15 | Feb 16 | (6.5) | (18.1%) | (6.7%) | 0.9% | (0.47) | 6.9 | Yuan devaluation, China slowdown, a Fed hike; MAT tax notices to 68 FIIs (BS, Aug–Sep 2015). | RBI cut 50 bp; the Fed held off. |
| 9 | Trump, demonetisation | Oct 16 | Dec 16 | (4.5) | (4.9%) | (1.7%) | 7.3% | 0.84 | 7.8 | US election and Fed hike bets; demonetisation at home. | A weak dollar in 2017. |
| 10 | Fed hikes, crude, rupee 74 | Feb 18 | Oct 18 | (7.8) | (5.8%) | (13.4%) | 9.0% | 0.44 | 11.1 | Fed at 2.25%, US 10y at 3.2%, crude above $84, rupee record low 74.39, LTCG back from Feb 2018. | Crude fell and the rupee recovered from Nov 2018. |
| 11 | Covid crash | Feb 20 | Mar 20 | (8.1) | (28.1%) | (5.0%) | 1.7% | (0.82) | 37.0 | Global dash for cash; record ₹62,000 cr outflow in March 2020. | Central-bank liquidity; record buying in Nov–Dec 2020. |
| 12 | Fed hikes, Ukraine | Oct 21 | Jun 22 | (33.6) | (10.4%) | (5.8%) | 11.1% | 1.44 | 21.2 | Fed hiking 75 bp at a time, dollar at a 20-year high, oil spike; India at about twice EM's P/E. | Commodities cooled, US yields eased; 2023 brought $20.7 bn. |
| 13 | China stimulus, earnings miss | Oct 24 | Feb 25 | (24.9) | (14.3%) | (4.1%) | 6.8% | 0.43 | (7.4) | 'Sell India, buy China', weak Q2 earnings, India's premium to EM near 90–110%. | Brief: a weaker dollar brought buying in Apr–Jun 2025; selling resumed in July. |
| 14 | AI rotation, war, oil | Jul 25 | Oct 26 | (43.1) | (11.2%) | (9.8%) | 4.9% | 0.92 | not yet | Money chased AI chip makers in Korea and Taiwan; single-digit Nifty earnings; US 10y above 5%; Iran war and oil above $100; rupee near 97. | Not yet. A few weeks of buying in Jul–Aug 2026 after the ceasefire. |
| 16 | Dates are inputs: change From and To and the numbers follow. Market moves run from the month-end before 'From' to the month-end of 'To'. Rupee vs $ is negative when the rupee lost value. | ||||||||||
| 17 | Sources for the reasons: Business Standard, 29 Dec 2008; 31 Dec 2011; 26 and 28 Aug 2013; 2 and 30 Sep 2015; 24 Apr 2015; 15 Dec 2016; 5 and 9 Oct 2018; 5 Apr 2020; 28 Dec 2022; 3 and 15 Nov 2024; 13 Oct 2026; 9 Jul 2026. | ||||||||||
| 18 | The pattern: big selling starts with global money getting dearer (Fed, dollar, yields) or a crisis; India's own weak spots (oil, inflation, earnings, an expensive market) decide how deep it goes. Tax changes caused wobbles of a week or two, not trends. | ||||||||||
What the price assumed, and the return a foreign investor needs.
| 1 | Price vs growth, and the bar a foreign investor measures India against | |||||
|---|---|---|---|---|---|---|
| 2 | Blue = sourced figure or assumption (see notes). Black = calculated. | |||||
| 3 | How to read this sheet. The top table is what the market paid and what companies earned. The second table asks two questions. First: what yearly return does a foreign investor need from India, once you add the safe US yield, a premium for risk and the rupee's usual fall? Second: what profit growth did the price quietly assume? The gap between that and what companies delivered is why the money left. | |||||
| 4 | Sep 2024 peak | Now (Sep–Oct 2026) | KOSPI now | Source | ||
| 5 | Nifty forward P/E (next 12 months' earnings) | 22.0 | 17.6 | 5.4 | India: NSE Market Pulse, Jul 2026 (22.0x at the Sep 2024 top); Bloomberg via Business Standard, 12 Sep 2026 (17.6x). Korea: FnGuide, end-Sep 2026. | |
| 6 | India's forward P/E premium to MSCI EM | 110% | 77% | NSE Market Pulse, Jul 2026 (110% at the peak); Bloomberg via Business Standard, 12 Sep 2026 (77%). Long-run average about 57%. | ||
| 7 | Nifty earnings growth delivered (FY25 / FY26) | 3.4% | 4.5% | JM Financial via Business Standard, 11 Jun 2026. Other brokers put FY25 between 1% and 8%; all agree on single digits. | ||
| 8 | Nifty earnings growth expected, FY27 | 15.2% | JM Financial, 20 Aug 2026. Q1 FY27 came in at +12.9%. | |||
| 9 | KOSPI operating profit growth, 2025 | 25.4% | Korea Exchange via JoongAng Daily, 2 Apr 2026. Chip makers' 2026 profits are up several-fold on top. | |||
| 11 | What did the price assume? | |||||
| 12 | US 10-year yield at the time (%) | 3.8% | 5.1% | |||
| 13 | Equity risk premium a foreign investor wants for India | 5.0% | 5.0% | |||
| 14 | Rupee's average yearly fall vs $, 2007–2026 | 4.0% | 4.0% | |||
| 15 | Return needed in rupees to clear that bar | 13.1% | 14.5% | |||
| 16 | Dividend payout ratio of the Nifty | 30% | 30% | |||
| 17 | Earnings yield (1 / forward P/E) | 4.5% | 5.7% | |||
| 18 | Earnings growth the price assumed, every year, forever | 11.7% | 12.8% | |||
| 19 | Earnings growth actually delivered | 3.4% | 4.5% | |||
| 20 | Gap | (8.3%) | (8.3%) | |||
| 22 | Read: at the top the market priced in about 12% growth a year, forever, and got 3–5%. India is cheaper now, but US yields went from 3.8% to above 5%, so the return a foreign investor demands went up as well. | |||||
| 23 | This is why a lower P/E alone has not brought them back: the bar moved too. Change B13 and B16 to test your own assumptions. | |||||
Four worlds for the next 12 months. The blue and yellow cells are yours to change.
