Ichimoku Strategies

Ichimoku Kinko Hyo — "one-glance equilibrium chart" — is a complete trend framework rather than a single indicator. Its five lines come from the same raw material: midpoints of past high–low ranges, not averages of closing prices. The **Tenkan-sen** is the midpoint of the last 9 bars' high and low, the **Kijun-sen** the midpoint of the last 26, and **Senkou Span B** the midpoint of the last 52. **Senkou Span A** averages Tenkan and Kijun. Both spans are plotted 26 bars ahead; the band between them is the **Kumo**, or cloud. The **Chikou Span** is the current close plotted 26 bars back.

Two structural properties explain most of its behaviour. Because the lines are range midpoints rather than moving averages, they flatten whenever the extremes of their window stop changing — a flat Kijun or Span B marks a level the market has balanced around, an equilibrium price rather than a trend line. And because the cloud is displaced forward, support and resistance for the next 26 bars are drawn before price arrives: the shape is known, the path through it is not.

## How the components produce signals

Most implementations draw on four readings. Price position relative to the cloud sets the broad bias — above is bullish territory, below bearish, inside undefined. The Tenkan/Kijun cross (the **TK cross**) is the trigger, conventionally weighted by whether it occurs above, inside or below the cloud. The Chikou Span confirms: the lagged close standing clear of the price action 26 bars ago means the move is not running back into recent structure. The cloud supplies context — its thickness stands in for how much past two-way trade sits at those levels, and the **Kumo twist**, Span A crossing Span B, marks a projected change of structure at a date already visible on the chart. The Kijun also serves as a trailing stop or pullback level.

## Main variants

- **Cloud breakout** — entry on a close beyond the Kumo, with the far edge or the Kijun as stop. - **TK cross** — the Tenkan/Kijun cross as trigger, filtered by cloud position. - **Kumo twist** — trading the projected Span A/Span B crossover, or reading it as an early bias change. - **Layered confluence** — three to five conditions required simultaneously; the "four-layer" formulations belong here. - **Ichimoku as a filter only** — the cloud gives direction while the trigger comes from elsewhere: momentum or flow oscillators, trend-strength measures, an ATR trailing system, a packaged signal overlay. - **Retuned or reduced** — non-standard lookbacks for 24/7 markets, or single-component use such as Kijun-only levels. - **Scalping adaptations** — the full structure on 1–15 minute charts, usually with a volatility filter and a fixed risk-multiple exit.

## What typically differentiates implementations

First, which line is load-bearing: a version gated on the cloud and one triggered by a TK cross are different systems sharing a chart. Second, the **offset convention** — platforms disagree on whether the displacement is 26 or 25 bars, and a version must say whether it reads the cloud at the current bar or at the leading edge. Third, execution timing: intrabar cross, confirmed close, or next-bar open. Fourth, the exit — opposite cross, Kijun trail, cloud edge, fixed multiple — which usually moves results more than the entry. Finally the settings, and whether the version can stand flat.

## Common mistakes

The characteristic error is Chikou-based lookahead. The test itself is legitimate in real time — today's close against price 26 bars ago — but a version that requires the lagged line to be clear of price and then books the entry at the bar where that line is drawn is entering 26 bars in the past. Any Chikou rule needs its evaluation bar written down explicitly.

Counting five Ichimoku conditions as five independent confirmations is the next: all five lines derive from the same high–low midpoints over overlapping windows, so they agree and fail together. Retuning the three lookbacks for a 24-hour market without moving the displacement projects the cloud over a horizon that no longer matches the Kijun. Beyond those: trading every TK cross inside a thick cloud, where the system is explicitly undefined; running it on Heikin Ashi or Renko series, whose prices were never tradable; and testing low-timeframe versions without costs.

## How to evaluate and backtest a version

Reproduce the five lines and every offset before trusting a result, then write down the evaluation bar and the execution bar for each condition — with a displaced series and a lagged one in play, this is where most unreproducible results originate.

