Homing Pigeon Candlestick Pattern: Rules, Trading and Tested Results

The homing pigeon candlestick pattern is a two-candle formation that appears in a downtrend: a long bearish candle, followed by a smaller bearish candle whose body sits inside the body of the first. Classic candlestick analysis classifies it as a bullish reversal pattern: sellers still push the market lower, but with far less force than the session before.

This guide explains how to identify the homing pigeon pattern, the psychology behind it, how traders set entry, stop loss and target, and how it differs from the bullish harami. It also answers a question that rules alone cannot: what actually happened next. We applied the same rule to full price history, which gave 2,432 occurrences on daily charts alone across US stocks, ETFs, indices, commodities, crypto and forex, and compared the outcome with random bars on the same assets.

What is the homing pigeon candlestick pattern?

In technical analysis, the homing pigeon is a two-line bullish reversal pattern from Japanese candlestick charting. The name comes from the image of a bird returning home: the second line "flies back" inside the first one. Both lines are bearish (black or red, closing under their opening level), which makes it unusual among bullish signals.

Definition and conditions

A homing pigeon pattern has four conditions:

• the market is in a downtrend before it;
• the first candle is a long bearish line;
• the second candle is also bearish, with a shorter real body;
• that smaller body lies inside the first one: it starts below the first open and ends above the first close.

Our test bench applies the same rule without judgment: after a decline (the last close before the setup is at least one ATR below the close 10 bars earlier, in the lower half of the last 40 closes), a long bearish line, then a short bearish line inside the first body. "Long" and "short" are measured against the 10 prior bars, as in the open source TA-Lib library.

What the homing pigeon pattern indicates

It is a pattern that indicates slowing bearish momentum. Day one shows the bears in full control. Day two still ends down, but it opens above the prior close and its body stays inside the first body, so the selling pressure is visibly smaller. Classical analysis reads it as a potential turn, not a confirmed one.

How to identify the homing pigeon pattern

Check four points in order.

Step 1: confirm the downtrend

There must be a downward move to reverse. A simple test: the setup comes after a fall over the previous 10 bars. Two such lines in a sideways range do not qualify.

Step 2: find a long first candle

The first line has a real body clearly larger than the recent average: the session where the bears look strongest. On the median daily occurrence in our bench, the two-bar formation spans 3.92% of price, or 1.35 times the average true range (ATR): a larger than usual move.

Step 3: check the second candle

Day two also closes under its own open, but it opens above the first close and closes above it too. It can be almost a doji, a very small line inside the first one.

Step 4: verify the open and close prices

Only the bodies matter, not the shadows. Compare the open and close prices: day two must open below where day one opened and close above where day one closed. The wicks may extend outside the first body. Wait for day two to close before you mark it.

Our bench is slightly looser than this rule: like TA-Lib, it also accepts a second body that shares one end with the first (for example a close exactly equal to the first close), as long as the second body is strictly smaller. Such setups count in our statistics, but the cases written up below all meet the strict rule.

The psychology behind the homing pigeon formation

The formation tells a two-day story. On day one, the bears dominate: price falls through the session and closes near the bottom, which confirms the bearish mood. On day two, the market opens higher, a sign that selling has paused. Sellers push price down again, but they cannot reach the prior close. The bar is bearish, yet smaller and contained.

Classical analysis interprets this as exhaustion: the same participants are still active, with less conviction, and some buying pressure absorbs the selling. If buyers then lift price above the top of the setup, the bears have lost control. That reading is plausible, but it is a story about two bars. Whether the market actually turns more often after this setup than after any random day is a question for data, answered in the reliability section.

Bullish vs. bearish homing pigeon: key differences

The classic version is bullish. Its mirror image can also be drawn as a bearish homing pigeon: in an uptrend, a long bullish candle followed by a smaller bullish candle contained in the first, read as a potential bearish reversal, with an entry under the setup instead of above it.

The bearish version is not part of the classic literature, and we have not tested it as such. Its closest tested relative is the bearish harami (a tall bullish line, then a small one inside it, after an uptrend). On daily data, its 2:1 setup reached the target 28.6% of the time against 27.5% at random, with an expectancy of −0.14 R before costs.

Homing pigeon vs. bullish harami: understanding the distinction

The homing pigeon is a special case of the bullish harami. Both need a prior decline, a tall bearish first line and a second body inside it. The harami compares bodies, while an inside bar compares full ranges (highs and lows). What separates the homing pigeon from the bullish harami is the color of day two:

• in a bullish harami, it can be bullish or bearish (the classic harami has a bullish one);
• in a homing pigeon, it must be bearish.