| 1 | Next 12 months, Nov 2026 to Oct 2027: four worlds | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2 | Edit the blue cells. Probabilities and India factors are judgement, shaded yellow. Everything else follows from the regression. | |||||||||||||
| 3 | How to read this sheet. This sheet imagines the next 12 months as four possible worlds. For each one you set how emerging-market shares do (a central case and how uncertain it is) and the India factor (India's own story, in $ bn). The model then works out foreign flows month by month: global push + India factor + momentum from the selling already done. The probabilities weight the worlds and must add to 100%. Change any blue or yellow cell: the results, the Simulation and the Cover all update. | |||||||||||||
| 4 | 1. AI boom holds, rates high | 2. US squeeze and AI bust | 3. After the squeeze: Fed cuts | 4. Soft landing, weak dollar | ||||||||||
| 5 | What happens | US 10y stays above 5%, the AI trade keeps winning, oil stays high. Today, continued. | Liquidity tightens, AI firms' cash flow disappoints, chip stocks crash. Risk-off everywhere. | The year after a crash: the Fed cuts, the dollar falls, money goes back to EM. | Inflation cools, the dollar slides, EM rallies, AI still leads. | |||||||||
| 6 | Closest year in the data | 2024 | 2008, 2022 | 2009, 2023 | 2017, 2025 | |||||||||
| 7 | Probability | 35% | 20% | 15% | 30% | How likely I think each world is. My judgement, not data. They must add to 100% (row 12). | ||||||||
| 8 | EM share return over 12 months, central case | 5.0% | (25.0%) | 20.0% | 15.0% | How much emerging-market shares rise or fall over 12 months in that world. | ||||||||
| 9 | EM share return, spread either way (std dev) | 15.0% | 12.0% | 12.0% | 12.0% | How uncertain that is. The simulation draws outcomes around the central case using this spread. | ||||||||
| 10 | India factor ($ bn over 12 months) | (15.0) | 5.0 | 5.0 | (15.0) | India's own story in $ bn a year, beyond what global conditions explain. For scale: 2025 was about −19, 2023 was +15. | ||||||||
| 11 | Why this India factor | Between 2024 (-10) and 2025 (-19): the drag fades slowly. | No precedent. India is the market with little AI in it, so it should be sold less than Korea and Taiwan. | 2009 and 2023 were +6 and +15. Funds are 32% underweight India, so there is room to buy. | 2025 again: a weak dollar helps EM, but the money goes to AI markets. | |||||||||
| 12 | Probabilities add up to | 100% | OK | |||||||||||
| 14 | From the regression (links) | |||||||||||||
| 15 | Constant ($ bn a month) | 0.102 | ||||||||||||
| 16 | Per 100% EM return ($ bn) | 18.790 | ||||||||||||
| 17 | Per $1 bn of prior-12m flow | 0.043 | ||||||||||||
| 18 | Monthly noise, std error ($ bn) | 2.72 | ||||||||||||
| 20 | Month by month (central case of each world) | |||||||||||||
| 21 | Month | Nov 26 | Dec 26 | Jan 27 | Feb 27 | Mar 27 | Apr 27 | May 27 | Jun 27 | Jul 27 | Aug 27 | Sep 27 | Oct 27 | 12 months |
| 22 | Already-sold flow still in the 12m window ($ bn) | (36.0) | (35.6) | (33.1) | (29.1) | (31.6) | (18.9) | (12.4) | (8.9) | (3.8) | (5.9) | (9.0) | (5.3) | |
| 24 | 1. AI boom holds, rates high | |||||||||||||
| 25 | EM return in the month | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | 0.4% | |
| 26 | FPI flow, prior 12 months ($ bn) | (36.0) | (38.2) | (38.4) | (37.1) | (42.3) | (32.5) | (28.5) | (27.3) | (24.4) | (28.6) | (34.0) | (32.8) | |
| 27 | FPI flow ($ bn) | (2.6) | (2.7) | (2.7) | (2.7) | (2.9) | (2.5) | (2.3) | (2.2) | (2.1) | (2.3) | (2.5) | (2.5) | (30.0) |
| 29 | 2. US squeeze and AI bust | |||||||||||||
| 30 | EM return in the month | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | (2.4%) | |
