Current Model Paper Track Record

Current model performance by strategy

Range fades and trend pullbacks use different target sizes and therefore different break-even win rates. Their forward results are reported separately instead of being blended into one marketing number.

Legacy model archive

Earlier-model outcomes remain preserved for auditability and are excluded from the current-model headline statistics above.

How outcomes are counted

WIN means the fixed final target was reached before the fixed stop. LOSS means the stop was reached first. TIMEOUT means neither was reached within the signal horizon. If both target and stop fall inside the same OHLC candle and order cannot be proven, the result is AMBIGUOUS and is excluded from win rate. Opposite-direction replacement signals can mark an unresolved signal CANCELLED. This avoids quietly treating uncertain results as wins.

Win rate = Wins ÷ (Wins + Losses). It is a historical forward record, not a forecast or guarantee.

All-time results by signal tier (legacy + current)

Recent current-model signal outcomes

Only signals created by the current model version are shown in this table. Earlier-model records remain preserved in the legacy archive summary above.

Learn how to audit the numbers

Read the detailed guide to forward tracking, decisive win rate, R-multiple, timeouts, ambiguous candles and sample-size limits.

Why this is separate from backtesting

The model also uses historical cohorts to decide whether a setup qualifies. Those validation statistics are research evidence, while this page is the forward-published track record. Keeping the two separate makes it harder to cherry-pick historical periods after the fact.