Ichimoku Cloud: How to Read Its Five Lines and the Cloud

Five lines and a cloud

The Ichimoku Cloud (Ichimoku Kinko Hyo) is an indicator created in Japan by Ichimoku Sanjin, whose real name was Goichi Hosoda, and it overlays five lines on the chart: the conversion line (Tenkan-sen), the base line (Kijun-sen), leading span A, leading span B and the lagging span (Chikou Span). The shaded area between the two leading spans is the cloud (Kumo). With so many lines it looks complicated, but the ingredients are simple. The four lines other than the lagging span are all combinations of 'the midpoint between the highest high and the lowest low over a set period', and the lagging span is just the close plotted further back. The reason for using a midpoint is that it is the center of the range over which buyers and sellers fought during that period. Price above the midpoint means it is trading in the upper half of that period's range; price below it means the lower half. So when you first look at Ichimoku, rather than memorizing line names, it is less confusing to treat it as a tool that asks whether price is above or below the midpoint of a short, a medium and a long period. A common mistake is to think that more lines mean more evidence. All five come from the same highs, lows and closes, so they are not independent pieces of evidence.

Conversion and base lines: the midpoint of a period

The conversion line is the highest high plus the lowest low of the last 9 bars, divided by 2; the base line is the same calculation over 26 bars. Where a moving average averages every close in its period, these lines look at only two values: the highest and the lowest. So they move only when a new high or low appears or when an old extreme drops out of the period, and in between they stay flat, as in the figure. When the conversion line is above the base line, the midpoint of the last 9 bars is higher than the midpoint of the last 26, which is read much like short-term moving averages stacked above long-term ones. The flat stretches are useful. A flat base line means the ends of the 26-bar range have not changed, which many people read as price moving back and forth inside that range. The thing to watch is why a line moved. A falling base line means one thing if a new low caused it and another if a high from 26 bars ago dropped out of the window. In the second case the line falls even if current price does not move at all, so it helps to get into the habit of finding the bar 26 bars back on the chart.

Leading spans: values shifted 25 bars ahead

Leading span A is (conversion line + base line) ÷ 2, and leading span B is the midpoint between the highest high and the lowest low of the last 52 bars. Both are plotted ahead of the bar they were calculated on. This is often described as '26 days ahead', but that counts the current bar as 1, so TradingView and this site's Ichimoku tool actually plot them 25 bars ahead. Compared with programs that shift them by 26 bars, the cloud looks one bar out of step, but the values of the lines are the same. Despite the name 'leading', they are calculated from bars that have already passed, so they do not look into the future. They are shifted forward to picture the idea that today's midpoint will act as a reference for a while. Reading the cloud drawn in the empty space on the right as a forecast of future prices is therefore a common mistake. The order of checks is this: first recall how many bars ago the cloud sitting under the current bar was calculated, then look at whether the cloud on the right is bullish or bearish, remembering that it only summarizes the last 25 bars. When comparing clouds from different programs, match the shift setting first.

The cloud's position, thickness and twists

Three things are read from the cloud. First, whether price is above, inside or below it; inside the cloud is often interpreted as a stretch where the direction has not been settled. Next is thickness. The farther apart the two leading spans are, the thicker the cloud, and one interpretation treats a thick cloud as support or resistance that price has trouble getting through. Last is which line is on top. A cloud with leading span A on top is called a bullish cloud, and one with leading span B on top a bearish cloud; the point where the two cross and the cloud changes is called a twist, and it is watched as a possible point where the trend may change. A thick cloud seems to hold because thickness means the short- and long-period midpoints are far apart, and many people traded in between, so orders near their break-even prices are thought likely to appear. Thin clouds, by contrast, are said to break easily. But thickness is measured in price, so a volatile instrument always looks like it has a thick cloud; it is better to judge it against the same instrument's past clouds. A twist reflects price changes that have already happened, given how the leading spans are calculated, so on its own it does not tell you the timing.

Lagging span: the close shifted 25 bars back

The lagging span involves no calculation. It is the current bar's close plotted 25 bars back, that is, at the 26th bar when the current bar is counted as 1. So the lagging span sitting above the price at that spot means the same thing as the current close being higher than the price 25 bars ago. This site's tool compares the current close with the close 25 bars ago. Because the line is pushed back, there is no lagging span over the last 25 bars of the chart, and the lagging span running into past price levels is sometimes interpreted as support or resistance. In the end, the lagging span is a way of making 25-bar momentum, the gap between the current price and the price 25 bars ago, visible. A common mistake is to treat it as an independent leading indicator; it is just the current close moved back, so all its information is already in the close. When judging it, also look at how far it sits from the bar at that spot. If they almost touch, current price is close to where it was 25 bars ago, and the verdict can flip easily over the next few bars. If the stretch 25 bars back swung sharply, the lagging span condition can change because of those past bars more than because of current price.