| 31 | FPI flow, prior 12 months ($ bn) | (36.0) | (37.1) | (36.1) | (33.6) | (37.4) | (26.2) | (20.8) | (18.2) | (13.7) | (16.4) | (20.1) | (17.1) | |
| 32 | FPI flow ($ bn) | (1.5) | (1.5) | (1.5) | (1.4) | (1.5) | (1.1) | (0.8) | (0.7) | (0.5) | (0.6) | (0.8) | (0.7) | (12.5) |
| 34 | 3. After the squeeze: Fed cuts | |||||||||||||
| 35 | EM return in the month | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | 1.5% | |
| 36 | FPI flow, prior 12 months ($ bn) | (36.0) | (36.3) | (34.6) | (31.3) | (34.3) | (22.2) | (15.9) | (12.3) | (6.9) | (8.5) | (11.2) | (7.1) | |
| 37 | FPI flow ($ bn) | (0.7) | (0.8) | (0.7) | (0.5) | (0.7) | (0.1) | 0.1 | 0.3 | 0.5 | 0.4 | 0.3 | 0.5 | (1.3) |
| 39 | 4. Soft landing, weak dollar | |||||||||||||
| 40 | EM return in the month | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | 1.2% | |
| 41 | FPI flow, prior 12 months ($ bn) | (36.0) | (38.1) | (38.1) | (36.7) | (41.7) | (31.7) | (27.5) | (26.2) | (23.1) | (27.1) | (32.3) | (30.9) | |
| 42 | FPI flow ($ bn) | (2.5) | (2.6) | (2.6) | (2.5) | (2.7) | (2.3) | (2.1) | (2.1) | (1.9) | (2.1) | (2.3) | (2.3) | (27.9) |
| 44 | Unit check: +$1 bn of India factor, nothing else | |||||||||||||
| 45 | FPI flow, prior 12 months ($ bn) | 0.000 | 0.083 | 0.170 | 0.261 | 0.355 | 0.454 | 0.557 | 0.664 | 0.776 | 0.893 | 1.014 | 1.141 | |
| 46 | FPI flow ($ bn) | 0.083 | 0.087 | 0.091 | 0.095 | 0.099 | 0.103 | 0.107 | 0.112 | 0.117 | 0.122 | 0.127 | 0.132 | 1.274 |
| 49 | Results | |||||||||||||
| 50 | 1 | 2 | 3 | 4 | Weighted | |||||||||
| 51 | FPI flow over 12 months ($ bn) | (30.0) | (12.5) | (1.3) | (27.9) | (21.6) | Net foreign flow over the next 12 months in each world. 'Weighted' averages them using the probabilities. | |||||||
| 52 | Of which, global push ($ bn) | 2.7 | (5.2) | 6.0 | 4.9 | The part that comes from global markets. | ||||||||
| 53 | Of which, India factor including its knock-on ($ bn) | (19.1) | 6.4 | 6.4 | (19.1) | The part that comes from India's own story, including its knock-on as buying or selling persists. | ||||||||
| 54 | Of which, momentum from selling already done ($ bn) | (13.7) | (13.7) | (13.7) | (13.7) | The part that is momentum from the $36 bn already sold in the last 12 months. | ||||||||
| 55 | India factor needed for zero net flow ($ bn) | 8.6 | 14.9 | 6.1 | 6.9 | How big India's own story would need to be for foreign investors to stop selling. Compare with history on Annual. | ||||||||
| 56 | Probability of net FPI buying (simulation) | 1% | 12% | 49% | 2% | 11% | Share of the 1,000 simulated years ending in net buying (from Simulation). | |||||||
| 58 | The global push is the EM-return effect plus the small constant, both scaled by the same knock-on as everything else. Momentum is what the $36 bn sold in the last 12 months keeps doing: the model finds selling persists for some months. | |||||||||||||
| 59 | The break-even row is the most useful one: how large a swing in India-specific sentiment each world needs before foreign investors turn net buyers. Compare it with the Annual sheet's history of the India factor. | |||||||||||||
| 60 | Scenario probabilities are our judgement as of 22 Oct 2026, not data. This is a scenario model for research and teaching, not a forecast to act on and not investment advice. | |||||||||||||
| 62 | Sensitivity: shift every world's India factor by the same amount | |||||||||||||
| 63 | Shift ($ bn) | Weighted 12m flow ($ bn) | ||||||||||||
| 64 | -10 | (34.3) | ||||||||||||
| 65 | -5 | (27.9) | ||||||||||||
| 66 | 0 | (21.6) | ||||||||||||
| 67 | +5 | (15.2) | ||||||||||||
| 68 | +10 | (8.8) | ||||||||||||
| 69 | +15 | (2.5) | ||||||||||||
| 70 | +20 | 3.9 | ||||||||||||
| 71 | +25 | 10.3 | ||||||||||||
1,000 simulated years from the scenarios. First 10 draws shown; all 1,000 are in the file.
