Delivery Percentage in Stocks: What It Does Not Tell You
Delivery percentage is one of the most quoted numbers in Indian retail research and one of the most consistently misread. The misreading is not subtle — it is a mistake about the denominator — and once you see it, most of what gets written about “high delivery stocks” falls apart.
This chapter covers what the number is, the exact exchange fields it comes from, the three things that structurally determine it before any investor decides anything, and what it can honestly support. If you want the current list rather than the method, the high delivery percentage screen is updated from the exchange file each evening.
What delivery percentage actually measures
When you buy a share intraday and sell it before the close, nothing moves into your demat account. The two trades net off in settlement. When you buy and hold past the close, the shares are delivered to you.
Delivery percentage is the share of a day’s traded quantity that ended in the second category:
DELIV_QTY ÷ TTL_TRD_QNTY × 100 = DELIV_PERThose are the literal column names in NSE’s sec_bhavdata_full file, published after each close. The exchange computes the figure; it is not a derived estimate.
The denominator is the day, not the company
Here is the mistake. “Tata Motors had 78% delivery” does not mean 78% of Tata Motors is held by anybody in particular. It means 78% of the shares that traded that day were delivered.
The day’s volume is usually a tiny fraction of a company. A stock can show 95% delivery on a session in which 0.02% of its equity changed hands. The percentage is large; the thing it is a percentage of is small.
This is why “high delivery percentage” lists that rank purely on the percentage are close to meaningless at the top. Without a volume or turnover floor, the highest figures belong to securities in which almost nothing traded — and 100% of almost nothing is 100%.
| What people think it measures | What it measures |
|---|---|
| Share of the company held long-term | Share of one day’s volume that settled as delivery |
| Investor conviction | Whether trades were closed out before the bell |
| Institutional buying | Nothing about who traded — the data is anonymous |
| A property of the stock | A property of a single session |
Three things that decide it before any investor does
This is the part almost no page on the subject states, and it changes which numbers are worth reading at all.
1. The series can make delivery compulsory
NSE trades securities in series, and the series is printed in the same file as the delivery figure.
BE and BZ are trade-to-trade. In these, intraday square-off is not permitted by exchange rule. There is no choice to make: every trade must be delivered. NSE does not even publish a delivery percentage for them — on the session of 23 September 2026 it printed a dash for all 260 such securities that traded. If you ever see a delivery list topped by BE or BZ names, the number did not come from NSE’s own field.
SM and ST are the SME platform. These are real shares in real companies, but they trade in large minimum lots among few participants, so delivery runs structurally high — the median SME row settles far above the median main-board row. High delivery there is the norm, not a signal.
EQ is the main board, and it is the only series where the number carries the meaning most readers assume.
2. ETFs settle at nearly 100% by construction
An exchange-traded fund is listed on the EQ series alongside companies, so a naive screen treats it as a stock. But a liquid debt or liquid ETF is bought to park cash, not to trade intraday, and settles at close to 100% delivery every single session.
On 23 September 2026, applying a ₹5 crore turnover floor and a 90% delivery cut to the whole NSE list returned fourteen securities. Exactly one — Westlife Foodworld — was a company. The rest were ETFs, REITs, InvITs and a government bond. A delivery screen that does not exclude them is, in practice, a list of ETFs.
3. Low volume inflates it
The fewer shares that trade, the more likely all of them were delivered. Any delivery ranking without a liquidity floor is sorted, in effect, by illiquidity.
What a normal delivery percentage looks like
Almost every article quotes a threshold — “above 60% is strong” — without saying what ordinary looks like. Measured across the 1,101 NSE shares liquid enough to rank on 23 September 2026:
| Percentile | Delivery |
|---|---|
| Median | 46.0% |
| 75th | 54.7% |
| 90th | 62.0% |
| 95th | 66.9% |
| 99th | 74.0% |
So a stock at 60% is above average and entirely unremarkable. A stock above 74% is genuinely in the top percentile of that session. Any single number offered as a universal threshold, without the distribution behind it, is not worth much — and the distribution shifts from session to session.
Why NSE and BSE print different numbers for the same stock
Each exchange reports only its own trades. A dually listed stock has two order books, two volumes and two delivery quantities, so the two percentages are computed on different denominators and will rarely agree.
Neither is wrong. When the two diverge sharply, it usually reflects where the intraday trading concentrated that day, which is itself a fact about the stock — but it is not an error to reconcile.
What it can and cannot support
It can reasonably support: an observation that, on a particular day, an unusually large share of trading was carried overnight rather than closed out. Read beside turnover, that distinguishes a day driven by intraday churn from one where positions were actually taken.
It cannot support: a claim about who bought. Delivery data is anonymous and carries no counterparty information whatsoever. If the question is who, the bulk and block deal disclosures name the client — that is a different exchange file, and the right one for the question.
It also cannot support a conclusion drawn from one session. A single day’s figure moves with expiry, index rebalances, results dates and a single large order. A stock whose delivery percentage is consistently elevated across many sessions is describing something; one day is noise.
The honest summary
Delivery percentage answers a narrow question accurately: of what traded today, how much was carried. It says nothing about the company, nothing about who traded, and — because its denominator is one day’s volume rather than the share capital — nothing about ownership.
Used with a liquidity floor, with the series in view, with ETFs excluded and across a run of sessions rather than one, it is a modest and useful description of participation. Used as a ranking on its own, it mostly discovers illiquid securities and exchange-traded funds.
Related reading
- High delivery percentage stocks today — the current screen, with the session’s full distribution printed above it.
- Bulk and block deals today — named counterparties, which delivery data never carries.
- High volume stocks today — participation measured against a stock’s own average.
- Upper circuit stocks today — the same exchange file, read for price bands.