Test the conditions in sequence rather than as a block: cloud filter alone, then the trigger, then each confirmation, so every contribution is attributable. Watch the trade count as you stack them — full confluence is rare, and thirty trades over a decade is a story, not a sample. Examine the parameter surface across the three lookbacks, treating the displacement as a separate axis. Segment by year and regime, read long and short separately, and apply realistic costs. Then benchmark: Tenkan and Kijun are Donchian channel midlines, so a plain Donchian or moving-average system on the same data is the honest baseline to beat.

The 20 decoded versions linked from this page vary along exactly these axes: which component triggers, which merely filters, how confluence is defined, and what closes the trade.

Strategies in this concept (27)

Frequently asked questions

What are the standard Ichimoku settings, and should they be changed for crypto or lower timeframes?

The original configuration is 9, 26, 52 with a 26-bar displacement, chosen when the Japanese trading week ran six days, making 26 roughly one month of sessions. That calendar logic does not carry over to a 24/7 market or a 5-minute chart, so retuned versions are legitimate. What matters is that the four numbers are coupled: the displacement is normally set equal to the Kijun period, and Senkou Span B to about twice it. Changing one lookback in isolation breaks the proportions the readings assume. Prefer a set that sits inside a broad region of similar behaviour over an isolated optimum.

Does the Ichimoku cloud repaint or look into the future?

The cloud does not repaint. Senkou Span A and B are computed from closed bars and then plotted 26 bars forward, so the shape ahead of price is already fixed and known — that is the design, not a leak. The lookahead risk sits with the Chikou Span. Comparing the current close with price 26 bars ago is available in real time, but a rule phrased as "the Chikou is clear of price" and then timestamped at the bar where the Chikou is drawn books a decision 26 bars before the information existed. Check which bar a version evaluates and which bar it executes on before comparing its results with anything else.

What is a Kumo twist, and does cloud thickness matter?

A twist is the point where Senkou Span A crosses Senkou Span B, flipping the cloud's colour. Because both spans are displaced forward, the twist is visible up to 26 bars before price reaches it, which is why some versions use it as an early bias change and others as a projected timing window. Thickness is the gap between the two spans, driven by the distance between the 52-bar midpoint and the shorter-term ones. A thin cloud is easier for price to pass through; a thick one represents a wide zone of past two-way trade. Versions that ignore thickness tend to take their worst signals inside the widest clouds.

Do all the Ichimoku signals need to align before taking a trade?

That is one design choice, not a requirement, and it has a cost. Requiring price above the cloud, a TK cross, a clear Chikou and a bullish twist simultaneously produces few signals, and few signals mean statistics you cannot lean on. It also overstates independence: the five lines are built from overlapping high–low midpoints, so they are correlated by construction and add less confirmation than their number suggests. The practical approach is to measure each condition's marginal contribution rather than assume more conditions is better.

Is Ichimoku a trend system, or can it be used in ranges?

It is a trend framework, and it says so itself: price inside the cloud is the explicit undefined state. In a range, price oscillates through the Kumo and the Tenkan and Kijun cross repeatedly without follow-through, so a version that trades every cross reverses into every swing. This is why many catalogued versions either require price to be clear of the cloud, add a separate trend-strength or volatility condition, or use Ichimoku only as a directional filter for a trigger generated elsewhere. When evaluating a version, segment the results by regime — if the gains come from a few trending stretches, the range behaviour is the part that needs work.

Why do decoded versions built on the same indicator differ so much?

Because the indicator is only one component. Two versions can both be Ichimoku strategies while disagreeing about which line triggers and which only filters, about execution timing, about settings and displacement, about the exit, and about whether Ichimoku is the core of the system or a single gate in a larger stack carrying other momentum, volume or volatility tools. When comparing the versions on this page, read those elements first — they explain the differences far better than the parameters do.

All strategy concepts · Explore all strategies