| 1 | 1,000 simulated years: how likely are FPIs to turn buyers? | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2 | Each row picks a world by its probability, draws an EM return around that world's central case, and adds the month-to-month noise the regression could not explain. | ||||||||||||||||||
| 3 | How to read this sheet. A Monte Carlo simulation runs the model 1,000 times. Each run picks one of the four worlds using its probability, draws a random but realistic emerging-market return around that world's central case, and adds the month-to-month surprises the model can't explain. The random numbers are fixed on the Draws sheet, so the results don't change every time the file opens. The chart shows how the 1,000 outcomes spread. | ||||||||||||||||||
| 4 | Summary | Spread of outcomes | |||||||||||||||||
| 5 | Average 12-month FPI flow ($ bn) | (21.7) | The average of all 1,000 simulated years. | From ($ bn) | To | Draws | Bucket | ||||||||||||
| 6 | Probability FPIs are net buyers | 11% | Share of runs where foreign investors ended the year as net buyers. | -60 | -50 | 23 | -60 to -50 | ||||||||||||
| 7 | Middle outcome, median ($ bn) | (22.6) | The middle outcome: half the runs were better, half were worse. | -50 | -40 | 111 | -50 to -40 | ||||||||||||
| 8 | 1-in-10 bad case, 10th percentile ($ bn) | (43.1) | A bad year: only 1 run in 10 was worse than this. | -40 | -30 | 196 | -40 to -30 | ||||||||||||
| 9 | 1-in-10 good case, 90th percentile ($ bn) | 0.7 | A good year: only 1 run in 10 was better than this. | -30 | -20 | 217 | -30 to -20 | ||||||||||||
| 10 | -20 | -10 | 189 | -20 to -10 | |||||||||||||||
| 11 | 1. AI boom holds, rates high | 2. US squeeze and AI bust | 3. After the squeeze: Fed cuts | 4. Soft landing, weak dollar | -10 | 0 | 146 | -10 to 0 | |||||||||||
| 12 | Probability of net buying, by world | 1% | 12% | 49% | 2% | 0 | 10 | 77 | 0 to 10 | ||||||||||
| 13 | Average flow, by world ($ bn) | (30.2) | (12.9) | (1.1) | (29.0) | 10 | 20 | 26 | 10 to 20 | ||||||||||
| 14 | Times picked | 311 | 205 | 158 | 326 | 20 | 30 | 7 | 20 to 30 | ||||||||||
| 15 | 30 | 40 | 0 | 30 to 40 | |||||||||||||||
| 16 | Pick thresholds (cumulative probability) | 0% | 35% | 55% | 70% | Outside -60 to +40 | 8 | ||||||||||||
| 31 | Draw | World | EM return, 12m | EM return, month | India factor, month ($ bn) | M1 | M2 | M3 | M4 | M5 | M6 | M7 | M8 | M9 | M10 | M11 | M12 | 12-month flow ($ bn) | |
| 32 | 1 | 1 | (0.2%) | (0.0%) | (1.2) | 1.6 | (0.5) | (8.3) | (5.5) | (1.5) | (6.5) | (1.3) | 1.3 | (2.9) | 1.4 | 1.1 | (2.4) | (23.6) | |
| 33 | 2 | 3 | 19.6% | 1.5% | 0.4 | (0.3) | 0.5 | (4.8) | (1.0) | (5.3) | (1.4) | (0.9) | (5.1) | (3.6) | (1.5) | 5.3 | (4.0) | (22.1) | |
| 34 | 3 | 2 | (23.5%) | (2.2%) | 0.4 | (3.9) | (2.6) | 1.1 | (0.5) | (0.4) | 0.4 | 0.9 | (6.4) | 0.5 | (0.8) | 0.3 | 1.7 | (9.8) | |
| 35 | 4 | 2 | (20.0%) | (1.8%) | 0.4 | 0.5 | (4.5) | (0.2) | (1.6) | (1.7) | 2.6 | (1.9) | (2.3) | (3.9) | (1.4) | 2.1 | (5.1) | (17.5) | |
| 36 | 5 | 2 | (22.0%) | (2.0%) | 0.4 | (8.3) | (3.6) | (0.6) | (3.1) | (5.2) | (2.6) | 1.3 | 2.9 | 3.3 | 1.0 | 0.1 | 1.9 | (12.9) | |
| 37 | 6 | 4 | 1.1% | 0.1% | (1.2) | (5.3) | 0.6 | (0.4) | (0.2) | (2.1) | (4.5) | (5.1) | 0.3 | 0.9 | (5.1) | (4.4) | (2.7) | (28.1) | |
| 38 | 7 | 4 | 25.5% | 1.9% | (1.2) | (3.3) | (4.8) | 0.1 | (0.8) | (4.4) | (4.0) | 1.0 | (5.3) | (2.7) | (0.9) | (1.8) | 0.4 | (26.3) | |
| 39 | 8 | 1 | (20.7%) | (1.9%) | (1.2) | (3.0) | (2.3) | (2.3) | (1.5) | (6.4) | 0.9 | (1.4) | (1.1) | (4.0) | (2.2) | 0.7 | (2.7) | (25.3) | |