Because the market still ends lower on day two, the homing pigeon can be read as weaker than a harami. Our data does not show a large gap. Over the same history, the bullish harami (13,897 trades) reached its 2:1 target 36.8% of the time against 36.3% on random dates, with an expectancy of +0.11 R; the homing pigeon (901 trades) reached it 34.3% of the time against 34.9%, with +0.03 R. In both cases the difference with chance is small.

When day two has no body at all, it becomes a harami cross. It is also different from the engulfing pattern, where day two covers day one instead of sitting inside it.

How to trade the homing pigeon pattern effectively

The classic plan buys a break above the high of the two bars, with a stop loss below their low and a target set as a multiple of the risk. This is the position our bench simulates on every occurrence.

Entry: wait for confirmation

The setup is a warning, not a buy signal. One plan waits for confirmation: a bullish candle that ends above day two, or a buy stop placed just above the high of the two bars. In our tests, that buy stop was triggered on only 37.1% of daily setups. In 62.6% of cases, price broke the bottom first and the setup was canceled before any entry; the small remainder touched neither level within the 20 bars our test allows. That is the main practical feature of this setup: because the first line is tall, the entry level is far away, and the decline resumes first in most cases.

Confirmation techniques using volume and indicators

A classic reading wants lower volume on day two than on day one, as a sign that selling pressure is drying up, and more activity on the confirmation bar. We measured one version of this: occurrences printed on volume above 1.5 times the 20-day average reached the 2:1 target 33.6% of the time, against 34.3% otherwise. The difference is negligible, and this breakdown is descriptive (not corrected for multiple testing). Technical indicators such as the RSI or a bullish divergence can serve as filters too; we have not tested them.

Trend and support filters

The classical reading gives more weight to a setup that forms on a support zone marked in advance (prior swing bottom, round number, rising moving average), or during a short decline inside a larger uptrend. We tested the trend filter: above the 200-day moving average, 35.3% of setups reached the 2:1 target with an expectancy of +0.04 R; below it, 34.0% with +0.03 R. The 200-day trend did not improve the result on its own.

Setting stop loss and profit targets

Position sizing and risk

The setup is wide (1.35 ATR on the median occurrence), so the distance from entry to exit is large. Size the position on that distance so that a full loss costs a small, fixed share of the portfolio, and never widen it once the position is open.

Common mistakes and limitations when trading the pattern

Ignoring the trend

The same two lines inside a range or an uptrend are not a homing pigeon. Without a prior decline, there is nothing to reverse.

Buying day two

Expecting a large edge

On daily charts, it does not differ measurably from chance. Treat it as a location with a defined risk, not as a forecast.

Advantages and limitations

• Advantage: clear structure, objective rules, and a natural exit point.
• Advantage: the entry and the stop both come from the pattern itself, so the risk is known before the trade.
• Limitation: rare, about one occurrence every 3,239 daily bars.
• Limitation: the entry is far away, and most setups are canceled before entry.
• Limitation: no measurable edge on daily stocks, ETFs or indices, and too few daily commodity cases to judge.

Homing pigeon reliability: results by market

Not on its own on daily charts: across 2,432 occurrences, the homing pigeon pattern was followed by a rise slightly more often than random bars on the same assets, but its average excess return over the rest of the market is within the margin of error. Crypto gave almost nothing to measure: no occurrence on daily charts, 1 on 4-hour charts and 62 on hourly charts. Here are the numbers from our test bench.

How to read the table

Read the columns in pairs. "Higher 10 bars later" is the share of occurrences where price was up 10 bars later; "random bars" is the same measure on random bars of the same assets. In the first row, 55.5% were followed by a rise, against 54.1% for random days: the typical stock rises more often than it falls, so a figure above 50% proves nothing alone. The excess return over the rest of the market after 10 days is +0.18%, with a 95% confidence interval from −0.19% to +0.60%. Zero is inside the interval: no measurable difference from the rest of the market. The excess return was close to zero in both the first half of the history (+0.26%) and the second (+0.11%), and the method is detailed in our methodology.

The "2:1 target hit" columns use the plan described above: entry above the top, exit below the bottom, target at twice the initial distance, compared with the same plan placed on random dates. Expectancy is the average result in R (multiples of the amount at stake), before fees and slippage: +0.03 R against +0.07 R on random dates in the first row, a gap our tests do not separate from noise. On the indices, commodities, hourly crypto and weekly lines, fewer than 30 trades were triggered (17, 6, 20 and 12): their 2:1 trade verdict reads Not enough data, and the gaps there are not interpreted.

