What is a dark cloud cover pattern? The usual answer — a bullish candle, then a bearish candle that opens above it and closes deep into its body. I suppose that is an answer too. But for me the dark cloud cover meaning sits in one number — how far the second candle closes below the middle of the first one. Below that middle, the buyers of the first day are losing money at the close.
Above it, you have two candles and nothing else.
This is a bearish pattern and only a bearish pattern. There is no bullish dark cloud cover — the mirror image has its own name, and I put it in the table below, together with the pattern people most often confuse it with.
| Pattern | Where candle 2 closes | Reads as |
|---|---|---|
| Dark cloud cover | Below the middle of candle 1, above its open | Bearish, after a rise |
| Bearish engulfing | Below the open of candle 1 | Bearish, deeper |
| Piercing line | Above the middle of a bearish candle 1 | Bullish, after a fall — the opposite of dark cloud cover |
Candle 1 is bullish in the first two rows and bearish in the third. The whole difference between the three patterns is one close against one line.
Dark cloud cover pattern anatomy
What is dark cloud cover pattern, in numbers? Two candles. The first is bullish — close above open, and better a long body after a few days of rising. The second opens at or above the first close, goes up a little, then sellers take over. It closes under the middle of the first body. Not under the first open — that would already be a bearish engulfing pattern.
The name is Japanese. Steve Nison brought it west in his 1991 book, and the picture is a sunny day covered by a cloud. You will also meet it written as dark cloud candlestick pattern or candlestick dark cloud cover — the same two candles.
Why does the middle matter so much? The middle of the first body is roughly the average price buyers paid during that candle. When the second candle closes below it, more than half of those buyers hold a loss. This is the whole reason the dark cloud cover candlestick is read as bearish — not the colour, not the shape, one close against one average.
One more thing, about the open. In Nison’s version the second candle opens above the high of the first — a gap up. Stocks do that between sessions. Forex almost never does. I checked the daily EUR/USD data for 2026, 213 trading days from 2 January to 27 October, and the open was above the previous high on 3 of them. Three. If you wait for the gap on a forex chart, you wait all year.
The forex version of the rule
This is where I disagree with most articles on the dark cloud cover candle pattern. They copy the stock-market definition with the gap, then show a forex chart where there is no gap at all.
For me the rule on a currency pair is simpler: the second candle opens at or near the first close, pushes above it, and still closes below the middle of the first body. The push above matters — it shows buyers tried and failed. The gap does not. On a market that trades 24 hours a gap only appears over a weekend.
I write trading robots, so I need every rule as four numbers per candle — open, high, low, close. For the robot the dark cloud cover candlestick pattern is three comparisons: close one above open one, close two below the middle of body one, close two above open one. There is no “looks bearish” in a log file. Either the comparisons are true or the pattern does not exist.
Written down this way, it also shows how little the pattern is — three inequalities on two bars.
EUR/USD, 27 and 28 January 2026
One real example, with prices. EUR/USD had risen from 1.1646 at the close on 19 January to 1.2042 at the close on 27 January. That day was the biggest candle of the month — open 1.1880, high 1.2081, close 1.2042, a 162-pip body.
On 28 January the pair opened at 1.2042, touched 1.2045 and closed at 1.1954. The middle of the previous body was 1.1961. The close was 7 pips under it. A bearish dark cloud cover pattern at the high of the year, by the forex version of the rule.
What came next? On 29 January the close was 1.1971, 17 pips higher — for one day the pattern looked wrong. Then 30 January closed at 1.1851 and 2 February at 1.1790. From the pattern close that is 164 pips in three trading days — about $1,640 on one standard lot, $16.40 on a 0.01 lot. The high of 1.2081 was not seen again in 2026, and the low of the year came in October at 1.1113.
I am not sure a stricter filter would have let this one through. Seven pips below the middle is a thin margin — a different broker’s daily close can sit on the other side of the line. The prices here are indicative daily rates from Pound Sterling Live, and an MT5 broker closes the daily bar at its own server time. Your terminal may show a close a few pips from mine.
This is normal, and it is also a warning: a pattern that depends on 7 pips depends on your broker.
Seven patterns, three worked
To test the rule I ran the same three comparisons over all of 2026 on the EUR/USD daily chart. Seven dark cloud cover candlestick patterns came out. For each one I noted how much of the first body the second candle covered, and where the close was five trading days later.
| Candles | Body covered | Close five days later |
|---|---|---|
| 12–13 January | 72% | +83 pips |
| 27–28 January | 54% | −148 pips |
| 20–21 April | 92% | −30 pips |
| 9–10 June | 73% | −33 pips |
| 31 Aug – 1 Sep | 71% | +31 pips |
| 8–9 October | 64% | +37 pips |
| 13–14 October | 93% | +76 pips |
EUR/USD daily, 2 January – 27 October 2026, indicative rates. Rule: candle 2 opens at or above candle 1 close, closes below its middle and above its open.