| 40 | 9 | 3 | 22.3% | 1.7% | 0.4 | 0.2 | (5.1) | (1.3) | 1.2 | (2.1) | (0.4) | (0.1) | (4.2) | 2.4 | (3.5) | 1.0 | (0.3) | (12.1) | |
| 41 | 10 | 1 | (0.1%) | (0.0%) | (1.2) | (2.5) | (0.7) | (0.5) | (3.7) | (3.3) | (4.8) | (6.3) | 0.1 | (5.2) | (4.6) | (1.9) | (6.5) | (39.8) | |
What moves monthly flows, fitted with Excel's LINEST on 225 months.
| 1 | What moves foreign money in and out of Indian shares | |||||
|---|---|---|---|---|---|---|
| 2 | Ordinary least squares on 225 months, Jan 2008 to Sep 2026, using Excel's LINEST. Dependent variable: FPI net equity flow in the month ($ bn). | |||||
| 3 | How to read this sheet. A regression measures, from 225 months of history, how much foreign flows move when other things move. Each 'coefficient' is the size of a link. The 't-stat' says whether the link is likely real (above 2, or below −2) or could be chance. 'R squared' says how much of the monthly ups and downs the model explains. The plain-English column says what each line means. | |||||
| 4 | Model used in the forecast: EM returns and recent selling | |||||
| 8 | Term | Coefficient | Std error | t-stat | In plain English | |
| 9 | Constant ($ bn a month) | 0.102 | 0.200 | 0.5 | The flow in a month when nothing else moves: a small trickle of money into India. | |
| 10 | EM return in the month | 18.790 | 3.016 | 6.2 | When emerging-market shares rise 1% in a month, foreign investors put about $0.19 bn into India (row 14). They buy when the whole EM mood is good. | |
| 11 | FPI flow over the prior 12 months | 0.043 | 0.012 | 3.6 | Momentum: after a year of selling, they tend to keep selling for a while. Selling doesn't snap back overnight. | |
| 12 | R squared | 0.18 | The model explains about 18% of the monthly ups and downs. The other 82% is India's own story and noise. | |||
| 13 | Std error of a month's flow ($ bn) | 2.72 | A typical month misses the model by about this many $ bn. | |||
| 14 | $ bn per 1% EM move | 0.188 | The coefficient in everyday units: $ bn per 1% move in EM shares. | |||
| 15 | Months used | 225 | Months of data used: January 2008 to September 2026. | |||
| 18 | Does the dollar or the US 10-year add anything once EM returns are known? | |||||
| 19 | Term | Coefficient | Std error | t-stat | In plain English | |
| 20 | Constant | 0.104 | 0.202 | 0.5 | ||
| 21 | EM return in the month | 19.034 | 3.987 | 4.8 | ||
| 22 | FPI flow over the prior 12 months | 0.043 | 0.012 | 3.6 | ||
| 23 | Dollar index change in the month | 0.897 | 10.928 | 0.1 | Once you know how EM shares did, the dollar adds nothing. Its effect works through EM share prices. | |
| 24 | US 10y yield change (pp) | -0.425 | 0.746 | -0.6 | Same for US yields, month to month. They matter, but through the mood of the whole market. | |
| 25 | R squared | 0.18 | ||||
| 27 | A t-stat above 2 (or below -2) means the effect is unlikely to be chance. R squared of about 0.18 means global factors explain under a fifth of any month's flow. | |||||
| 28 | The rest is India-specific (earnings, valuation, the rupee, what else is on offer) and noise. The Annual sheet adds the monthly residuals up by year and calls the total the India factor. | |||||
| 29 | Caveats: flows and EM prices move together within a month, so the EM coefficient is association, not proof of cause. Months are not independent, so treat t-stats as indicative. | |||||
The raw data. Latest 24 months shown; the file has every month since January 2007.
