Ichimoku on the 15-Minute Chart Tested: 4 of 10 Gold Runs Passed Every Check, and 0 of 10 on EURUSD

Ichimoku is usually sold as a system for daily charts, yet one of Google’s own suggestions for it is “ichimoku settings for 15 min chart”. Earlier I ran 329 Ichimoku cloud breakouts on EURUSD M5: they made +0.29 pips a trade before any spread, less than a third of a 1-pip cost. Here I built the 15-minute chart from the same data and asked a narrower question: does any standard Ichimoku entry make money on M15 after the spread, and does it hold up?

How did I test Ichimoku on the 15-minute chart?

• Data: 16,926 five-minute EURUSD bars from June 3 to August 26, 2026, joined into 5,642 fifteen-minute bars, and 13,736 gold bars from June 16 to August 26, joined into 4,577.
• Settings: the default 9/26/52 for the five rules below, plus seven other popular settings in the grid further down. What each line means is in my Ichimoku settings guide for MT4.
• Cost: 1 pip a trade on EURUSD, $0.30 an ounce on gold.
• Entry: at the close of the signal bar. One trade at a time: a signal during an open trade is skipped.
• Exit A: stop 1.5 ATR (14), target 3 ATR, closed after 48 bars (12 hours) if neither is hit.
• Exit B: close after 12 bars, 3 hours, with no stop or target.
• Checks: positive after cost; positive in both halves, before and after July 15; and above the 95th percentile of 300 runs of random entries with the same exit and the same number of trades.

The five rules:

• Cloud breakout: a bar closes outside the cloud after closing inside or on the other side.
• TK cross: Tenkan-sen crosses Kijun-sen.
• TK cross with the cloud: the same cross, taken only when price closes above the cloud for a buy or below it for a sell.
• Kijun cross: a bar closes across the Kijun-sen.
• Strong signal: the TK cross with the cloud, plus the Chikou span above price 26 bars back for a buy, plus a green cloud ahead. I read Chikou as it is known live, today’s close against the close 26 bars ago. Chikou is drawn 26 bars back, so on a history chart the point over a candle is a later close; reading it there would look into the future.

Which Ichimoku rules made money on gold M15?

Dollars per ounce after a $0.30 cost. “Random top 5%” is the 95th percentile of 300 runs of random entries with the same exit and number of trades. Gold rose from $4,351 to $4,675 in the test, so random entries lost little: the median random run lost $0.29 to $0.61 an ounce a trade.

The pattern is clear on gold: the TK cross is the entry that worked, and the cloud made it better as a filter. Adding “price on the right side of the cloud” cut the trades from 133 to 79 and lifted the result from +$2.50 to +$4.59 an ounce. Both longs and shorts made money: +$5.34 and +$3.83 a trade.

Most trades still lost. The best rule won 46.8% of its trades, the TK cross alone 40.6%. With a target twice the stop that is enough. The worst drawdown of the TK cross with the cloud was $88.60 an ounce, about 19 average trades.

Why did nothing pass on EURUSD M15?

Pips per trade. The other four runs, all with the 12-bar exit, lost between 0.26 and 2.00 pips a trade after cost.

Ichimoku on EURUSD M15 was not useless. Most rules made something before the spread. But the median 15-minute ATR was about 4 pips, so a 1-pip spread takes a quarter of a typical bar, and the edges were around 1 pip. Two runs ended positive. Both lost in the first half, and only the TK cross with the ATR exit beat random entries. That is a pass on one check of three.

The TK cross with the cloud, the best rule on gold, lost on EURUSD. The same rule behaving differently on two markets in the same 10 weeks is the warning I gave in gold vs EURUSD with one rule: an Ichimoku result belongs to its market, not to the indicator.

Is Ichimoku better on M15 than on M5?

Eight settings: 9/26/52, 7/22/44, 6/13/26, 10/30/60, 20/60/120, 8/22/44, 12/24/120 and 5/15/30. Pairs with fewer than 30 trades are left out, which removed one gold M15 pair. On gold M5 the default Kijun cross ended at exactly $0.00 a trade, counted as not positive.

Yes, on both markets. On EURUSD M5 every one of the 40 pairs lost after cost, and only 2 made money before it. On M15, 26 of 40 made money before cost. The slower chart gives Ichimoku the same thing it gives other trend tools: bars big enough that the spread is a smaller share of them. On M5 a 1-pip spread is half of a typical EURUSD bar, on M15 a quarter, as the ATR settings guide measures timeframe by timeframe. Ichimoku was built for slower charts, and the 5-minute chart is where it struggled most.

What does the cloud look like on M15?

On both markets price spent about 14% of the time inside the cloud: 14.4% on EURUSD M15 and 14.5% on gold M15. That is why a cloud breakout is not rare: price leaves the cloud often, and most exits are not the start of a trend. The median cloud on M15 was 1.3 ATR thick on EURUSD and 1.19 on gold, about the same as on M5. The cloud works better as a filter than as a trigger. My Ichimoku vs plain trend following test found the same on M5 before costs: the TK cross alone lost, and the same cross on the right side of the cloud turned positive. On gold M15 that filtered cross was the best rule after costs too.

What are the limits of this result?

• 10 to 12 weeks of data. Gold covers June 16 to August 26, 2026, a period when it rose $324 an ounce. Shorts made money too, which helps, but a falling or flat gold may behave differently.
• Small samples. The TK cross with the cloud had 79 trades, only 23 of them in the first half. Treat +$4.59 as a sign of an edge, not as a number to expect.
• The exit matters. The same TK cross with the cloud made +$4.59 with the ATR exit and +$1.37 with a 12-bar exit, which lost in the second half.
• Bar data, not ticks. When a stop and a target sit in the same bar, I took the stop first. My tick data vs spread test shows how much bar order can change a result.

If you want to trade the gold rule, test it on a demo account first; how long to demo trade on MT4 explains why a few weeks of demo can still mislead, and backtest vs demo account shows how far the two drifted for 12 tuned setups.

Frequently asked questions

Does Ichimoku work on the 15-minute chart?

On gold M15 four of ten rule and exit runs passed every check in my test, led by the Tenkan/Kijun cross on the right side of the cloud at +$4.59 an ounce a trade. On EURUSD M15 none passed after a 1-pip spread.

What is the best Ichimoku strategy for M15?

In this test, the Tenkan/Kijun cross taken only when price is above the cloud for a buy or below it for a sell, with a 1.5 ATR stop and a 3 ATR target, on gold.

Is the Ichimoku cloud breakout a good entry on M15?

Not here. It made +$0.98 an ounce on gold with the ATR exit but lost in the first half, and it lost after cost on EURUSD with the same exit.

Is Ichimoku better on M15 or M5?

M15. On EURUSD M5 all 40 setting and rule pairs lost after cost; on M15, 26 of 40 made money before cost and 4 after it. On gold, positive pairs went from 9 of 40 on M5 to 26 of 39 on M15.

Does the Chikou span help on M15?

Read honestly, as today’s close against the close 26 bars ago, the strong signal made money on gold and lost on EURUSD. Read off the chart, where Chikou is drawn 26 bars back, it looks far better than it is.