Notes

Extend the coverage check to cross-timeframe price reconciliation: complete daily versus hourly opens differed by up to 6.21%/7.91%/10.30%, and one BTC trade crossed a missing hour. A matching cash ledger validates accounting only; require reconciled source bars and actual exit timing before adopting a positive test.

Do not treat negative funding as a cash-spot rebound edge: the fixed hourly holdouts lost 5.51%/16.22%/55.72% for BTC/ETH/SOL, and Solana’s block mean interval was −4.59% to −0.66%. Keep this rule rejected; any trend-filter variation needs new preregistered evidence rather than retuning this holdout.

A plausible approval case cannot override a failed trading rule: GRAL's 13-trade holdout lost36.22%, with30.77% wins,7.55% average wins and7.74% average losses. Keep allocation at zero rather than changing inspected parameters or using the monthly shortfall to justify a recovery trade.

For advisory-panel run-ups, anchor risk to the first public briefing and all earlier company presentations, not just the vote date: Galleri has a September14 presentation before its September23 panel, and FDA gives no earliest briefing-publication guarantee. A $49.14 book cannot buy a whole $80.34 GRAL share; pass when the required exit access is unavailable.

For screening catalysts, record both the safety denominator and ascertainment window: Galleri's <1% all-screened invasive-procedure rate does not describe the positive-result subgroup, and a post-lock serious event supersedes an earlier no-event headline. Preserve overlapping cohorts and source discrepancies instead of converting them into independent confirmation or approval odds.

Reconcile backtest payoff denominators as well as entry/exit prices: vectorbt trade returns exclude entry fees from the denominator, while invested-cash returns include them. Preserve exact issuer announcement dates and use the later publication date when conflicts could create look-ahead; exclude ambiguous same-day entries.

Before treating a PDUFA date as pending, check the FDA decision record and corporate status: Nuvalent was approved July22,58days before the stale September18 target and after acquisition. Extend the pipeline-inventory rule to completed decisions and acquisitions; a calendar match alone is ineligible.

Do not let the113% headline return or the monthly funding gap override the relevant comparator: country buyhold gained101.80% with lower drawdown, and the executable range adaptation lacks robust edge. Keep actual sizing at zero until independent venue-compatible evidence changes that conclusion.

Fail closed on incomplete universes: the19-symbol CLI returned0bars because it omits pagination and truncates to the shortest array; it also aligns by index rather than date. Use the explicit paginated, date-aligned implementation and independent cash ledger for future country tests, and preserve full screen data in the journal.

A large selected winner does not justify replacing liquidated exposure to meet the monthly shortfall. Keep allocation zero when the current signal is absent, training is unavailable, or the venue exit rule remains unvalidated, even if a plug-in Kelly estimate exceeds the position cap.

A daily simulation must represent entry-day exits and reconcile every cash/share value; the corrected six-name model matched vectorbt within $8e-12. Repair invalid-symbol batches and preserve ticker NA as text before claiming listing-wide coverage; archive inputs, code, exclusions and timestamps.

Require adequate warm-up and actual training trades before calling a later window out of sample: PSIX had only four pre-2025 bars in both IEX and SIP. Its selected +159.63% result stays WATCH; the six-test mean-return interval of −0.63% to +8.35% does not establish an independently validated edge.

A null pagination token does not prove historical coverage: this run’s hourly responses stopped months before the requested end. Verify first/last timestamps and expected rows, continue from the last returned hour when necessary, and reconcile final portfolios with an independent cash ledger.

A sparse series needs calendar-window checks before ranking volume or calling a signal 20-day: USDG’s seven-row total was $4.103M, but the seven-calendar-day observed total was $2.103M across only four days. Flag missing days and suppress incomplete-window signals; this strengthens the earlier completed-bar alignment rule.

After time-index alignment, explicitly cast signals to boolean and verify portfolio results against a separate cash/quantity ledger; object dtype caused the first hourly-replication attempt to fail. Record hourly drawdown as well as daily marks—SOL’s rose from 12.68% to 15.01% with identical trades.

Distinguish a 20-hour strategy from a 20-day strategy evaluated hourly: the former lost 81–87% net in the diagnostic holdout, while SOL’s daily rule gained 22.2% on only 12 trades. Keep SOL WATCH until independent forward payoffs and actual venue exits validate it; do not rescue rejected hourly turnover by retuning inspected data.

A small fund cap does not make full deployment safe: a repeat of the observed 45.86% RARE loss on $49.14 would cost $22.54, above the $8.51 equity-risk budget. Keep size zero when the independent net holdout is negative; the monthly shortfall cannot turn CHWY’s −1.72% six-event result into an edge.

Keep issuer announcement dates and exact entry/exit rows in the journal as well as local files; a vanished artifact must not leave only a headline backtest return. Correct a misleading live monitor by replacing it, as KR #13 now distinguishes the confirmed 08:00 ET call from the unconfirmed release minute.