Delivery percentage is the share of a day's traded quantity that actually changed hands and settled into someone's demat account, rather than being bought and sold again the same day. If a stock traded ten lakh shares and four lakh were delivered, delivery was 40%. The other 60% was intraday trading that netted out before the close.
Volume tells you how busy a stock was. Delivery tells you how much of that activity was people actually taking the shares home. The exchange publishes it every evening in its end-of-day file, and we show it on every F&O stock page - RELIANCE, for instance, or any company in the stock directory.
Why it is worth reading separately from volume
A day with enormous volume can mean two very different things.
It can mean a crowd of intraday traders chasing a move, buying and selling the same shares several times over and closing flat by 3:30. That lifts volume and leaves ownership exactly where it was.
Or it can mean shares genuinely moving from one set of owners to another, which is what a real change of view about a company looks like.
Price and volume alone cannot tell those apart. Delivery can. High delivery on a strong day says the buying was taken home. Low delivery on the same day says it was mostly traded and handed back.
How we show it: against the stock's own normal
A delivery percentage means little in isolation, because normal varies enormously between stocks. A large, widely held company where most volume is long-term investors might deliver 50% on an ordinary day. A heavily traded favourite of intraday traders might deliver 20%. Neither number is good or bad.
So we compare each day with that same stock's average over the previous 20 sessions, and label it:
- High when today's delivery is at least 30% above its own 20-session average.
- Low when it is at least 30% below it.
- Normal in between.
The label is a convenience for reading the page, not a statistical finding. The two numbers it is built from sit right beside it, so you can judge the gap yourself.
The honest limits
It says shares were taken home, not why. A fund rebalancing into an index, a promoter adding to a stake, and a large block changing hands all raise delivery without saying anything about where the price goes next. Large deals are listed separately on Bulk and Block Deals, which often explains an odd delivery day.
It is less meaningful on thin days. A stock that barely traded can show a high delivery percentage on very little quantity. Read it together with volume, never instead of it.
It does not tell you direction on its own. High delivery on a falling day means sellers were getting out for real. Pair it with what the price did.
It is end of day. The exchange publishes delivery after the close, so during market hours the page shows the last session's figure, labelled with its date.
What is a good delivery percentage?
There is no universal good number. Compare the stock with its own recent average rather than with other stocks. A day well above that stock's normal delivery is the useful signal; a high percentage that is ordinary for that particular company is not.
Does high delivery mean the price will go up?
No. It means more of the day's trading was genuine change of ownership. That is true of determined sellers as much as determined buyers. It is a description of how the trading happened, not a forecast.
Is delivery percentage available for F&O contracts?
No. Delivery is a cash-market concept: shares settling into a demat account. It is published for the stock itself. Futures and options are settled differently and have no delivery figure of their own.
What you can and cannot test with it
Delivery percentage is not one of the conditions our backtester can test today, and we would rather say so than imply otherwise. The nearest thing it can test is participation measured by volume: "buy when volume is well above its own average on a rising day" is a rule you can replay across years of end-of-day data, with every charge deducted. The free plan covers it: Unlimited backtests on the free plan. What Free limits is DATA - three years, the Nifty 50, no options - never the number of runs The history behind it: Ten years of end-of-day equity historyThe universe holds no delisted companies, so every long-run result carries survivorship bias. We say so on the results page.
EdgeTest is analysis only. It does not give investment advice, does not place orders, and is not registered with SEBI