Ichimoku Cloud

One chart, five lines, a cloud into the future

Ichimoku Kinko Hyo — literally "one-look equilibrium chart" — is unusual in almost every dimension. It's the only classical indicator that plots forward in time. It's the only one that bundles five distinct components onto a single chart. And it's the only major indicator in this course whose coverage in our reference library (Murphy, Kaufman, Bulkowski) is mostly absent.

Kaufman has a short section with the formulas and a basic rule set. Murphy doesn't mention Ichimoku at all in our edition. Bulkowski likewise. Most of what's taught today as "Ichimoku" — the Kumo twist, Chikou confirmation, the signal-strength hierarchy — comes from Japanese-language primary sources and retail-trading traditions outside the classical-Western TA canon. We'll cover the math where our books support it, and flag where we're working from tradition rather than direct citation.

Origin

Kaufman's terse attribution:

Developed by Goichi Hosada, published in his book in 1969 but developed in the 1930s. — Perry Kaufman, Trading Systems and Methods

That's essentially all Kaufman says about Hosoda. The rest — his pen name Ichimoku Sanjin, his team of assistants, the 30+ years of refinement — is historical tradition outside our library.

The name parses as:

• Ichimoku — "one look" / "at a glance"
• Kinko — "equilibrium" / "balance"
• Hyo — "chart" / "graph"

A chart that shows balance at a glance. Hosoda's goal was a single image from which a trader could immediately see trend direction, momentum, support/resistance, and future bias without needing separate indicators. Whether it actually delivers on that promise is a different question — but that was the design intent.

The five components

Kaufman's reproduction (kaufman.txt:13904):

1. Tenkan-sen (Conversion Line)

Midpoint of the last 9 bars' high-low range. Fast-moving. Think of it as a very short-term price anchor.

2. Kijun-sen (Base Line)

Midpoint of the last 26 bars' range. The slower anchor. Price in a healthy uptrend tends to pull back to Kijun and bounce.

3. Senkou Span A (Leading Span A)

The midpoint of the two running anchor lines, then displaced 26 bars forward.

4. Senkou Span B (Leading Span B)

Midpoint of the last 52 bars' range, also displaced 26 bars forward. Slower than Senkou A — responds to a longer window.

5. Chikou Span (Lagging Span)

Yes, behind. Chikou is today's close drawn at the timestamp of the bar that occurred 26 bars ago. This sounds useless but lets you visually compare "where is price now vs where price was 26 bars ago" — a crude momentum check.

Kaufman's phrasing note: Kaufman writes Tenkan and Kijun as "MA of highs + MA of lows", which is slightly different from the canonical midpoint of the N-bar range. Standard modern implementations use (max(H) + min(L)) / 2 over the lookback, and that's what our ichimoku() function computes. If you see numeric discrepancies between our playground and Kaufman's chart, this is why.

The Cloud (Kumo)

The area between Senkou A and Senkou B, projected 26 bars into the future.

When Senkou A > Senkou B → the cloud is bullish (usually rendered green). When Senkou A < Senkou B → the cloud is bearish (usually rendered red).

Kaufman's clean framing:

By projecting the cloud forward, traders can easily see support and resistance areas as prices approach them.

This is Ichimoku's signature feature. Every other indicator you've learned plots on the current bar or earlier. Ichimoku's cloud tells you where future S/R zones will be before price gets there. The cloud 5 bars ahead of the current bar is already computed based on today's data.

Note on rendering: lightweight-charts doesn't natively fill an area between two lines. Our playground below draws Senkou A (green line) and Senkou B (orange line) instead; the "cloud" is the region between them. Professional platforms (TradingView, MotiveWave) fill it.

Default periods — and a caveat

The canonical 9 / 26 / 52 / 26 periods were Hosoda's choices. Our books do not explain why. The widely-cited folk reasoning — that 26 = one month of Japanese trading when markets traded 6 days a week — is common on trading forums but not sourced here. Treat that explanation as tradition, not citation.

Some modern practitioners adapt the periods for 5-day weeks (e.g., 7 / 22 / 44 / 22), arguing that Hosoda's choices were artifacts of his era. Purists hold to 9 / 26 / 52 / 26 regardless. There's no rigorous testing in our library either way — if you want data on whether adapted periods perform better, you won't find it in Murphy, Kaufman, or Bulkowski.

Play with it

Toggle All 5, Tenkan/Kijun, Cloud only, or Chikou only to isolate each layer. Switch market to sideways and watch the cloud collapse to near-flat; switch to up or down and watch it open wide.

Notice:

• Tenkan vs Kijun behaves like a short-term / long-term MA crossover pair (9 vs 26 bars).
• Cloud projects 26 bars beyond the last price bar — look at the right edge of the chart; the cloud continues past where price ends.
• Chikou trails to the left — it's today's close drawn 26 bars ago.
• All 5 is dense. Hosoda's "one-look" works only after you've trained your eye to parse the stack.

Signals — Kaufman's rules

Kaufman gives a two-tier rule set (kaufman.txt:13931):

Macrotrend filter (the cloud)

The 'cloud' serves the purpose of a long-term trend filter. Positions are recommended only in the direction of the cloud, long when green, short when red.

Buy when the cloud turns from red to green or when the price moves above the cloud when the cloud is green (bullish). Sell short when the cloud turns from green to red or when the price moves below the cloud when the cloud is red (bearish).