The hourly crypto result

On 62 hourly occurrences across 50 crypto pairs since 2017, only 20 opened a trade. The 2:1 setup reached its target 10.0% of the time against 25.5% for the same plan on random dates (−0.70 R against −0.24 R before costs), but with fewer than 30 trades these figures are shown, not interpreted: the 2:1 trade verdict reads Not enough data. The 10-bar move is rated no clear edge: an excess return over the rest of the market of −0.25%, within the margin of error, and −0.45% once trading costs are deducted.

Forex lines

On hourly and 4-hour forex charts the homing pigeon forms almost only at the weekly open, across the weekend gap: forex trades around the clock, so the second line can only open inside the first body after the market has been closed. Those rows measure the weekend gap, not the pattern, so we do not publish them. Daily forex had too few cases to judge either measure (only 2 occurrences, and none opened a position): both its verdicts read Not enough data.

Homing pigeon case studies

The gallery in the Real examples section shows real homing pigeons picked by our bench. The three cases below are cards from that gallery, on daily charts of a stock index and two US ETFs (prices adjusted for dividends), each checked bar by bar against the strict rule above, with what happened next.

Homing pigeon pattern example 1: canceled the next session (KOSPI)

On June 10 and 11, 2015, Korea's KOSPI index printed the setup on its daily chart: a bearish line from 2,079.13 to 2,051.32, then a bearish line that opened at 2,065.07 and ended at 2,056.61, inside the first body. The two bars spanned 2,049.13 to 2,080.03. The very next session, June 12, the index fell to 2,044.87, below the bottom, so the setup was canceled before any entry. Ten sessions after the setup, the KOSPI was 1.4% higher, and it traded above 2,080.03 only from June 23, after the cancellation.

Example 2: canceled in a falling market (SPY)

On July 2 and 3, 2008, the SPY ETF on the S&P 500 printed the setup: a bearish line from 92.42 to 90.55, then a bearish line from 91.21 to 90.64, inside the first body. The two bars spanned 89.69 to 92.68. On July 7, SPY fell to 88.92, below the bottom, before it ever reached 92.68, so the setup was canceled. The decline went on to 86.13 on July 15; ten sessions after the setup, SPY was 0.3% lower than the setup's final price.

Example 3: canceled before entry (XLRE)

On March 20 and 23, 2026, the XLRE real estate ETF printed the setup: a bearish line from 40.87 to 39.64, then a bearish line from 40.23 to 39.93, inside the first body. The two bars spanned 39.46 to 40.95. XLRE never reached 40.95 before it broke 39.46 on March 25, so the setup was canceled. Ten days after the setup, XLRE was 2.7% higher: a canceled setup does not mean the market kept falling. Most daily setups end this way, as the 62.6% cancellation rate shows.

Tools: homing pigeon screener and charts

TA-Lib, the free technical analysis library, includes a CDLHOMINGPIGEON function, and stock screeners can offer a ready-made scan. Check the definition your screener uses: some ignore the prior decline, some compare full ranges instead of bodies. Before you use it with real money, run a simple backtesting exercise: mark the last 50 occurrences on your market, apply the same entry, stop and target, and compare the results with the same trade on random dates.

FAQ about the homing pigeon pattern

Summary of key points

• The homing pigeon is a two-candle bullish reversal pattern: in a downtrend, a long bearish candle, then a smaller bearish candle with its body inside the first one.
• It differs from the bullish harami only by the color of day two.
• On 2,432 daily occurrences, 55.5% were higher 10 days later, against 54.1% for random days, but the average excess return over the rest of the market is within the margin of error.
• On crypto, the bench found almost no occurrence: none on daily charts and only 20 trades on hourly charts, too few to judge.
• Most daily setups (62.6%) are canceled before entry because price breaks the low first.

Risk warning. This page is educational and is not investment advice. No candlestick pattern guarantees a result, and any position can lose the full amount at stake.

About the author and the data. Written by Amaury, founder of PatternsWizard, who is passionate about finance, quantitative trading and data analysis. He built the PatternsWizard test bench, which detects each setup with fixed, published rules on full market history: crypto, forex, US stocks, ETFs, indices and commodities. Trades and win rates are compared with random bars on the same assets, excess returns with the rest of the market, with error margins and a correction for multiple testing, and the whole pipeline is checked on random-walk prices, where nothing can be predicted. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.