Ichimoku Settings for MT4: What to Change and What to Leave Alone

Ichimoku Kinko Hyo ships with MetaTrader 4 and needs no installation, no DLL and no paid version. It also has more moving parts than any other built in indicator, which is why most people either avoid it or change its numbers on the first day. This guide covers where it lives, what each input actually does, and which settings survived measurement on real data.

Where is Ichimoku in MT4?

Insert → Indicators → Trend → Ichimoku Kinko Hyo. It draws on the price chart, not in a subwindow, unlike Stochastic or ADX. The dialog has exactly three numeric inputs.

Note what is missing: there is no “Apply to” price option and no method choice. Every line is built from highs and lows, never from closes. That is a real difference from moving averages, and it is why Ichimoku lines flatten out in a way an EMA never does — when the 26-bar range stops expanding, Kijun becomes perfectly horizontal.

What are the five lines on the chart?

• Tenkan-sen — fast line, drawn at the current bar.
• Kijun-sen — slow line, drawn at the current bar. A flat Kijun is the single most useful visual in the whole system: it says the 26-bar range has not moved.
• Senkou Span A — the midpoint of Tenkan and Kijun, shifted 26 bars forward.
• Senkou Span B — the 52-bar midpoint, also shifted 26 bars forward. The area between A and B is the cloud.
• Chikou Span — today’s close, shifted 26 bars backward. It is the current price plotted in the past, nothing more.

The cloud ahead of price is not a forecast. It is built entirely from data that already exists and then displaced forward, which means it repaints nothing but also predicts nothing. Understanding that one sentence removes most of the mysticism around this indicator.

Which settings actually tested better?

I ran the cloud breakout entry on four different parameter sets across the same 16,926 EURUSD 5-minute bars, holding each signal two hours.

The 10 / 30 / 60 row deserves attention, because it is the change you will see recommended most often on forums — round the numbers up a bit for modern markets. On this data it destroyed the edge completely, moving from +0.29 to -1.71 pips per trade. A 15% parameter change flipped the sign. That is the clearest warning in this article: these numbers are not robust, and any setting you cannot justify with a measurement is a guess.

The 20 / 60 / 120 row is the interesting one. Doubling everything gave the best result of the four, and it fits what I keep finding elsewhere: on 5-minute forex data, slower is usually better, because most short-horizon signals are noise. But note the cost — 224 signals against 329, a third fewer opportunities. I would not switch to it on one quarter of data. I would test it on my own pair first.

The setting that is not in the dialog

Hold time mattered more than any parameter. Identical 9 / 26 / 52 signals produced -0.44 pips at 30 minutes, +0.29 at two hours and -0.45 at eight. Nothing in the settings window changes that, and no amount of tuning Tenkan will fix a hold length that does not match the signal. The full breakdown is in my cloud breakout test.

Three rules that keep the chart usable

• Use the cloud as permission, not as a trigger. Its measured job is to say which direction you are allowed to trade. In my comparison test, adding nothing but “price on the correct side of the cloud” turned a Tenkan/Kijun system from -392 pips to +308.8.
• Do not add another trend indicator. Ichimoku already contains a fast line, a slow line, a range projection and a lag line. A moving average on top is a sixth view of the same price. Filtering a trend signal with another trend signal made results worse in every test I have run, including my moving average settings guide.
• Turn off what you do not use. If your rules never mention Chikou Span, set its colour to None. A line you do not act on is not neutral — it is something you will eventually rationalise a trade with.

Frequently asked questions

What are the best Ichimoku settings for MT4?

The default 9 / 26 / 52 on this data, with 20 / 60 / 120 as the only variant that measured better. Avoid 10 / 30 / 60: it was the worst of four at -1.71 pips per trade, despite being the most frequently recommended change.

Does Ichimoku repaint?

No. Senkou Span A and B are drawn 26 bars ahead of price, which looks like a forecast but is only past data displaced forward. Chikou Span is today’s close drawn 26 bars back. Once a bar closes, no Ichimoku value on it ever changes.

Why does Ichimoku use 9, 26 and 52?

They come from the Japanese trading week of the 1930s, which ran six days. 26 was roughly a month of sessions and 52 two months. Nothing about them was derived for currency markets or for intraday bars, which is exactly why they are worth testing rather than trusting.

Can I use Ichimoku on a 5-minute chart?

You can, and I did, but hold your expectations where the data puts them: the best default-setting result was +0.29 pips per trade before spread. It was designed for daily charts, and my results suggest slower parameters suit fast charts better than faster ones do.

Should I trade the Kumo twist?

I did not test the twist — the point where Senkou A crosses Senkou B — in this run, so I will not give you a number for it. What I would note is that the twist happens 26 bars ahead of price, so acting on it means acting on a projection whose input is already two hours old on a 5-minute chart.

Is Ichimoku better than the DeMARK trend line?

Different jobs. Ichimoku describes a range and a direction over 26 to 52 bars. A DeMARK trend line marks a specific structural level. I use the second for entries and the kind of context Ichimoku gives for permission, which is the same division that worked in the tests above.