AI · US Semiconductors — Track record
Every elapsed decision is graded against what prices actually did — hit-rates and errors per signal family, plus the cost of the decisions this book chose not to take.
121 graded 0 pending
1W calibration 0.09 How far along the 1-week path a decision typically gets: the fraction of the decided move realized after one week, measured across this book's graded decisions. Feeds a bounded conviction adjustment back into future decisions.
Scorecard
Hit-rates per signal family.
| Signal family | Dir | n | Hit 1W | Hit 4W | Hit 12W | Med err 1W | Stop-first | Bias | Status |
|---|---|---|---|---|---|---|---|---|---|
| NoTrade | sell | 66 | 50% | 27% | — | — | 0% | 0 | ● active |
| FailedBreakdown_Up_20D | sell | 18 | 33% | 17% | — | — | 0% | -3 | ● active |
| RallySell_BearTrend | sell | 8 | 25% | 0% | — | — | 0% | -5 | ● active |
| BreakdownDown_20D | sell | 7 | 60% | — | — | — | 0% | +2 | ● active |
| FailedBreakdown_Up_20D | buy | 7 | 14% | 14% | — | -3.1% | 86% | -7 | ● active |
| FailedBreakout_Down_20D | sell | 5 | 20% | 25% | — | — | 0% | -6 | ● active |
| AnticipationUp_VCP | buy | 2 | 50% | 0% | — | +5.5% | 50% | 0 | ○ building sample (2/5) |
| AnticipationUp_VCP | sell | 2 | 50% | 100% | — | — | 0% | 0 | ○ building sample (2/5) |
| BreakoutUp_52W | buy | 2 | 50% | 0% | — | +6.4% | 50% | 0 | ○ building sample (2/5) |
| ReversalUp_Oversold | sell | 2 | 0% | 0% | — | — | 0% | 0 | ○ building sample (2/5) |
| MomentumContinuation_Up | buy | 1 | 0% | 0% | — | -12.6% | 100% | 0 | ○ building sample (1/5) |
| ReversalDown_Overbought | sell | 1 | 0% | 100% | — | — | 0% | 0 | ○ building sample (1/5) |
Counterfactuals Counterfactuals grade decisions that were NOT executed — expired queues, no-chase rules, untaken optional trims — against what the market then did. They are kept strictly separate from the calibration cells above.
What the decisions we didn't take would have done.
- Optional trims not taken — 22 cases , median would-have 1W move +0.5% → verdict: saved
- Queue expiry (order lapsed unfilled) — 3 cases , median would-have 1W move +3.7% → verdict: cost
- Anticipate slot (pre-breakout buy-stop) — 12 cases , median would-have 1W move +1.3% → verdict: cost
- No-chase queue (queued a pullback instead of chasing) — 15 cases , median would-have 1W move +7.3% → verdict: cost
Counterfactuals are advisory evidence for the book's learnings reviews — never blended into the calibration cells.