Three of seven were lower after five days. That is close to a random result, and it agrees with what Thomas Bulkowski found on stocks — in his Encyclopedia of Candlestick Charts the dark cloud cover acts as a bearish reversal 60% of the time, which he rates as poor.
The one pattern that paid well came after a real rise. It was the only one of the seven where the first candle closed above the close four days earlier. The other six formed inside flat weeks. A dark cloud over a flat field is just weather.
The body size matters as well. The June pattern had an 11-pip first body. The daily range on EUR/USD is usually 60–80 pips, so an 11-pip body is noise, and the second candle covered it with 8 pips. If the robot is allowed to count that, it counts everything. I would set a minimum body — half of the 14-day average true range or so — before the pattern is allowed to exist.
Trading the dark cloud cover
So how to trade dark cloud cover pattern signals on a real account? For me there are four decisions, and the pattern itself makes none of them.
The first decision — where the stop goes. Above the high of the two candles is the usual answer. In the January case that is 1.2081, 127 pips above the pattern close — about $1,270 on one standard lot, $12.70 on a 0.01 lot. A stop above the second candle only, at 1.2045, would be 91 pips, and on 29 January it would not have been hit.
I take the wider one anyway. The narrow stop is where the first pullback takes you out.
The second decision — what you want from the trade. With a 127-pip stop and a two-to-one target you need 254 pips, which is 1.1700. EUR/USD reached that level on 2 March, five weeks after the pattern. The first 164 pips came in three days and the rest took a month. If you cannot hold a position for a month, the two-to-one target is not yours, and a one-to-one target at 1.1827 was done on 2 February.
The third decision — confirmation. Many traders wait for the next candle to close lower before they sell. In January that next candle closed higher, so the confirmed entry came on 30 January at 1.1851 — 103 pips worse than the pattern close. Confirmation is not free. It buys you fewer false patterns and it costs you the first third of the move, every time.
The fourth decision — the trend on the higher timeframe. This is the one the robot cannot make from two candles. I prefer a trend filter that is already on the chart before the pattern appears, not one I add afterwards to justify the trade. Our free Smart Trend Trading System for MT5 is one way to see the direction. A 50-day moving average is another.
What matters is that the filter was chosen before the cloud, not because of it.
Dark cloud cover vs bearish engulfing
The question I see most often — dark cloud cover vs bearish engulfing, which one is stronger? The difference is one line, the open of the first candle. Close below it, bearish engulfing. Close above it, dark cloud cover. By that logic the engulfing is the stronger of the two, the sellers took back the whole day and more.
For me they are the same event at two depths. In a robot I treat the depth as a parameter, not as two names. Half the body, the full body, 120% of the body — one input, three patterns.
And the opposite of dark cloud cover is the piercing line — a bearish candle, then a bullish candle that opens at or below its close and finishes above its middle. Same logic, same checks, other direction. Both belong to the same family as the hammer and the doji: one or two candles that say the last move ran out of buyers, or out of sellers, for now.
Common mistakes with the pattern
The first mistake — reading dark cloud cover candlestick patterns on five-minute charts. On M5 the pattern appears twenty times a day and the spread alone can make it. The pattern was described on daily candles, and my seven cases are daily candles. H4 is the lowest I would go.
The second — no rise before the cloud. Six of my seven cases had no rise behind them, and none of those six fell more than 33 pips in the next five days. A reversal pattern needs something to reverse.
The third — no minimum body. An 11-pip cloud on EUR/USD is not a cloud.
The fourth, and the one I see most — treating the two candles as a forecast rather than a trigger. The pattern says that on one day sellers won back half of what buyers made. It does not say the trend has ended. The 29 January close, 17 pips higher, is the true picture of what a dark cloud cover candle can do the next day: nothing.
So the dark cloud cover pattern for me is a middle line, a close under it, a rise before it and a body big enough to matter. The more of these you check before you press sell, the steadier the trade — and the fewer clouds you see.
Daily EUR/USD open, high, low and close for 2026 — Pound Sterling Live, indicative rates, retrieved 7 November 2026. Pattern definition — Steve Nison, Japanese Candlestick Charting Techniques, 1991. Pattern statistics on stocks — Thomas Bulkowski, Encyclopedia of Candlestick Charts, 2008, summary at thepatternsite.com.
Signal2forex.com - Best Forex robots and signals
