The two numbers
Every trading day, two related but distinct numbers get published for each stock:
- Traded volume — total shares that changed hands during the day, including intraday round-trips
- Delivery volume — shares that were actually transferred to demat accounts (i.e. positions held overnight)
Their ratio — delivery volume divided by traded volume, expressed as a percentage — is the delivery percentage.
Why volume alone is misleading
Imagine a stock trades 10 million shares in a day. Sounds like high activity. But:
- Scenario A — 8 million of those were intraday trades (bought and sold by speculators within the day). Only 2 million were genuine position changes.
- Scenario B — 6 million were genuine position changes. Only 4 million were intraday.
Same headline volume. Very different market state. In Scenario A, traders dominate; the move could reverse the next day. In Scenario B, real positions changed hands; the move has more conviction behind it.
Volume alone can't distinguish these. Delivery percentage does.
Reading the combination
| Volume | Delivery % | Interpretation |
|---|---|---|
| High | High | Genuine institutional activity. Strongest signal. |
| High | Low | Speculative churn. Often fades within days. |
| Low | High | Low overall activity but what little there was, was real. Often pre-cursor to bigger moves. |
| Low | Low | Dead day. No meaningful information. |
The high-volume, high-delivery combination is what the platform's signals and conviction engine watch for. It's the textbook footprint of institutional buying.
What "high" actually means
There's no universal threshold; each stock has its own baseline. Useful frames:
- Volume ratio (today / 20-day average) > 2x — high volume by the stock's own standard
- Delivery percentage 1.5x the stock's 30-day average — high delivery by the stock's own standard
The platform's scanner signals — specifically volume spike — use ratios rather than absolute numbers for this reason.
How conviction uses both
The conviction engine has two distinct core factors that use these signals differently:
- Delivery factor uses delivery percentage combined with the delivery-day's price change. It answers: was there real position-building or distribution?
- Volume factor uses today's volume ratio combined with today's price change. It answers: is unusual activity happening right now?
These don't double-count. Volume can spike without delivery (intraday traders piling in); delivery can be high without exceptional volume (slow institutional buying over many days). The conviction engine reads both signals as independent factors and weights them accordingly.
When delivery data updates
Volume updates in real time (live during market hours). Delivery percentage is end-of-day only — it's published in the NSE/BSE bhavcopy file after market close, typically by 18:00–19:00 IST.
This means:
- During market hours: only intraday volume is visible
- Post-market (after ~19:00): delivery numbers for the day finalise
- The conviction engine recalculates with the new delivery data after market hours
A stock's conviction score can shift meaningfully between the market close and the evening update on a high-delivery day — the delivery factor was unknown at close and locked in afterwards.
A practical use
When you see a stock up 4% during market hours on heavy volume, hold off on conclusions until delivery prints. The next morning, the conviction strip on the stock page will show whether the move had delivery confirmation.
- Delivery confirms — the move was real. Conviction often shifts toward Bullish.
- Delivery doesn't confirm — the move was speculative. Conviction often stays flat or weakens.
This single check separates real moves from speculative ones with high reliability.
Open Delivery analysis