What is a bullish engulfing pattern?

Start with what a single candle is. Four numbers, the open, the high, the low and the close, compress an entire trading session. The real body, the distance from open to close, is the verdict: how much ground one side actually kept. The wicks are the record of attempts that failed, prices that traded during the session but could not survive to the close. On its own, one candle can only ever describe one session's argument.

A bullish engulfing spans two, and that is the whole of its advantage. Session one is a down session: sellers took ground and kept it into the close. Session two opens at or below where session one finished, which means the sellers begin with everything they had won still in hand and, for a moment, a little more. Then the buying starts. By the close, price is above where session one opened. Not partly back. All the way back, and past the starting line.

Say that as an auction rather than as a shape. On day one, supply overwhelmed demand for six and a quarter hours, and the closing print recorded the loss. On day two, the market opened conceding that loss, and then a bid arrived that absorbed everything the sellers could produce, took price back through the entire prior session, and closed above the level at which the selling had begun. Whoever was short into that second close is under water on the whole of the prior day's work. Whoever was buying has repriced two sessions in one.

This is why the pattern feels different from the one-bar candles even before you know any statistics about it. A doji or a hammer is a candle that suggests. It shows a fight that ended level, or a probe that was refused, and it leaves the resolution to whatever comes next. A bullish engulfing does not suggest. Inside its own two bars, the reversal of control has already happened and been recorded at a closing price. It is the difference between a witness saying the argument was getting heated and a witness saying one side conceded and left the room.

Which brings the honest limit into view immediately, because it follows from the same logic. The pattern proves a completed takeover across two sessions. It does not prove a takeover of the trend, the week or the quarter. A market can hand control over for two bars and take it straight back on the third. Everything else on this page is about the difference between those two claims, and about the three disciplines that keep the first from being mistaken for the second.

The left panel is worth dwelling on, because it contains the detail most descriptions skip. The engulfing session did not begin with buying. It began with more selling, and the day's low was made early. That is typical, and it is part of why the pattern carries weight: the buyers who reversed the session did not have an easy start. They had to absorb a further push lower and then work price back across the entire prior range before the close. A wide green body is the arithmetic result of that work, not a decoration on it.

Loose definitions are where this pattern goes wrong, so here it is precisely. Three conditions, and one deliberate exclusion.

First, session one must be a down session: it closed below where it opened. Without that, there is no ground taken and therefore nothing to take back, and the formation degenerates into two up candles in a row. Second, session two must open at or below session one's close. This is the condition most often dropped, and dropping it changes what the pattern means: if session two opens above the prior close, part of session one's decline was never reversed, only avoided. Third, session two must close at or above session one's open, which is the moment the takeover becomes complete rather than partial.

The exclusion is the one that trips people: the test is on the real bodies, not the wicks. The engulfing candle's body must cover the prior candle's body. Its high may sit below the prior high and its low may sit above the prior low, and the pattern still stands. This is not a technicality; it follows from what a body and a wick each mean. The wick is a price that was reached and rejected, an offer nobody would live with. The body is the ground each side finished the session actually holding. A reversal of control is a reversal of held ground, so it is measured where the ground is.

That last row is the hinge of this whole page. The four conditions above define the pattern completely, and satisfying all four tells you only that the shape is present. Everything that decides whether the shape is worth a second of attention, where it printed, how much larger the second body was, how many people were involved, sits entirely outside the definition. A page that gives you the four conditions and then a promise has given you a shape test and called it a method.

One practical note on the opening condition. On liquid Indian instruments, the pre-open call auction sets the opening print, and on most days it lands within a few paise of the prior close rather than a visible distance below it. So a textbook bullish engulfing on a daily chart usually looks like the second candle starting almost exactly where the first one finished, not like a dramatic gap down. The condition is satisfied; it simply does not announce itself. Chart software that draws the two bodies flush against each other is showing you a valid pattern, not a broken one.

This is the argument that belongs to this pattern and to no other common formation, so it is worth stating carefully. Single-candle patterns describe a single session's internal struggle and then hand the question to the future. Two-candle patterns can describe a relationship between sessions, and a relationship is where a change of control becomes visible.

Consider what each formation has actually established by the time it closes. A doji has established that one session opened and closed at the same level: maximum effort, zero displacement, a tie. It is a picture of indecision, and a tie resolves in either direction. A hammer has established more: within one session, price was driven down and then bought all the way back to close near the high, so a probe lower was refused. But notice what a hammer leaves standing. The previous session's decline is untouched. The hammer says today's attempt to go lower failed; it says nothing about yesterday.