Shorter-term (TK cross, filtered by cloud)

Trades are taken only in the direction of the cloud. Buy when the Conversion Line (Tenkan) moves above the Base Line (Kijun). Sell short when the Conversion Line moves below the Base Line.

In other words: the cloud determines regime, the Tenkan/Kijun cross fires the entry. Same structure as ADX + MA crossover — Ichimoku's cloud plays the role ADX usually does.

Signals NOT in our library

The following are widely taught in Ichimoku practice but not attested in Kaufman, Murphy, or Bulkowski. Presented as tradition rather than citation:

Chikou confirmation

Compare the Chikou Span to the price bar 26 periods ago. If Chikou > that old price, it's bullish confirmation (today's close exceeds price 26 bars prior). If Chikou < that old price, bearish. This three-way check (price vs cloud / TK cross / Chikou vs historical price) is the classical signal-strength hierarchy.

Kumo twist

When Senkou A crosses Senkou B (looking forward 26 bars), the cloud changes color. Some practitioners treat a future-dated twist as a leading signal — the cloud's color will flip before price gets there, giving advance warning.

Kaufman's rule "buy when the cloud turns from red to green" is effectively a Kumo twist, just phrased without the Japanese term.

Signal strength hierarchy

Classical Ichimoku practice grades setups by how many components agree:

All four aligned = strong. Three of four = moderate. Two = weak. This hierarchy is tradition, not in our library.

What Kaufman does warn about

Kaufman flags that Ichimoku has essentially no rigorous English-language statistical testing:

There have been very few comprehensive studies published since 1990.

Immediately after his Ichimoku section he cites Colby and Bulkowski as the rare exceptions in pattern testing — but neither has any Ichimoku content. That's an honest statistical vacuum. Whether Ichimoku "works" in the Bulkowski-backtested sense is an open question in our reference library. It's widely used; its edge is unquantified.

Why the design might still be useful

Even without statistical validation, Ichimoku has two legitimate properties:

1. It combines regime filter + momentum trigger + forward projection in a single visual. Most traders approximate this by stacking a regime indicator (ADX) + an oscillator (RSI/MACD) + support/resistance lines. Ichimoku bakes all three into one chart.
2. The cloud's forward projection is a real feature, not cosmetic. Knowing where future S/R will be before price arrives gives a planning edge — you can identify upcoming decision points 26 bars in advance.

Neither property proves it works. Both make it a legitimate tool in the toolbox, especially for traders who prefer a holistic chart read over stacked indicator panels.

Hidden traps

• Thinking "Ichimoku" is one indicator. It's five, plus the derived cloud. Each component works differently and carries different signals.
• Reading forward cloud as price prediction. The cloud 26 bars ahead is computed from today's data — it's where today's Tenkan/Kijun/ranges project. Price doesn't have to go there.
• Ignoring Kaufman's phrasing note. If you're implementing Ichimoku yourself and using "MA of highs + MA of lows" instead of "N-bar high + N-bar low midpoint," your numbers won't match TradingView's. Canonical implementation uses range-midpoint.
• Trading Tenkan/Kijun crosses without cloud filter. Even Kaufman's simplified ruleset filters TK crosses through cloud direction. Raw TK crosses without regime context whipsaw badly.
• Using period adaptations without backtesting. The 9/26/52/26 defaults are Hosoda's. "Modern" 7/22/44 variants lack statistical validation in our library. Changing periods without testing is curve-fitting on tradition.
• Expecting Ichimoku to work on thin instruments. Like every range-based indicator, Ichimoku assumes honest high-low behavior. On illiquid stocks where a single block trade defines the daily range, the components are noisy.
• Treating cloud thickness as a precise S/R strength measure. Tradition says thicker cloud = stronger S/R, and it's plausible (thicker means Senkou A and B disagree more, implying a wider balance zone) — but again, not tested in our library.

Quick check

What you now know

• Ichimoku Kinko Hyo — "one-look equilibrium chart" — was developed by Goichi Hosoda in the 1930s and published in 1969. Five components stacked on one chart.
• Five lines: Tenkan (9-bar mid), Kijun (26-bar mid), Senkou A and B (projected 26 bars AHEAD), Chikou (today's close, plotted 26 bars BEHIND).
• The Cloud (Kumo) is the area between Senkou A and B, projected forward. Unique among indicators — shows future S/R zones before price arrives.
• Two-tier signals (Kaufman): (1) cloud direction = regime filter, (2) Tenkan/Kijun cross = entry trigger. Only take trades where both agree.
• Chikou confirmation and Kumo twist are additional signals from tradition but not attested in our library. Use carefully.
• No rigorous backtesting in Murphy / Kaufman / Bulkowski. Kaufman himself flags that few comprehensive indicator studies exist post-1990. Ichimoku's edge is widely claimed but not quantified here.
• 9 / 26 / 52 / 26 defaults are Hosoda's — the folk "6-day-week" rationale is tradition, not sourced in our books. Modern 7/22/44 adaptations exist but lack testing.
• Kaufman's "MA of highs + MA of lows" phrasing differs slightly from the canonical "N-bar range midpoint" formulation — our implementation uses the canonical form.

Next: Williams %R — Larry Williams's simple inverted Stochastic, or a pair of remaining oscillators (CCI, ROC). Short, rich, closing out the More Indicators unit.