EMA vs Ichimoku
Michael Whitman
·
Trading since 2017, full-time since 2020
·
Published 2026-09-08
·
Updated
·
Reviewed
The exponential moving average is one smoothed line that weights recent closes most. Ichimoku is a five-part framework whose two fast lines, tenkan and kijun, are midpoints of 9- and 26-session high-low ranges, so its crossover is built from extremes rather than from closes.
A single smoothed line against a whole framework. The fairest test is to put the framework’s own crossover beside the moving-average crossover of the same lengths. Ichimoku’s tenkan and kijun use 9 and 26 sessions; a 9/26 exponential pair uses the same windows and only swaps what is measured.
What each one is
The exponential moving average is one line. Each new close moves it by a fixed fraction of the gap between that close and the old average: 2 ÷ (N + 1), so a fifth for a 9-session line. The exponential moving average covers it.
Ichimoku bundles five lines into one display. StockCharts credits it to the journalist Goichi Hosoda, in a 1969 book. Tenkan-sen is the midpoint of the 9-session high-low range, kijun-sen of the 26-session range; two more lines, the slower one a 52-session midpoint, form the cloud, plotted 26 sessions ahead; the fifth is the close plotted 26 sessions back. Ichimoku covers all five.
Their fast lines differ in one way that matters. An average of closes changes every session. A midpoint changes only when a new extreme arrives or an old one leaves the window.
Where they differ
What moves the line. Every close moves an exponential average. Tenkan can sit unchanged for days while price moves inside its range, then jump when the range’s high or low drops out.
How many questions are answered. The average gives one level. Ichimoku adds the cloud and the lagging line, a support-and-resistance read and a confirmation check in one display.
What is drawn forward. The cloud is shifted 26 sessions ahead; no moving average is. The shift moves past data forward on the chart; it does not forecast anything.
What has to be learned. One length against five components, which is the main reason people end up using only part of Ichimoku.
Where the two crosses agree
On SPY they crossed at almost the same rate, a little under ten times a year each, and most crosses of one had a partner cross of the other within a few sessions. Both lag, both whipsaw in sideways markets, and both cost a round trip per signal acted on.
Which one to use: EMA or Ichimoku?
Use the exponential pair when you want the crossover to respond to closes. It reacted first more often in this sample, and it has one number per line to understand.
Use Ichimoku when you will read all five parts. Used whole it is a coherent framework, and the tenkan-kijun cross is only one of its readings.
Use the tenkan-kijun cross when you want a signal that waits for a range extreme. It ignores closes that stay inside the old high-low range, which is a filter, and a slower one.
And when you only want a crossover, the exponential pair is the simpler tool. A quarter of a framework is not a simpler framework.
Why a midpoint line waits
A midpoint only moves when an extreme does. Tenkan is (highest high + lowest low) ÷ 2 over nine sessions. A rally that has not yet made a new nine-session high, and still has an old low in the window, leaves tenkan held down by that low.
So the cross often waits for a date, not for price. It comes when the old low ages out of the window, which can be several sessions after closes have already turned.
A worked example: September 2026
The exponential pair crossed first. On 21 September 2026 the 9-session EMA of SPY closes rose above the 26-session EMA, on a close of $773.50.
Tenkan stayed below kijun for five more sessions. Its nine-session window still held the 16 September low. On 28 September tenkan was $762.37 against a kijun of $762.45. On 29 September the 16 September low left the window, the lowest low in the nine sessions became $757.97, and tenkan jumped to $766.555, above kijun: the Ichimoku cross, on a close of $764.20.
Same move, two dates. The six-session gap here is wider than the usual one; in the full sample the median gap was two sessions.
The original data
The sample and the lines. SPY daily bars from Yahoo Finance, highs, lows and closes as traded, crosses counted from 3 January 1994 to 2 October 2026. Tenkan and kijun per StockCharts’ formulas, with the current session in each window. EMAs use 2 ÷ (N + 1), seeded in 1993. A cross is dated to the session the sign of the gap flips; a session where the two lines are exactly equal keeps the previous sign.
How often each crossed. 322 tenkan-kijun crosses, 9.83 a year, and 314 crosses of the 9/26 EMA pair, 9.59 a year. Tenkan and kijun were exactly equal on 306 sessions, which an average of closes almost never is.
Matching the crosses. Each tenkan-kijun cross was paired with the nearest EMA cross in the same direction within 20 sessions, nearest pairs first, each cross used once. That matched 246 pairs; 144 crosses, from one indicator or the other, had no partner.
Which came first. In the 246 pairs the Ichimoku cross came later 162 times, on the same day 36 times and earlier 48 times; the median gap was two sessions later. Against a coin toss, 48 earlier and 162 later is far outside chance (exact binomial, p < 0.0001). The median was two sessions for upward pairs and one for downward pairs.
What the later date cost in price. The median distance between SPY’s close on the two cross dates was 0.63% of price. That is a description of the gap, not a result of trading either cross. Every cross, its partner and both closes are in the SPY tenkan-kijun and EMA crosses file.
When it fails
The failure is treating the cloud alone as the filter and calling it Ichimoku. Above the cloud reads as bullish, so longs are taken. In a range the cloud flattens and price crosses it again and again, with the same whipsaws a single line produces, later.
The second failure is reading the forward-shifted cloud as a forecast. It is past data drawn ahead.
A third is expecting the midpoint lines to track closes. They wait for extremes, which is why the cross so often came second.
A fourth is stacking more indicators on top. Ichimoku already answers several of the questions they would add.
And a fifth is optimizing the periods. They were designed as a set, and tuning tenkan and kijun to make the cross fire earlier on past data mostly turns them into a slower copy of the exponential pair that already crossed first.
Related
The exponential moving average covers the one-line, close-weighted average. Ichimoku covers all five components and how they relate. And tenkan-sen covers the nine-session midpoint that drives the Ichimoku cross.
If you use the tenkan-kijun cross, know what moves it: a new high or low entering the window, or an old one leaving. An exponential average moves with every close. Decide which of those you want your signal to wait for before you compare the two on a chart.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.