A bullish engulfing says something about yesterday. It says yesterday has been undone. That is a categorically different claim, and it is why the two-bar structure is not a cosmetic difference. The one-bar patterns are evidence about an attempt. The engulfing pattern is evidence about an outcome.

Read the last column across all four rows and the pattern's honest position becomes clear. Every formation on the chart leaves something unproved. What distinguishes the engulfing pair is not that it removes the uncertainty; it is that it moves the uncertainty forward. The doji and the hammer leave you wondering whether the reversal will begin. The engulfing pair leaves you wondering whether the reversal, which has demonstrably begun, will continue. That is a better question to be holding, and it is the entire dividend of the second bar.

It also explains a practical difference in how the three are traded. With a doji or a hammer, waiting for a confirming close is close to mandatory, because without it you are acting on an attempt. With an engulfing pattern the case for entering at the close of the second bar is genuinely stronger, since the evidence the confirming candle would provide has partly been provided already. That does not make waiting wrong. It makes the choice a real trade-off between price and evidence rather than a foregone conclusion, and the sections below give you the two variables, location and magnitude, on which the trade-off actually turns.

Here is the discipline that does the most work, and the one most articles reduce to a single throwaway line about trends. The pattern claims that control changed hands. That claim can only be true if one side held control in the first place. Everything about how much a bullish engulfing is worth follows from that one sentence, and the figure below is built to make it inescapable: the identical pair, drawn at the same scale, in three different places.

After a genuine decline, the pattern is at its most informative. Every session going in belonged to the sellers, and then one session took the whole of the last one back. There is a real transfer of control to point at, and the people who were positioned for continuation are now positioned wrongly. This is the setting the textbooks are describing when they call it a reversal pattern, and it is the only setting in which that label is even coherent.

Inside a sideways range, the same pair means almost nothing, and this is where the majority of them print. A range is a standing stalemate: neither side holds the ground, so a candle that takes the ground back has taken back nothing anybody owned. Up sessions and down sessions alternate inside a range by construction, which means the raw ingredients of an engulfing pattern are constantly present. Trading every valid pattern therefore means trading mostly range noise, and the exhaustive count later on this page shows exactly how the arithmetic falls out.

In the middle of an uptrend, after a small dip, the pattern is real but the label is wrong. Buyers already had control, gave a little back over a session or two, and then took it again. That is a continuation, and reading it as a reversal signal produces the strange error of treating a bullish pattern as evidence that something bearish has ended, when nothing bearish was happening. This case is far more common than it sounds, because a healthy uptrend is a sequence of exactly these dips and recoveries.

There is a refinement worth carrying. The longer and more one-sided the decline going in, the more the pattern is worth, for the same reason a doji matters more after a long run: there is more accumulated conviction to overturn, and more participants positioned in the direction that just failed. A pattern that reverses one lazy down session inside a drifting market has reversed almost no commitment. A pattern that reverses the last session of a long, relentless slide has reversed the session in which the sellers were most confident. Both satisfy the definition identically.

The second discipline is magnitude, and it has two halves that must be read together: how much larger the engulfing body was, and how many people were involved. The definition sets no minimum for either, which is precisely why they have to be checked by hand.

Take the body first. A second body that exceeds the first by a hair satisfies the rule and describes almost nothing: buyers finished the session a few paise better than the sellers had started it. A second body two or more times the size of the one it swallows describes something else entirely. The buyers did not merely match the prior session's selling; they absorbed all of it and kept going well past the point where it began. That surplus is the actual content of the phrase "buyers overwhelmed the sellers", and it is measurable as a simple ratio: the second body divided by the first.

Volume is the credibility check on the whole story. The pattern claims a full session of selling was absorbed. Absorbing supply requires buyers, in quantity, so a decisive engulfing candle should print on volume above the recent average, and the more it exceeds it the better the claim looks. The identical candle on volume well below average is describing something much duller: not a crowd overwhelming another crowd, but a thin session in which a modest amount of buying met a modest amount of resting supply and pushed price further than the size involved really justified. If reading participation is unfamiliar territory, our guide to volume in trading covers it properly; here it is enough to treat it as the question of whether anybody was actually there.

Two further quality checks are worth building into the habit, and both are unusual enough that most explanations omit them. The first is the size of the body being engulfed. A ratio test can be satisfied trivially if the prior body was almost nothing to begin with: swallowing a near-doji is not a takeover, it is a formality, and it will pass a body-ratio filter with a spectacular number attached. The prior session has to have been a real session of selling for its reversal to mean anything. The second is where the second candle closed within its own range. A close right at the high of the engulfing session says the buying was still in force at the bell; a close well inside the range says the move was being sold into before the session even finished.

