← All validation receipts
Public Validation Receipt

Veritas — 2026-08-03 — Day 40 of public tracking

Post-market outcome: predicted first-15-minute bands vs actual first-15-minute trade. Misses are included. This record is permanent and unedited.

Observed: 2026-08-03T08:46:18.562990+05:30

Bands held
3/3
Direction correct
0/3
NIFTY 50
BAND HIT
Predicted band (low)23,708.82
Predicted band (high)24,029.88
Predicted midpoint23,869.35
Actual open (09:15-09:25 Median)23,904.80
Band resultHIT ✓
Direction (bias: NEUTRAL)Not asserted
Confidence44.7%
Failure modeNeutral call — direction not asserted, band used as range only

The NIFTY 50 first-15-minute price (23,904.80) landed near the midpoint of the predicted band (23,708.82 – 24,029.88). The band held — the model's first-15-minute range estimate was consistent with the actual first-15-minute trade. Directional bias was NEUTRAL — the model did not assert a direction for this session. On neutral calls the band alone is the accountability measure, and direction is explicitly not part of the validation score. Model confidence was very low at 45% — the model effectively signalled that it was unsure about this session. On sub-48%-confidence days the band width tends to be wider and the directional read carries less weight in the overall score.

BANK NIFTY
BAND HIT
Predicted band (low)55,892.71
Predicted band (high)57,020.99
Predicted midpoint56,456.85
Actual open (09:15-09:25 Median)56,796.95
Band resultHIT ✓
Direction (bias: BULLISH)Incorrect ✗
Confidence26.2%
Failure modeDirection miss — band held but bias was wrong

The BANK NIFTY first-15-minute price (56,796.95) landed near the upper end of the predicted band (55,892.71 – 57,020.99). The band held — the model's first-15-minute range estimate was consistent with the actual first-15-minute trade. The directional bias was BULLISH, but the actual first-15-minute move was bearish — a direction miss (failure mode: SIGN_MISS). The range still held, so the band served its purpose as the primary accountability surface even though the directional signal was wrong. Model confidence was very low at 26% — the model effectively signalled that it was unsure about this session. On sub-48%-confidence days the band width tends to be wider and the directional read carries less weight in the overall score.

SENSEX
BAND HIT
Predicted band (low)75,872.11
Predicted band (high)76,906.95
Predicted midpoint76,389.53
Actual open (09:15-09:25 Median)76,515.10
Band resultHIT ✓
Direction (bias: NEUTRAL)Not asserted
Confidence33.3%
Failure modeNeutral call — direction not asserted, band used as range only

The SENSEX first-15-minute price (76,515.10) landed near the midpoint of the predicted band (75,872.11 – 76,906.95). The band held — the model's first-15-minute range estimate was consistent with the actual first-15-minute trade. Directional bias was NEUTRAL — the model did not assert a direction for this session. On neutral calls the band alone is the accountability measure, and direction is explicitly not part of the validation score. Model confidence was very low at 33% — the model effectively signalled that it was unsure about this session. On sub-48%-confidence days the band width tends to be wider and the directional read carries less weight in the overall score.

About This Receipt

This is a permanent, unedited record of what AION Veritas predicted before the 2026-08-03 market open versus what the market actually did. Validation price is the 09:15–09:25 IST median first-15-minute window. Receipts are published the same day, after market close, and are never revised.

How Veritas works → • Full receipt archive →