| RSI divergenceH1 | 19 instruments · H1 / H4 / D1~235,000 trend bars |
55.5% reversal vs 49.9% base · z +4.3Holds on 17 of 19 symbols |
| Correction depthStop guide | Same sample |
Median 1.3 ATR · 80% under 2.8 ATRBecame the 2.8 ATR stop guide table |
| CTA horizons in agreementAll four | Same sample |
+0.20 ATR over 20 barsTrend persistence — not a reversal |
| COT positioning3-year speculator index | 15 instruments · weekly COT2020 – 2026 |
Below 25 and above 90 shift 13-week odds by 10 – 14 pp (t 2.1 – 2.9)Context row — not an entry signal |
| COT extremes52-week index | 15 instruments2018 – 2026 · 251 episodes |
52 – 53% against a 50% baseNo usable reversal signal — reported as such |
Wyckoff target · CORR line · setup range lines Regression & pivot channels | Same sample |
No better than mirrored / offset controlsKept as drawing context only, never as a signal |
| Chart patternsDouble / triple top & bottom · H&S | 13 instruments · H1 / H4 / D13,207 confirmed patterns |
49.9% vs 50.6% for a random entryso PATTERNS is drawn as structure context |
| Measured rangeVOLATILITY 3D · RANGE surface | 13 instruments · H1 / H4 / D1out of sample · 1,500-bar window |
79.7 – 81.3% of closes inside the 80% zone · 50.1 – 50.5% inside the 50% zoneCalibrated — the zones mean what they say |
| Volatility forecast modelsRW20 · EWMA · GARCH · HAR-RV | Per symbol, livelast 250 forecasts · each from past data only |
Ranked by QLIKE in the window — e.g. gold H4: EWMA −7.5%, GARCH +38.8% at 20 daysThe best model differs by symbol — the window shows which one |
| News — size of the moveNEWS IMPACT | 12 symbols · M152022 – 2026 · every release with ≥ 20 cases |
First 15 min after a Fed decision ×5 – 6.5 a normal 15 min, CPI and payrolls ×3 – 4Shown per release, in pips / points for your symbol |
| News — MT5 good / bad flagDirection | ~1,700 USD releases per symbol2021 – 2026 |
48 – 52% in the "expected" directionNot used for direction |
| News — direction after a surpriseActual vs forecast | 12 symbols · M15~340 big surprises each |
Big surprise: 55 – 60% textbook direction (GBP/USD 59.5%) · small: ~50%Shown as a scenario where significant — never as a signal |
| Cone pathKalman push → pullback → target | 13 instruments · H1 / H4 / D189,203 historical paths |
Direction 50.1% vs 50.3% base · channel holds the close 1 – 13%Drawn and labelled as an illustration |
| Probability map±1σ · 5–95% · target close / touch | 7 currency pairs · D11971 – 2026 · walk-forward, past data only |
Stated vs real frequency: ±1σ 1.3 pts · percentiles 1.6 · close beyond 0.8 · touch 0.9Size of the move, not direction |
| Volume profile levelsPOC · VAH · VAL · naked POC | 8 instruments · ~44 daysprofile from one-minute bars · 0.978 vs tick profile |
Bounce 50 / 50 / 53% vs 49% at random prices · "80% rule" 41% vs 41%Shown as context — never as a signal |