Now the uncomfortable arithmetic, which is also the most useful thing on this page. The bullish engulfing pattern is a shape test, and shape tests are satisfied far more often than markets actually turn. Any honest account has to put those two frequencies next to each other.

The figure below does it by counting rather than asserting. It draws 90 sessions of an illustrative price series, then applies the definition mechanically to every adjacent pair of bars in it and shades each pattern the rule finds. Nothing is chosen by eye and no outcome is scored. The count is simply how often the shape occurs.

One instrument at that rate produces two or three of these a quarter, which already matches or outnumbers the genuine turning points in most instruments over the same stretch. Now scale it the way a real screen works. Forty names on a watchlist, each producing a pattern every twenty-two sessions or so, generates roughly nine a week: the shape arrives on most trading days, often more than once, in every market condition. Trends that actually reverse do not arrive most days. The supply of the pattern and the supply of the event it is supposed to herald are not remotely the same size, and no amount of pattern quality closes that gap. It is closed by refusing most of them.

Which is what the cascade in the figure is showing. Each discipline is a refusal, and the refusals compound. Location alone removes the patterns that claim a takeover where there was nothing to take. The body ratio removes the ones that qualify on a technicality. Volume removes the ones that happened in an empty room. What survives is a small number of patterns that describe what the pattern is supposed to describe, and the honest way to hold that is not as a filter that improves your odds by some quoted amount, but as a filter that stops you from acting on a shape that was always going to be there.

This is also the honest answer to the question everyone actually wants answered, which is how often the pattern works. The answer is that the question is badly formed. There is no stable success rate to quote, because the number depends entirely on which definition of the pattern you use, which market and period you test, what you count as success, how long you hold, and what you charge yourself in costs. Every published figure is a report on one particular set of those choices. What is stable, and what is worth carrying, is the structure: the pattern is common, real reversals are not, and the gap between those two facts is closed by discipline rather than by the pattern.

Everything above inverts cleanly, which is a good sign that the mechanism is real rather than a story fitted to a shape. A bearish engulfing pattern is an up session followed by a down session that opens at or above the prior close and closes at or below the prior open. Where the bullish version records buyers taking back a full session of selling, the bearish version records sellers taking back a full session of buying.

Two asymmetries are worth knowing about, and both are about market behaviour rather than about the pattern. Declines tend to be faster and more volatile than advances, so bearish engulfing patterns after a long rally often print with wider ranges than their bullish counterparts, which mechanically widens the stop and therefore shrinks the position that a fixed risk budget allows. And an advance can end without any dramatic candle at all, simply by flattening into a range and drifting, so the absence of a bearish engulfing at a high is not evidence that the high is not a high. Patterns are one way a turn can be recorded, not the only way turns happen.

A pattern that cannot be turned into a plan with an exact invalidation price is entertainment. This one can, and the arithmetic is short enough to do in your head.

The stop belongs below the low of the pattern, which on almost every bullish engulfing is the low of the engulfing candle itself. The reasoning is exact rather than conventional: the reading is that the deepest probe lower was met and completely reversed. If price trades back below that low, the reversal was not completed after all, and the interpretation that justified the position has been falsified. There is nothing left to hold on to except hope, which is not a position management technique.

That stop placement has a consequence that most pages never carry through, and it is the reason the arithmetic matters. The width of the engulfing candle sets the position size. Work it in rupees on an instrument near ₹500, with every figure illustrative. A wide engulfing candle prints with a low of ₹484 and a close of ₹512. Entering at the close puts the stop distance at roughly ₹28 a share, so a trader who has decided to risk ₹5,600 on the idea can hold 200 shares and not one more. Now take a compact version of the identical setup: low ₹496, close ₹508, stop distance roughly ₹12. The same ₹5,600 risk budget now permits 466 shares. Same logic, same rupees at risk, and more than double the position, purely because the second candle was narrower.

The most dramatic engulfing candle on the screen is the one that earns the smallest position. Size from the stop, never from the account, and the drama prices itself.

The entry decision is a genuine trade-off, and the two-bar structure is what makes it one. Taking the close of the engulfing candle gets the better price and accepts the risk that the takeover stalls on the very next bar. Waiting for a following session to close above the engulfing candle's high pays a worse price for evidence that the move extended past the pattern itself, which is the same discipline a breakout demands before it is treated as real, and for the same reason: the level being cleared is what the previous participants were defending. Neither choice is wrong. Choosing one by habit, without pricing the difference, is.

