Ichimoku Bearish

Overview

Ichimoku Kinko Hyo (“equilibrium chart at a glance”) is a multi-component indicator that paints a forward-projected “cloud” (Kumo) from price-range midpoints. This preset enters SHORT when price is trading below the cloud AND the Tenkan-sen (9) crosses below the Kijun-sen (26), confirming downside momentum against a bearish structural backdrop.

Original 2026-05-04 sweep (1,680 combinations × 50 coins, IS/OOS split) selected SL 15% / TP 15% and reported WR 51.7% and PF 1.55. A fresh out-of-sample re-test on 2026-06-28 (product engine, regime-robustness + 5-axis adversarial kill) found the SHORT result to be bear-beta — profit driven by a falling market, not a regime-independent cloud edge. The backtest is reproducible; we no longer present this as a confirmed live directional edge.

How It Works

1. Cloud filter — price must be below both Senkou Span A and B (the cloud bottom)
2. Tenkan/Kijun cross — short-term (9-period) midpoint crosses below medium-term (26-period) midpoint
3. Confirmation — Chikou Span (lagging close, displaced −26) clear of cloud
4. Entry — SHORT on bar close after the cross
5. Exit — TP 15% / SL 15% / 12-bar (2-day on 4H) timeout

Why It Works (Thesis)

Crypto downtrends are often multi-week. The cloud catches regime; the Tenkan/Kijun cross catches timing. Combining them filters out bear-trap countertrends that break simpler moving-average systems.

With a symmetric 15%/15% structure, the original result was WR-based: a 51.7% win rate. At the time the IS→OOS increase in profitable coin count (38→40 of 50) looked like evidence the edge was not regime-specific. The 2026-06-28 re-test reverses that read: both the IS and OOS windows were falling markets, so a SHORT book wins on direction. Decomposed against regime, the cloud-plus-cross filter does not show a standalone edge — treat this thesis as unconfirmed.

Results (2-year backtest, IS/OOS split, measured 2026-05-04)

Metric	IS (May24–May25)	OOS (May25–May26)	Combined
Total trades	725	900	1,625
Win rate	~52%	~52%	51.7%
Profit factor	2.07	1.83	1.55
Coins profitable	38/50	40/50	—

IS PF (2.07) is higher than OOS (1.83). These original numbers are reproducible, but the 2026-06-28 re-test attributes the profit in both periods to bear-beta rather than a regime-independent edge — so they are not evidence of a confirmed directional edge.

Caveats

• Failed fresh out-of-sample regime-robustness (2026-06-28). The earlier “not regime-specific” claim was reversed — the SHORT result is bear-beta. Status downgraded from verified to testing for this reason.
• Symmetric 15%/15% means a losing trade is as large as a winning one — capital management is critical.
• Not live-tracked on OKX. Backtest only — “verified” would mean the backtest is reproducible, not that the live directional edge is confirmed.
• 4H signals fire less frequently than 1H variants. Expect 1–3 signals per coin per week.