Three Black Crows Candlestick Pattern: Confirmation vs Late Entry

Define each candle before interpreting the pattern

TradingView describes three consecutive long bearish bodies, opens inside the previous body, and closes at or near the low. For an explicit three-bar detector, label the candles 1, 2, and 3 from oldest to newest. Require close < open for each, then close 1 > close 2 > close 3. Require the second and third opens to lie strictly between the previous close and open. Long body and near the low need measurable limits. One illustrative specification uses the mean absolute body of the 20 bars before candle 1 as a fixed baseline: every body must exceed that baseline, and every lower wick must be no more than 10% of its candle's range. These are example research settings, not TradingView's published detector parameters or an established optimum.

• Body: open minus close for each bearish candle. Descending closes do not require decreasing body sizes.
• Open: previous close < current open < previous open for candles 2 and 3. Decide separately whether equality is allowed.
• Lower wick: close minus low. Divide by high minus low to measure proximity to the low; reject zero-range bars.
• Upper wick: high minus open. A separate cap, such as 20% of the range, is an optional filter, not a substitute for checking the lower wick.
• Prior advance: define it using bars before candle 1. For example, compare the preceding close with the close ten bars earlier; record this as a chosen context filter.

A shrinking-body example still needs three lower closes

Assume a prior advance and a frozen average body of 3 price units. In this hypothetical sequence, bodies measure 5, 4, and 3.5 units. They shrink but each remains longer than the baseline. Candle 2 opens at 117, inside 115 to 120; candle 3 opens at 115, inside 113 to 117. The closes descend from 115 to 113 to 111.5. The lower wicks are 0.4, 0.4, and 0.3, about 6.9%, 8.3%, and 7.3% of their respective ranges. All pass the illustrative 10% limit. If the third low were 109 with its other prices unchanged, the lower wick would be 2.5 / 6.3, about 39.7%, and the sequence would fail that rule. A smaller third body alone neither invalidates the pattern nor proves exhaustion.

A reversal near a top and a late short are different setups

After an advance, the sequence can mark the start of a bearish reversal hypothesis. After a long decline, the same candles provide no evidence of a newly broken uptrend. Record the prior trend before inspecting the three bars; otherwise the decline within the pattern can be mistaken for its context. Consider a hypothetical short fill at 111.5 in the example above, with an exit trigger at the first high of 120.4. The price distance at risk is 8.9 units. If support identified before entry is at 110, only 1.5 units remain before that level, about 0.17 times the risk before costs. The shape passes, but it provides little room under that particular exit plan. Support may break or hold; the calculation is a comparison of distances, not a price forecast.

Confirmation costs time; liquidity determines the fill

The third candle must close before its final body and wick can be checked. An intrabar marker can disappear if price rebounds before that close. Waiting for bar close addresses this unfinished-candle change; it does not guarantee that every script avoids every form of repainting. Entering after the third close, selling a later break of the third low, and placing a limit order on a rebound are different hypotheses. The breakdown order may fill lower after a gap. The rebound order can remain unfilled or fill just before a recovery. A chart's last price is not proof that the intended position size could trade there. Check spread and available liquidity for the instrument and session, and include commissions and slippage in the study. For stock shorts, also check borrowing availability and costs. A sharp rebound produces losses even when the historical pattern remains correctly identified. The SEC notes that a stock short has potentially unlimited losses; a stop trigger does not guarantee its execution price. Above-average candle volume is not proof of institutional selling or sufficient liquidity at the later entry.

Compare exits after fixing the entry and extension rule

An extension filter can quantify how much of the move has already occurred. For an illustrative test, divide the first open minus the third close by the 14-period ATR measured immediately before candle 1. Require a positive ATR and compare all signals with a version that skips values above 2. Keep that threshold separate from the pattern definition. Then hold the detector, filter, entry, position size, market, timeframe, and dates fixed while comparing exits. One experiment can use a stop above the first high and a target two times the initial price risk below entry. Another can keep the same stop and close after five completed bars if the stop has not triggered. Specify whether the entry bar counts and how simultaneous conditions are handled. Neither exit is a recommendation. Use ordinary price candles, document fill assumptions, and include costs. TradingView warns that non-standard charts can use synthetic prices that distort strategy results. Report trade count, net results, drawdown, and the signals removed by the filter, then evaluate the selected version on a separate date range. This page does not supply a measured win rate or a live scanner.