On 2026-08-03 NSE and BSE replaced the last-30-minute VWAP close with a closing auction for F&O names. Nobody has history for this mechanism, so this page is the history: the dislocation between each stock's 15:15 price and its official close, every session, both exchanges, from day one. Before each expiry session exists I commit measurement bars publicly; the number publishes here either way. The page regenerates automatically every evening from the same pipeline, whether the result flatters the forecast or not.
Dislocation = median |official close vs 15:15 last trade| across the F&O universe. The ratio is against a diffusion benchmark: the auction print sits 20 minutes after the 15:15 anchor, and prices move in 20 minutes even in a healthy market. The benchmark asks how big the gap should be from ordinary price movement alone; a ratio near 1.0 means the auction adds nothing on top. Verdicts require the raw number inside the committed band and the diffusion ratio below 1.3 (the materiality line, fixed here ahead of every future session).
| date | index | bars committed | measured | diffusion ratio | verdict |
|---|---|---|---|---|---|
| 2026-08-04 | NIFTY | no bars on file for this session (pre-streak or committed only in-thread) | 40.5bp | 1.93x | — |
| 2026-08-06 | SENSEX | no bars on file for this session (pre-streak or committed only in-thread) | 12.4bp | 0.56x | — |
| 2026-08-11 | NIFTY | 15-22bp raw, diffusion ratio ~1.0 (committed 2026-08-10 in-thread, before the print) | 21.6bp | 1.08x | INSIDE |
| 2026-08-13 | SENSEX | 15-25bp raw, diffusion ratio ~1.0 (committed 2026-08-12 in-thread, before the session existed) | 17.9bp | 0.83x | INSIDE |
| date | exchange | median |bp| | p90 |bp| | names |
|---|---|---|---|---|
| 2026-08-03 | BSE | 9.6 | 72 | 208 |
| 2026-08-03 | NSE | 52.7 | 114 | 210 |
| 2026-08-04 | BSE | 16.3 | 77 | 208 |
| 2026-08-04 | NSE · expiry | 40.5 | 97 | 210 |
| 2026-08-05 | BSE | 11.8 | 55 | 208 |
| 2026-08-05 | NSE | 27.0 | 71 | 210 |
| 2026-08-06 | BSE · expiry | 12.4 | 69 | 208 |
| 2026-08-06 | NSE | 18.5 | 63 | 210 |
| 2026-08-07 | BSE | 15.4 | 91 | 208 |
| 2026-08-07 | NSE | 17.5 | 54 | 210 |
| 2026-08-10 | BSE | 14.3 | 65 | 208 |
| 2026-08-10 | NSE | 20.8 | 58 | 210 |
| 2026-08-11 | BSE | 12.7 | 65 | 208 |
| 2026-08-11 | NSE · expiry | 21.6 | 52 | 210 |
| 2026-08-12 | BSE | 17.0 | 71 | 208 |
| 2026-08-12 | NSE | 59.2 | 158 | 210 |
| 2026-08-13 | BSE · expiry | 17.9 | 65 | 208 |
| 2026-08-13 | NSE | 24.2 | 60 | 210 |
| 2026-08-14 | BSE | 15.6 | 60 | 208 |
| 2026-08-14 | NSE | 16.9 | 45 | 210 |
| 2026-08-17 | BSE | 15.5 | 58 | 208 |
| 2026-08-17 | NSE | 22.3 | 55 | 210 |
1. Bars for a future expiry are committed publicly before the session exists. Changing them afterwards restarts the claim at zero. Standing commitment, ratified 2026-08-14: every future expiry on either exchange is called at 15-25bp raw and diffusion ratio ~1.0 until a revision is published here before the session it first applies to.
2. The number publishes whatever it is. An OUTSIDE verdict stays on this page forever.
3. Prices are last trades at 15:15 from minute bars against raw exchange bhavcopy closes; corporate-action jumps are excluded above 2000bp. Fuller honesty labels live in the methodology write-ups.
4. The series is append-only. History does not get revised; errors get corrected in the open with the correction noted.
Measurement, not advice. Nothing here is a recommendation to trade anything. Method and the rest of the record: the audit practice · the graveyard. Contact via the site.