01 · Frozen record
Scores stay frozen once recorded.
The evaluation window shown here was scored under methodology v2.12. That record is frozen and is not retroactively rescored.
Public evidence ledger · Evidence building · Insufficient admitted history — publishing resumes when the cohort reaches the N≥30 floor (~Oct 19, 2026). · as of
An auditable track record.
Insufficient completed 3M history to publish a headline edge claim.
This page reads the latest persisted track-record run. It uses score bands as recorded on the snapshot date, completed future windows only, and a date-matched SPY benchmark.
Evidence window: May 16, 2026-May 21, 2026 + Jul 21, 2026 onward.
Reader outage: May 22, 2026-Jul 20, 2026. That interval remains excluded. Accrual resumes with rows recorded from Jul 21, 2026 onward. Nothing was backfilled.
Evidence protocol
Recorded live. Never backfilled.
02 · Reader outage
The outage hit the display, not the data.
May 22-July 20 remains outside the evidence cohort. New rows accrue prospectively from July 21. Nothing in the outage interval is reconstructed.
03 · Publication rule
We publish only completed windows.
Only completed windows publish. Estimates below N 30 stay withheld. At current sample sizes, band separation is not yet statistically meaningful.
Evaluated rows
1,200
Unique tickers
240
Sample window
May 16, 2026 - May 21, 2026
Benchmark
SPY
Band monotonicity
Average forward return by band
Low N marked where N < 30
1Mfirst resumed cohort completes Aug 20, 2026; the table publishes that band only at N 30 or above.
3Mfirst resumed cohort completes Oct 19, 2026; the table publishes that band only at N 30 or above.
6Mfirst resumed cohort completes Jan 17, 2027; the table publishes that band only at N 30 or above.
| Band | 1M mean / excess / hit | 3M mean / excess / hit | 6M mean / excess / hit |
|---|---|---|---|
STRONG N 5 | Insufficient History N 5 below the minimum 30 observations. Estimate withheld. SPY N 5 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded |
BUY BIAS N 157 | +2.11% Excess +1.25% · Hit 50% · Beat SPY 45% N 157 · SPY N 157 · Sharpe-style 0.07 · Sortino-style 0.11 | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded |
NEUTRAL N 997 | +1.87% Excess +0.86% · Hit 53% · Beat SPY 48% N 997 · SPY N 997 · Sharpe-style 0.06 · Sortino-style 0.10 | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded |
SELL BIAS N 39 | +0.09% Excess -0.50% · Hit 54% · Beat SPY 46% N 39 · SPY N 39 · Sharpe-style -0.07 · Sortino-style -0.09 | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded |
AVOID N 2 | Insufficient History N 2 below the minimum 30 observations. Estimate withheld. SPY N 2 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded | Insufficient History N 0 below the minimum 30 observations. Estimate withheld. SPY N 0 · Open windows excluded |
1M ordering
Monotonic in evaluated bands
BUY BIAS / NEUTRAL / SELL BIAS
3M ordering
Insufficient history
No checked bands yet
6M ordering
Insufficient history
No checked bands yet
Completed cohort curve
STRONG band versus SPY
Equal-weight completed 3M cohorts are compounded from 100. This is a research curve, not an executable portfolio, and losing stretches stay visible.
Insufficient completed curve
The STRONG basket versus SPY curve appears after enough completed 3M cohorts exist.
Signal decay
STRONG band by holding window
1M · N 5Insufficient
Insufficient history
3M · N 0Insufficient
Insufficient history
6M · N 0Insufficient
Insufficient history
Public evidence ledger
Recent evaluated hits and misses
Completed STRONG-band windows only
No completed STRONG-band examples are old enough to evaluate yet. Open windows stay excluded.
These examples are selected from completed STRONG-band observations only. Both positive and negative outcomes remain visible.
Super Score evidence
Convergence track record
Super Score convergence history is not yet persisted with completed forward-return windows, so this panel stays in evidence-building mode.
Insufficient persisted Super Score history
Honesty panel
How to read this page
- benchmark
- SPY adjusted forward returns are matched to every matured score date; an incomplete match blocks a new publication.
- dividends
- Uses adjusted daily bars where the upstream price source provides them.
- risk free
- Risk-free rate is 0% for the displayed cohort Sharpe/Sortino-style ratios.
- weighting
- Band cohorts are equal-weighted by ticker snapshot observation.
- open windows
- Only fully elapsed 1M, 3M, and 6M windows are evaluated.
- survivorship
- Historical snapshot rows remain in the cohort even if a ticker later leaves the current universe.
- null handling
- Open score windows are excluded. A missing date-matched benchmark for any matured score window blocks a new publication, so the prior result stands.
- point in time
- Each row uses the Veridion band stored on that snapshot date.
- transaction costs
- Gross research returns; no transaction costs or slippage are deducted.
Past performance does not guarantee future results · not investment advice. This is a measurement of historical score behavior, not a promise or transaction instruction.
Traceability
- Open forward-return windows are excluded.
- Rows without realized future prices are excluded, never filled.
- SPY benchmark comparisons require a matching benchmark row on the same score date.
- Full methodology: Veridion Score methodology.