Notice that only one of the five rungs is about the pattern at all. That is the correct proportion. The formation's contribution is to hand you a specific, defensible invalidation price on a chart where such prices are otherwise a matter of opinion, and that contribution is genuinely valuable. Everything after it is risk management, and risk management is what determines whether a sequence of these trades is survivable, regardless of how any individual one resolves.

Because the shape is unmistakable and the story is satisfying, this pattern attracts confident misuse. Four errors account for nearly all of it, and each one maps to a discipline covered above. None of them is a charting mistake; every one is a decision made a step too early or a step too loosely.

All four have the same root, which is treating the shape as the conclusion rather than as the beginning of a check. The shape is free. It costs a scanner nothing to find and it will find plenty. What is scarce is the combination of a shape and a context in which the shape's claim is even possible, and the discipline to decline everything else.

A page that sold you certainty would end differently, so here is the part that gets left out. Candlestick patterns have been tested carefully by academics, and the results for patterns traded in isolation are poor. The benchmark study is Marshall, Young and Rose (2006) in the Journal of Banking and Finance, which tested candlestick signals on Dow Jones Industrial Average component stocks from 1992 to 2002 against a bootstrap null, meaning randomly generated price series built to share the same statistical properties as the real ones. The strategies added no significant value. The same team then ran the harder test in Marshall, Young and Cahan (2008), in the Review of Quantitative Finance and Accounting: the 100 largest Tokyo Stock Exchange stocks from 1975 to 2004, candlesticks examined in the market where the technique originated, across three decades. No value in the full period, none in the sub-periods, none in bull markets or bear markets. Later work across other markets reads mixed at best, and in the mixed results the added filters, holding rules and cost assumptions tend to be doing much of the work.

Read that finding precisely, because it concerns a specific object: the isolated pattern, traded mechanically, every occurrence, no location test, no magnitude test, no sizing discipline. That object has no demonstrated standalone edge, and any source promising you a percentage for the bullish engulfing pattern is quoting one particular test as though it were a property of the shape. But notice what the finding is not. It is not a test of someone who declines every pattern that prints inside a range, requires a decisive body on real participation, sets the invalidation price before entering, and sizes from the stop. That framework exists because the naked pattern carries no authority.

So what is the bullish engulfing actually good for? Two things, and they are worth more than the promise it is usually sold with. First, it is the clearest single picture on a chart of control changing hands, and after a sustained decline that picture is genuine information about who is now positioned wrongly. Second, and more practically, it hands you an exact price at which your reading is proven false, which is the input risk management actually runs on and the thing most chart observations never supply. Reading it that way, context before pattern, magnitude before entry, size from the invalidation price, is exactly the order the method we teach is built around, and it is the difference between using the pattern and being used by it.

The whole of it compresses into three sentences. A bullish engulfing says that one full session of selling was reversed by one full session of buying, completely, and that is more than any single candle can tell you. It does not say that the decline is over, only that its last session has been taken back. Ask where it printed, ask how decisively, ask who was there, and let the pattern's own low tell you the price at which you were wrong.

• Steve Nison, Japanese Candlestick Charting Techniques (New York Institute of Finance, 1991). The reference text that carried candlestick grammar into Western practice. It is the source of the classical engulfing criteria used on this page: a prior trend, opposite-coloured real bodies, and a second real body that engulfs the first, measured on the bodies rather than on the high-to-low ranges.
• Marshall, Young and Rose (2006), Journal of Banking and Finance 30(8), 2303 to 2323. Candlestick technical trading strategies: Can they create value for investors? Dow Jones Industrial Average component stocks, 1992 to 2002, tested against a bootstrap null: no evidence of value in candlestick trading signals taken in isolation. ideas.repec.org
• Marshall, Young and Cahan (2008), Review of Quantitative Finance and Accounting 31(2), 191 to 207. Are candlestick technical trading strategies profitable in the Japanese equity market? The 100 largest Tokyo Stock Exchange stocks, 1975 to 2004: no abnormal returns in the full period, in any ten-year sub-period, or in bull and bear markets separately. link.springer.com
• Exchange session mechanics. A daily candle on an NSE or BSE instrument compresses the 09:15 to 15:30 continuous session. The pre-open call auction sets the opening print, which is the mechanism that allows a session to open at or below the prior close and satisfy the engulfing condition without a visible gap on the chart.
• The counts on this page. The pattern census is computed by applying the definition mechanically to a synthetic 90-session price series generated for this page, not to any real instrument. It measures how often the shape occurs and how many survive the location, magnitude and volume disciplines. It scores no outcomes and implies no success rate.