Weighted Moving Average vs Ichimoku

Michael Whitman · Trading since 2017, full-time since 2020 · YouTube creator, 137K subscribers · Published 2026-09-08 · Updated 2026-10-19 · Reviewed 2026-10-05

One of these is a single line with one setting. The other draws five things and projects two of them into the future. The difference that matters most is what each one listens to: the weighted average hears every close, and Ichimoku’s lines hear only the extremes of a range.

What each one is

The weighted moving average scales each close by its age. In a nine-session version the newest close is multiplied by 9, the one before by 8, down to 1 for the oldest, and the total is divided by 45, the sum of 1 to 9, so the newest close carries 20.00% of the weight and the oldest 2.22%. The weighted moving average covers it.

Ichimoku draws two fast lines from 9- and 26-session range midpoints, a cloud whose slower edge is the 52-session midpoint, projected 26 sessions forward, and a lagging line behind price. Its fastest line, tenkan-sen, is the middle of the nine-session range, the highest high plus the lowest low, halved, per StockCharts. Ichimoku covers each component.

The inputs differ before the arithmetic does. Every moving average is a weighting of closes. Ichimoku’s lines are statements about range, not about where each bar finished.

Where they differ

Whether the tool can say nothing. The weighted average is always sloping one way or the other. Ichimoku can put price inside the cloud, a state that means no trend.

What makes each line move. A new close moves the weighted average every session. Tenkan moves only when a new nine-session high or low prints, or when the old one drops out of the window.

Level against zone. The line is one price. The cloud is a band whose width comes from the recent range.

Backward against forward. Ichimoku shifts its cloud 26 sessions ahead of price; nothing in the moving average family does.

What the weighted line and the cloud share

Both are computed from bars that have already closed, the cloud’s forward shift included. Both lose their way in a sideways market: the line crosses price back and forth, and price drifts inside the cloud.

Which one to use: Weighted Moving Average or Ichimoku?

Run Ichimoku when you want a reading of no trend at all. Only the cloud can report that state, and no choice of weights gives a moving average that reading.

Do not buy the cloud as a whipsaw detector for the line without checking. On bitcoin’s daily chart the weighted line was crossed about as often on cloud days as on any other day, as the count below shows.

Run the weighted average when you want each close to count. It never ignores a session; Ichimoku’s lines ignore every close that stays inside the old range.

And when screen space is short, the line wins. On a chart already carrying structure work, one more line is affordable and a cloud is not.

Bitcoin’s no-trend days and the weighted line

The question. One way to pair the two uses the cloud as a filter: stand aside from the line while price is inside the cloud, because that is when the line chops back and forth. That claim can be counted on a market that trades every day of the week.

How often the cloud said no trend. From 4 December 2014, the first day with a full cloud, to 4 October 2026, bitcoin closed inside its cloud on 668 of 4,323 days (15.45%), above it on 2,114 (48.90%) and below it on 1,541 (35.65%). The share was steady by year, between about 10% and 19% in every full year from 2015 to 2025, and 25.6% so far in 2026.

How often the line was crossed. A cross means the close switched sides of the nine-day weighted average from one day to the next. Inside the cloud that happened 133 times, 19.91 per 100 days. Above the cloud it was 20.29 per 100 days and below it 23.95, so 21.83 per 100 outside the cloud overall. The gap between inside and outside is within chance (Fisher exact test, p = 0.283). The line’s own turns, a day when its slope changed sign, came at 15.87 per 100 days inside and 16.50 outside (p = 0.733).

How long the no-trend readings lasted. The 668 inside days came in 163 separate spells. The median spell lasted 2 days, 56 lasted a single day and 17 lasted ten days or more. The longest ran 27 days, from 20 July to 15 August 2026, and the weighted line was crossed 7 times inside it.

A worked example: SPY, 14 and 17 August 2026

The nine closes. From 4 to 14 August 2026 SPY closed at $771.33, $769.79, $768.56, $773.26, $773.03, $770.56, $772.49, $777.88 and $776.34.

The weighted average on 14 August. Multiply each close by its weight, 1 for 4 August up to 9 for 14 August, add the nine products and divide by 45: $773.68.

Tenkan on 14 August. The highest high in the same nine sessions was $779.37, on 13 August; the lowest low was $760.52, on 4 August, when SPY’s low sat 1.40% under its close. Tenkan = ($779.37 + $760.52) ÷ 2 = $769.945, about $3.74 below the weighted line, because one deep intraday low set half its value.

Then the window moved. On 17 August SPY closed at $772.67 and the 4 August session left the window. Tenkan’s lowest low became $767.46, from 6 August, so tenkan jumped to $773.415, up $3.47 in a day. The weighted average barely moved, to $773.70, because the close that left carried a weight of 1 and the new close sat close to the line.

Same nine sessions, two different events. The weighted average responded to closes; tenkan responded to a single low aging out. Neither number is the wrong one: they answer different questions.

The original data

The bitcoin sample. Bitcoin’s daily bars from Yahoo Finance, each covering one UTC calendar day, from mid-September 2014 through 4 October 2026, with the incomplete 5 October bar dropped. The first 78 days only build the cloud and the first day-to-day comparison, so the count runs over 4,323 days from 4 December 2014.

The rules, as coded. Weighted average: the last nine closes weighted 9 down to 1, divided by 45. Tenkan and kijun: the midpoint of the 9-day and 26-day high-low range. Span A: the average of the two; span B: the 52-day midpoint. Both spans are drawn 26 days ahead, so the cloud on any day was built from the bar 26 days earlier. Inside means the close sat between the two spans, edges included.

What the counts say and do not say. Inside the cloud, 133 crosses in 668 days; outside, 798 in 3,655. That is one market on one interval, and nothing here measures what price did after a cross. Every day’s close, weighted average, tenkan, kijun, both cloud edges, the cloud state, the cross and the turn are in the bitcoin weighted average and cloud file.

The SPY worked example. The nine closes, highs and lows behind the August 2026 example, both lines on each day and the weight each close carried are in the SPY weighted average and tenkan file, taken from Yahoo Finance daily bars. The weighting moved the line $1.10 above the simple average of the same nine closes, $772.58.

When it fails

The characteristic failure is reaching for the line when the cloud goes quiet. Price inside the band is a reading: the framework found no direction in the recent range. Swapping in a line that cannot express that discards the only information you had.

A second failure is assuming the cloud and the line fail on the same days. On bitcoin the line’s crosses were spread almost evenly across cloud states, so the cloud did not mark its choppy stretches.

A third is counting both as independent confirmation. They share their price history, so agreement is not two votes.

A fourth is reading a tenkan jump as a market move. As on 17 August, it can be an old extreme leaving the window.

And a fifth is placing a stop at a cloud edge or the line. Neither measures how far price travels in a bar.

Related

The weighted moving average covers linear weighting of closes. Ichimoku covers the framework and its no-trend reading. And tenkan-sen is the line that jumped on 17 August.

This page is educational, not financial advice. Test every idea on your own charts before risking money.