| 1 | Monthly data, Jan 2007 to Oct 2026 | ||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2 | Blue = raw data as published. Black = calculated. One row per month; market levels are month-end closes. | ||||||||||||||||||||||||
| 3 | How to read this sheet. One row per month since January 2007. Blue numbers are raw data: NSDL flows and month-end market prices. Black columns are calculated from them. Hover over any column header for a plain-English note on what it is. | ||||||||||||||||||||||||
| 4 | Month | FPI equity, net ($ mn) | FPI equity, net (₹ cr) | Dollar index (DXY) | US 10y yield (%) | US 3m bill (%) | USD/INR | Nifty 50 | KOSPI | EM proxy: EEM ($) | S&P 500 | Year | FPI flow ($ bn) | EM return, month | FPI flow, prior 12m ($ bn) | DXY change, month | US 10y change (pp) | Nifty in $ | Model: fitted ($ bn) | India factor: residual ($ bn) | FPI flow, 12m to date ($ bn) | FPI flow, next 12m ($ bn) | DXY change, next 12m | EM return, next 12m | Brent crude ($) |
| 219 | Nov 24 | (2,563) | (21,612) | 105.74 | 4.18 | 4.37 | 84.56 | 24,131 | 2,456 | 43.26 | 6,032 | 2024 | (2.6) | (2.7%) | 9.9 | 1.7% | (0.11) | 285.4 | 0.0 | (2.6) | 6.2 | (14.6) | (5.9%) | 25.6% | 72.94 |
| 220 | Dec 24 | 1,828 | 15,446 | 108.49 | 4.57 | 4.21 | 85.79 | 23,645 | 2,399 | 41.82 | 5,882 | 2024 | 1.8 | (3.3%) | 6.2 | 2.6% | 0.40 | 275.6 | (0.3) | 2.1 | 0.1 | (18.9) | (9.4%) | 30.8% | |
| 221 | Jan 25 | (9,043) | (78,027) | 108.37 | 4.57 | 4.19 | 86.65 | 23,508 | 2,517 | 42.72 | 6,041 | 2025 | (9.0) | 2.2% | 0.1 | (0.1%) | (0.00) | 271.3 | 0.5 | (9.6) | (5.8) | (13.8) | (10.5%) | 38.3% | 76.76 |
| 222 | Feb 25 | (3,977) | (34,574) | 107.61 | 4.23 | 4.19 | 87.33 | 22,125 | 2,533 | 43.21 | 5,954 | 2025 | (4.0) | 1.1% | (5.8) | (0.7%) | (0.34) | 253.4 | 0.1 | (4.0) | (10.0) | (7.4) | (9.3%) | 44.8% | 73.18 |
| 223 | Mar 25 | (401) | (3,973) | 104.21 | 4.25 | 4.19 | 85.54 | 23,519 | 2,481 | 43.70 | 5,612 | 2025 | (0.4) | 1.1% | (10.0) | (3.2%) | 0.02 | 275.0 | (0.1) | (0.3) | (14.6) | (19.7) | (4.1%) | 30.0% | 74.74 |
| 224 | Apr 25 | 510 | 4,223 | 99.47 | 4.18 | 4.19 | 85.17 | 24,334 | 2,557 | 43.76 | 5,569 | 2025 | 0.5 | 0.1% | (14.6) | (4.5%) | (0.07) | 285.7 | (0.5) | 1.0 | (13.1) | (26.7) | (1.4%) | 46.2% | 63.12 |
| 225 | May 25 | 2,344 | 19,860 | 99.33 | 4.42 | 4.23 | 85.52 | 24,751 | 2,698 | 45.52 | 5,912 | 2025 | 2.3 | 4.0% | (13.1) | (0.1%) | 0.24 | 289.4 | 0.3 | 2.0 | (7.7) | (32.5) | (0.4%) | 50.7% | 63.90 |
| 226 | Jun 25 | 1,690 | 14,590 | 96.88 | 4.23 | 4.19 | 85.46 | 25,517 | 3,072 | 48.24 | 6,205 | 2025 | 1.7 | 6.0% | (7.7) | (2.5%) | (0.19) | 298.6 | 0.9 | 0.8 | (9.2) | (39.3) | 4.4% | 41.8% | |
| 227 | Jul 25 | (2,052) | (17,741) | 100.03 | 4.36 | 4.24 | 87.71 | 24,768 | 3,245 | 48.56 | 6,339 | 2025 | (2.1) | 0.7% | (9.2) | 3.3% | 0.13 | 282.4 | (0.2) | (1.9) | (15.1) | (35.1) | (0.2%) | 32.0% | 72.53 |
| 228 | Aug 25 | (3,994) | (34,993) | 97.77 | 4.23 | 4.04 | 88.17 | 24,427 | 3,186 | 49.86 | 6,460 | 2025 | (4.0) | 2.7% | (15.1) | (2.3%) | (0.13) | 277.1 | (0.0) | (4.0) | (20.0) | (28.0) | 1.7% | 34.4% | 68.12 |
| 229 | Sep 25 | (2,702) | (23,885) | 97.77 | 4.15 | 3.85 | 88.70 | 24,611 | 3,425 | 53.40 | 6,688 | 2025 | (2.7) | 7.1% | (20.0) | – | (0.08) | 277.5 | 0.6 | (3.3) | (29.6) | (29.1) | 3.8% | 25.1% | 67.02 |
