Ichimoku Kumo Breakdown Short: Sell Beneath a Bearish Cloud

Ichimoku Kinko Hyo answers three questions in one picture, and this strategy will not sell unless all three agree. Where is price? Beneath the Kumo — the cloud is the market's accepted value area, and trading under it is what a bearish regime looks like. What is the cloud itself saying? Senkou Span A below Senkou Span B, so the cloud projected ahead is genuinely bearish rather than being somewhere price happens to sit under. And what is momentum doing right now? Tenkan-sen below Kijun-sen, the fast conversion line having given way to the slower base line. The short is covered when price climbs back into the cloud, on the view that a reclaimed value area ends the breakdown thesis. Every distance test is measured in ATR rather than in price, so "well below the cloud" carries the same meaning on BTC as it does on XRP. The 9 / 26 / 52 / 26 Ichimoku periods are held at their canonical values on purpose, and the tunable parameters are the entry and exit thresholds instead. Risk, exactly as the pinned backtest executed it: a hard -10% stoploss, which a short position needs because its loss is otherwise unbounded, and a flat +10% ROI target that closes any single trade reaching it. This is a short-only design built for a falling market. Over the pinned in-sample window buy-and-hold was about -62%, and a long-only strategy could not post a positive absolute return there at all. What is published here is the UNTUNED configuration, and that is a deliberate choice worth explaining. A 100-epoch hyperopt over the in-sample leg did find a stronger setting: it lifted in-sample return from +0.11% to +1.43% and profit factor from 1.02 to 1.41. But on the held-out out-of-sample leg that same setting was worse on every measure — return +0.22% to +0.16%, profit factor 1.41 to 1.35, Sharpe 2.13 to 1.27 — and the platform's own overfit score rose from 0 to 39. In-sample is the number a listing displays; out-of-sample is the number that tells you whether it will keep working. The tuned version looked far better on the page and slightly worse in the only test that had not been optimised against, so the untuned one is what ships.