Sanyaku Kouten and Sanyaku Gyakuten

When all three conditions below point up, the state is called Sanyaku Kouten, and when all three point down, Sanyaku Gyakuten. This site's Ichimoku tool judges it on the last closed bar. In the figure, the conversion line and lagging span conditions are met first, and all three come together on the bar where price clears the cloud. All three conditions are met only after price has already risen, so Sanyaku Kouten is closer to a confirmation that the market has tilted to one side, and is hard to treat as a starting signal. The usual order of checks is the cloud position first, then whether the conversion line is above or below the base line, then the lagging span. The cloud reflects the longest period, so it sets the direction of the larger trend, and the other two conditions show whether the shorter trend agrees. It is rare for all three to be met for the first time on the same bar; they usually fall into place one at a time. Some people count two out of three as 'almost Sanyaku Kouten', but price often fails to clear the cloud and is pushed back, so by definition it is still mixed.

• The conversion line is above the base line
• The close is above the cloud (higher than the top of the cloud plotted at the current bar)
• The lagging span is above the price 25 bars ago (current close > close 25 bars ago)
• All three reversed is Sanyaku Gyakuten; anything in between is mixed

The numbers 9, 26 and 52

9, 26 and 52 are explained as roughly a week and a half, a month and two months, based on a six-day week from the days when markets also traded on Saturdays. Applied as-is to crypto, which trades 24 hours a day, seven days a week, 26 daily bars are 26 days, not a month. So some people favor a setting that stretches the periods to 20, 60 and 120 with a shift of 30 (an actual 29-bar shift in this site's tool), while others keep 9, 26 and 52 because the default is what the most people watch. There is no verified conclusion that either one is better, so the realistic approach is to switch between the two settings in the tool and compare whether the verdicts match. Longer periods make the lines slower, so the verdict changes less often but turns later. A common mistake when choosing numbers is to pick the periods that seem to fit the recent chart and then believe that setting is right. A setting fitted to the past few months is no guarantee for the next few. Changing the bar length has the same effect. Twenty-six 4-hour bars are a little over four days, so the same 9, 26 and 52 look at entirely different periods depending on the bar length.

The limit: it reacts late

Every Ichimoku line is either a midpoint of past highs and lows or a close shifted back, so the conditions change only well after price has turned. Leading span B, a 52-bar midpoint, is especially slow. In sideways markets, price moves in and out of the cloud and the conversion and base lines cross often, so the verdict keeps changing. With so many lines, it is also easy to find at least one plausible reason in any situation, which is something to be careful about, and the verdict comes out differently for each bar length. How price moved after Sanyaku Kouten was not measured in this article. Reacting late is less a defect than a result of the design. Using midpoints makes the lines move less on small swings, and the price paid is lagging behind sharp reversals. Right after a sudden drop, for example, every line may still point to the previous trend. In such moments, rather than assuming the trend is intact because the verdict has not changed, it is more practical to work out how far price would have to move for the verdict to change.

How it looks different in crypto and stocks

Crypto trades around the clock, so there are almost no gaps between bars, and a move through the cloud looks continuous from bar to bar. Stocks build up news overnight after the regular session, so the next open can start far from the previous close, and price often jumps over the cloud in a single bar. This is especially common in stocks such as Samsung Electronics and SK Hynix, which react to overseas semiconductor news, and in big US tech stocks that report earnings after the close. A cloud crossed by a gap means no trading happened inside it, so reading the cloud as support or resistance may fit less well. Korean stocks have a daily price limit that caps a single day's move, while US stocks have no daily limit but have mechanisms that briefly halt trading during sharp moves. The periods mean different things too. Stocks trade five days a week, so 26 daily bars are a little over a month, whereas in crypto they are 26 days. Keep in mind that the same setting covers different calendar spans in the two markets.

Reading it on a live chart

The last bar on a live chart has not closed yet. Each time its high, low or close changes, the conversion and base lines, the leading spans calculated on that bar and plotted 25 bars ahead, and the lagging span all move with it. If the forming bar rises above the cloud and then drops back just before the close, the Sanyaku Kouten you saw for a moment is never recorded. That is why this site's Ichimoku tool judges on the last closed bar. On a live chart, keep the shape of the forming bar separate from the verdict on closed bars, and note in advance the price above or below which this bar must close for the verdict to change; the screen can swing without shaking you. The shorter the bar length, the more often bars close and the more often the verdict changes.

A practical checklist

Working through Ichimoku in the order below makes it harder to miss something despite all the lines. Each step organizes the current state; none of them makes the decision for you.

• Check the bar length and settings first (whether it is 9, 26 and 52, and how many bars the shift is)
• On the last closed bar, see whether price is above, inside or below the cloud
• See whether the conversion line is above or below the base line, and whether the base line is flat or sloping
• Look at the distance between the lagging span and the price 25 bars ago to judge whether the verdict could flip easily
• Compare the cloud's thickness with the same instrument's past clouds
• Note the price at which this bar would have to close for the verdict to change