| 230 | Oct 25 | 1,656 | 14,610 | 99.80 | 4.10 | 3.72 | 88.63 | 25,722 | 4,108 | 55.30 | 6,840 | 2025 | 1.7 | 3.6% | (29.6) | 2.1% | (0.05) | 290.2 | (0.5) | 2.2 | (16.7) | 65.07 | |||
| 231 | Nov 25 | (425) | (3,765) | 99.46 | 4.02 | 3.70 | 89.34 | 26,203 | 3,927 | 54.32 | 6,849 | 2025 | (0.4) | (1.8%) | (16.7) | (0.3%) | (0.08) | 293.3 | (0.9) | 0.5 | (14.6) | 63.20 | |||
| 232 | Dec 25 | (2,515) | (22,611) | 98.28 | 4.16 | 3.55 | 89.77 | 26,130 | 4,214 | 54.71 | 6,846 | 2025 | (2.5) | 0.7% | (14.6) | (1.2%) | 0.15 | 291.1 | (0.4) | (2.1) | (18.9) | 60.85 | |||
| 233 | Jan 26 | (3,976) | (35,962) | 96.99 | 4.24 | 3.57 | 91.68 | 25,321 | 5,224 | 59.10 | 6,939 | 2026 | (4.0) | 8.0% | (18.9) | (1.3%) | 0.08 | 276.2 | 0.8 | (4.8) | (13.8) | 70.69 | |||
| 234 | Feb 26 | 2,497 | 22,615 | 97.61 | 3.96 | 3.58 | 91.01 | 25,179 | 6,244 | 62.58 | 6,879 | 2026 | 2.5 | 5.9% | (13.8) | 0.6% | (0.28) | 276.7 | 0.6 | 1.9 | (7.4) | ||||
| 235 | Mar 26 | (12,724) | (117,775) | 99.96 | 4.31 | 3.60 | 94.36 | 22,331 | 5,052 | 56.79 | 6,529 | 2026 | (12.7) | (9.3%) | (7.4) | 2.4% | 0.35 | 236.7 | (2.0) | (10.8) | (19.7) | ||||
| 236 | Apr 26 | (6,474) | (60,847) | 98.08 | 4.39 | 3.58 | 94.92 | 23,998 | 6,599 | 63.99 | 7,209 | 2026 | (6.5) | 12.7% | (19.7) | (1.9%) | 0.08 | 252.8 | 1.6 | (8.1) | (26.7) | 114.01 | |||
| 237 | May 26 | (3,450) | (32,963) | 98.91 | 4.45 | 3.59 | 95.00 | 23,548 | 8,476 | 68.60 | 7,580 | 2026 | (3.5) | 7.2% | (26.7) | 0.8% | 0.06 | 247.9 | 0.3 | (3.8) | (32.5) | 92.05 | |||
| 238 | Jun 26 | (5,157) | (49,340) | 101.19 | 4.42 | 3.73 | 94.79 | 23,866 | 8,476 | 68.41 | 7,499 | 2026 | (5.2) | (0.3%) | (32.5) | 2.3% | (0.04) | 251.8 | (1.3) | (3.8) | (39.3) | 72.92 | |||
| 239 | Jul 26 | 2,123 | 20,200 | 99.80 | 4.75 | 3.68 | 95.68 | 24,384 | 6,595 | 64.09 | 7,490 | 2026 | 2.1 | (6.3%) | (39.3) | (1.4%) | 0.33 | 254.8 | (2.8) | 4.9 | (35.1) | 90.12 | |||
| 240 | Aug 26 | 3,104 | 29,631 | 99.43 | 4.76 | 3.73 | 95.38 | 24,080 | 6,820 | 67.02 | 7,686 | 2026 | 3.1 | 4.6% | (35.1) | (0.4%) | 0.01 | 252.5 | (0.5) | 3.7 | (28.0) | 90.49 | |||
| 241 | Sep 26 | (3,750) | (35,861) | 101.45 | 5.29 | 4.03 | 96.05 | 22,620 | 6,838 | 66.79 | 7,652 | 2026 | (3.8) | (0.3%) | (28.0) | 2.0% | 0.54 | 235.5 | (1.2) | (2.6) | (29.1) | 103.53 | |||
| 242 | Oct 26 | (5,261) | (49,870) | 101.59 | 5.15 | 4.06 | 94.79 | 22,652 | 5,534 | 63.57 | 7,757 | 2026 | (5.3) | (4.8%) | (29.1) | 0.1% | (0.15) | 239.0 | (36.0) | 112.22 | |||||
| 244 | Sources: FPI flows: NSDL FPI Monitor, year-wise report, equity, net (fpi.nsdl.co.in), read 22 Oct 2026. October 2026 is to the 21st and estimated (see the comment on its cell). | ||||||||||||||||||||||||
| 245 | Market data: Yahoo Finance monthly closes for DX-Y.NYB, ^TNX, ^IRX, INR=X, ^NSEI, ^KS11, EEM, ^GSPC, BZ=F (Brent), downloaded 22 Oct 2026. The Oct 2026 row is the 22 Oct close. EEM (iShares MSCI EM ETF, price only) stands in for MSCI EM. | ||||||||||||||||||||||||
| 246 | Fitted and residual columns use the model on the Regression sheet, Jan 2008 to Sep 2026. Next-12m columns stop where a full 12 complete months is not yet available. | ||||||||||||||||||||||||
The record
I read 62 Business Standard articles on every foreign selling wave, from the 2008 crash to this month. These 13 passages carried the argument. They are quoted as published; the highlighting is mine.
