Dark Cloud Cover Candlestick: Gap and Penetration Rules
Dark cloud cover is a potential bearish reversal after an advance: a long bullish candle is followed by a bearish candle that opens higher and closes below the first body midpoint. TradingView requires the second open above the previous high; StockCharts uses the previous close. State the gap convention before comparing labels.
Key Takeaways
- The midpoint is (first open + first close) / 2. A lower wick crossing it does not satisfy a closing-price test.
- Keep a strict high-gap detector separate from a close-gap or no-gap variant.
- A close at the midpoint fails a strictly-below rule. A full-body engulfing receives a separate label in the examples here.
- Pattern completion, later confirmation and a trade entry are separate decisions. None establishes a profitable strategy by itself.
Define the trend and first candle before measuring penetration
TradingView describes a first body larger than average and offers SMA50 or SMA50/SMA200 trend filters. Its help page does not specify every implementation detail. For the reproducible examples below, use the following custom research settings, rather than assuming they reproduce every built-in label.
| Test | Example rule | Worked input |
|---|---|---|
| Prior uptrend | On the bar immediately before candle 1: close > SMA50 > SMA200 | 99 > 96 > 92 passes; the averages require sufficient earlier history |
| Bullish first candle | First close > first open | 110 > 100 passes |
| Long first body | First body exceeds the mean absolute body of the preceding 20 bars, excluding candle 1 | 10 > an assumed historical mean of 3 passes |
- The 20-bar body average and pre-pattern trend sampling are declared example choices, not a universal definition.
- If history is insufficient, mark the trend or body comparison unavailable. Do not treat missing averages as zero.
- A two-candle shape during a decline does not satisfy this bearish reversal setup. Disabling the trend filter changes the question being tested.
Calculate a valid dark cloud cover step by step
Use hypothetical candle 1 with open 100, high 111, low 99 and close 110. Candle 2 has open 112, high 113, low 103 and close 104. The first body is 10 and its midpoint is 105. Penetration is measured down from the first close: (110 - 104) / (110 - 100) = 60%. The second close remains inside the first body, so this example has partial coverage.
| Condition | Calculation | Result |
|---|---|---|
| High gap | 112 > 111 | Pass |
| Bearish second candle | 104 < 112 | Pass |
| Below first body midpoint | 104 < (100 + 110) / 2 = 105 | Pass |
| Partial body coverage | 100 < 104 < 105 | Pass |
| Bar | Open | High | Low | Close |
|---|---|---|---|---|
| Candle 1 | 100 | 111 | 99 | 110 |
| Candle 2 | 112 | 113 | 103 | 104 |
Hollow bodies close at or above the open. Filled bodies close below the open.
Candle 2 opens at 112 above the previous high of 111 and closes at 104 below the body midpoint of 105. Earlier trend and body-average history is assumed by the stated example settings and is not shown in this two-bar diagram.
Strict gap, close gap and continuous-market variants
The two cited definitions use different opening thresholds. For candle 1 above, an open of 110.50 is above its close but below its high: it passes the StockCharts opening test and fails the TradingView opening test. An open exactly at 110 passes neither strict greater-than test. A no-gap rule is a custom variant, even when it is useful for studying continuously traded markets.
| Gap mode | Opening rule | Example |
|---|---|---|
| TradingView high gap | Second open > first high | 112 > 111: passes |
| StockCharts close gap | Second open > first close | 110.50 > 110: passes |
| Custom no-gap variant | Second open >= first close, with no tolerance below it | 110 >= 110: passes only this relaxed opening rule |
- Keep all other body, trend and close conditions fixed when comparing gap modes.
- If allowing an opening tolerance, record its size and unit, such as one minimum price tick; do not silently round away a failed gap.
- Use the same symbol, timeframe, data feed and regular or extended session settings. Those choices determine the OHLC being tested.
Near misses and the bearish engulfing boundary
These cases reuse the first candle and the passing context above. Here dark cloud cover is reserved for partial body coverage, with the second close above 100. At or below 100, record full body coverage separately so an engulfing candle is not counted twice. This is an explicit classification convention; a detector checking only the midpoint may allow overlap.
| Second-candle change | Calculation | Classification |
|---|---|---|
| Close 105, open 112 | (110 - 105) / 10 = 50% | Midpoint equality: fails strictly below |
| Close 106, open 112 | (110 - 106) / 10 = 40% | Insufficient penetration |
| Low 103 but close 106, open 112 | Low < 105; close > 105 | Wick penetration alone fails |
| Open 111, close 104 | Open equals the first high | Fails strict high gap; passes close gap |
| Open 110, close 104 | Open equals the first close | Only the declared no-gap variant passes |
| Close 99, open 112, low 98 | 112 > 110 and 99 < 100 | Full body coverage: bearish engulfing |
- For the close-99 case, the second low is adjusted to 98 so its OHLC remains valid.
- A close above 105 does not turn these bullish-then-bearish candles into a piercing line. Piercing has the opposite candle order and follows a decline.
- The gap concerns the opening price. The whole second candle need not remain above the first high; it must trade back into the first body.
Confirmation and failure conditions are additional rules
Wait for candle 2 to close before evaluating its final penetration. For a separate confirmation experiment, require the very next candle to close below the second low of 103. A next close of 102 qualifies; a low of 102 followed by a close of 104 does not. Waiting changes both signal timing and potential execution price. It does not guarantee continued selling.
- Illustrative cancellation rule: cancel the pending setup if price trades above the two-candle high of 113 before confirmation, or if the next candle fails the confirmation close.
- The second high is 113, not its opening price of 112. A stop derived from the high must use the high, including its wick.
- If one historical bar crosses both confirmation and cancellation levels, OHLC alone may not reveal event order. Specify a conservative ordering rule or inspect lower-timeframe data.
- Mark nearby support and the distance to it before testing a short entry. A valid shape can still form too close to support for the chosen exit plan.
- Volume or momentum filters are optional research conditions. Volume alone does not identify institutional selling, and no win-rate advantage is established here.
- Specify entry timing, position size, exit rules, spread, commissions, borrow costs where relevant and slippage. A stop order can fill beyond its selected level.
Check the same definition on TradingView and Pineify
On TradingView, search Indicators for Dark Cloud Cover and inspect the selected trend settings. Compare labels only after choosing a gap convention and checking completed candles. For a visual review, upload a clear chart to Pineify AI Chart Analysis with the symbol, timeframe, session and the two candles marked. Include readable OHLC values when a one-tick difference matters; image interpretation can miss those details.
- For a custom detector, give Pine Script AI Coding Agent the trend sample bar, body-average window, gap mode, penetration threshold, partial-body boundary and close requirement.
- Use the worked case and near misses as acceptance cases. Check generated source and chart labels before attaching alerts or testing a strategy.
- This guide does not scan live markets or execute orders. A chart review does not replace exact OHLC comparisons.
Pineify next step
Review your marked two-candle setup
Use AI Chart Analysis to review the visible pattern and surrounding price structure. For exact gap and penetration rules, ask the Coding Agent to draft a detector, then verify its output against the numerical cases above.
Educational information, not investment advice. All numbers and research rules in the examples are hypothetical. Patterns can fail, trading can result in losses, and historical results do not guarantee future performance.