FII and DII data: how the daily numbers work
7 min read · data: NSE provisional FII/DII cash-market data, 248 sessions Jan 2024 – Sep 2026 (with gaps) · descriptive study, not advice
Every evening, business channels flash two numbers: FIIs sold ₹2,000 crore, DIIs put in ₹3,000 crore. These are among the most watched figures in Indian markets. This guide explains who these groups are, where the numbers come from, and what our own data says they can and cannot tell you.
It is a reading guide, not advice about any trade.
Who are FIIs and DIIs?
- FII stands for Foreign Institutional Investor. The official term today is FPI, Foreign Portfolio Investor. These are overseas funds, banks, insurers and other investors registered with SEBI to invest in Indian markets.
- DII stands for Domestic Institutional Investor: Indian mutual funds, insurance companies, banks, and pension and provident funds that invest through them.
Both groups are large. On the 248 days in our data, the typical day’s net figure (the size, ignoring the direction) was about ₹2,000 crore for FIIs and ₹2,900 crore for DIIs.
Where the evening numbers come from
After the close, NSE publishes provisional figures for FII/FPI and DII trading in the cash market: the total value they bought, the total value they sold, and the difference (the net). The figures combine trades on NSE, BSE and MSEI, in ₹ crore.
Three details matter:
- Provisional means early. The numbers come from exchange trade data on the same day and can differ from final figures.
- Cash market only. They leave out futures and options, where foreign funds are often very active.
- There is a second source. NSDL publishes FPI flow data too. It works on a different basis (it includes primary-market deals like IPOs and placements, and counts by reporting date), so NSDL’s number for a day often differs from the exchange’s evening number. Neither is wrong. They answer different questions.
On our pages you will see these flows written as “put in” (net positive) or “took out” (net negative), with ▲ and ▼ arrows.
What 248 days of our data show
We hold both numbers for 248 trading days between January 2024 and September 2026 (our collection has gaps, so this is not every day). We lined them up with the Nifty 50’s daily move.
1. FIIs and DIIs usually move in opposite directions. On 77% of days, one group put money in while the other took money out. When foreign funds take money out, domestic funds often absorb the shares, and the reverse.
2. Flows and the same day’s Nifty move together. On days FIIs were net positive, Nifty closed higher 66% of the time. On days they were net negative, it closed higher 44% of the time. The link is real but loose (a correlation of about 0.31). It also does not show which causes which: a rising market can attract flows just as flows can lift a market.
3. Today’s flow says very little about tomorrow. After an FII net-positive day, Nifty rose the next day 54% of the time. After a net-negative day, 46%. Across all days it was 48%. The next-day correlation was about −0.05, close to zero. In our sample, the evening FII number was mostly a description of the day that just ended.
4. Foreign funds were net negative on most days in this period (about 73% of our days). That reflects a long stretch of foreign outflows in 2024–2026, not a permanent rule. A different period could look very different.
How to read the numbers well
- Read both together. “FIIs took out ₹3,000 crore” means less if DIIs put in ₹3,500 crore the same day.
- Look at a run of days, not one. A single day is noise. A month of steady flow in one direction says more about the backdrop.
- Remember what is missing. Derivatives positions, primary-market deals and retail investors are not in the evening cash number.
- Do not map it to one stock. The figure is a market total. It does not say which stocks the funds traded.
Our Daily Wrap reports the day’s flows in plain words, and the Market page shows the backdrop.
The limitation
Our sample is 248 days with gaps, and it covers a period dominated by foreign outflows. The comparisons are simple counts, not a model. Flow data describes what institutions did. It does not predict what the market will do next.
Related: How to read market breadth · Bulk vs block deals