29 Dec 2008
FIIs pull out $13 bn in '08, may start buying in H2-09
During a year when Indian stock market lost its valuation by more than half, foreign institutional investors (FIIs) have pulled out an estimated $13 billion from domestic bourses.
The headline's $13 billion was an estimate made before the year closed; NSDL's final count is $11.97 billion.
31 Dec 2011
Sensex sheds 24.6% in 2011 outlook for New Year bleak
…a combination of near double-digit inflation, high interest rates, slowing domestic growth and policy inaction turned off investors already shaken by global headwinds.
26 Aug 2013
Sensex back to March 2009 for FIIs
So, while in absolute terms, the Sensex has only dropped to 18,558.13, for a foreign investor, the losses are about 57 per cent.
24 Apr 2015
Notices to 68 FIIs to recover Rs 602 cr tax towards MAT: Sinha
…the Income Tax department had slapped notices on foreign portfolio investors (FPIs), saying they have to pay 20 per cent MAT on untaxed capital gains made by them over the past three years.
5 Oct 2018
Markets stare at worst-ever FPI sell-off in 2018: All you need to know
The sell-off this year has come when the Fed has embarked upon delivering the biggest annual rate hike in more than a decade.
28 Dec 2022
FPIs sets record, offloaded Rs 1.2 trillion worth of Indian stocks in 2022
Apart from global monetary tightening, volatile crude, rising commodity prices along with Russia and Ukraine conflict led to an exodus of foreign money in 2022.
3 Nov 2024
FPIs withdraw record Rs 94,000 cr from Indian equities in October
…the worst-ever month in terms of outflows, triggered by the elevated valuation of domestic equities and attractive valuations of Chinese stocks.
3 Mar 2025
Will cutting capital gains tax alone bring back FIIs to Indian equities?
Merely cutting or even abolishing capital gains tax… will not be enough to bring foreign investors back to the Indian stock market, analysts believe.
23 Dec 2025
Year-ender 2025: India stocks lag EM, Asia in worst showing in decades
…the index's underperformance versus the MSCI Emerging Markets Index is the worst seen since 1993, according to Bloomberg data.
13 Apr 2026
Why do FPIs keep selling? From high valuations to a changed narrative
This amounts to almost 1 per cent of market capitalisation, exceeding the selling witnessed even during the global financial crisis (GFC).
12 Jun 2026
Deep dive: Hits & misses of Q4FY26 earnings season - JM Financial
…at the start of FY26, earnings were expected to grow 12–15 per cent, whereas actual growth was only 4.5 per cent…
11 Oct 2026
Nifty 50 trails Asian peers; Kospi and Taiex nearly double, shows data
South Korea and Taiwan rode a chipmaker rally on artificial-intelligence (AI) demand.
13 Oct 2026
Record FPI outflows top ₹3 trillion in CY26, nearly double 2025 tally
India, he said, has very little to offer on that front, so it has become the market FPIs sell to fund those positions. "The weak rupee is adding to it, as it cuts dollar returns…"
NSDL FPI Monitor, monthly flows January 2007 to October 2026 (October estimated to the 21st). Yahoo Finance month-end closes for the dollar index, US yields, the rupee, the Nifty 50, the KOSPI and the won, the MSCI EM fund (EEM), the S&P 500 and Brent crude. NSE Market Pulse and Bloomberg via Business Standard for valuations; JM Financial for Nifty earnings; the BofA Asia Fund Manager Survey for positioning.