Rejected thesis
Bullish RSI / MACD divergence (single-name timing trigger)
Code: research/bullish_divergence.py
Report: research/bullish_divergence_report.json (n=41,288)
Headline: a pivot low at lower price than the previous pivot low but with the oscillator at a HIGHER reading is a textbook bullish divergence. Both flavours (RSI and MACD histogram) produce material edge even unconditionally — and amplify cleanly when stratified by trailing 252d drawdown:
| Pattern × condition | n | 40d edge | hit @ 20d |
|---|---|---|---|
| rsi_div all | 9,869 | +6.78 % | 77 % |
| rsi_div dd ≤ -30% | 4,079 | +11.05 % | 78 % |
| rsi_div OOS 2020+ dd ≤ -30% | 2,467 | +12.86 % | 80 % |
| macd_div all | 31,419 | +5.33 % | 76 % |
| macd_div OOS 2020+ dd ≤ -30% | 5,445 | +11.06 % | 77 % |
This is one of the strongest standalone short-horizon signals in the repo — bigger raw edge than every other chart pattern we've tested (beats coil_breakout, PEAD, bear_flag, even unconditional gap patterns). The OOS amplification on drawdown is on top.
The reason oscillator divergences work where oscillator extremes don't: a divergence requires two pivots — i.e. price has actually established a swing-low structure — whereas an extreme reading can fire anywhere mid-slide.
| Sample | 41,288 |
|---|---|
| Validator | research/bullish_divergence.py |
| Verdict | Rejected |
Every result here is reproducible from the script named above. Reports are in the repository.
- Volatility squeeze
- Pre-FOMC drift
- Uptrend pullback
- Momentum continuation
- Industry-relative momentum (at 4-week lookback)
- Earnings-reaction reversal (anti-PEAD)
- Beat-persistence pre-earnings drift
- Sector ETF mean-reversion
- 52-week-high breakout
- Volume accumulation (above-200DMA +...
- Cross-sector momentum
- Analyst-upgrade cluster