The falling wedge: reading it, and why it fails

This article explains the falling wedge chart pattern and the evidence on how it behaves. It is educational and general, not personalised investment advice. The entry, stop and target methods described are conventions found in the technical-analysis literature, shown so readers understand them, not instructions to act on, and nothing here is a recommendation to buy, sell, or hold any security.

The falling wedge is one of the most frequently cited “bullish” chart patterns on the internet, and one of the more modest performers in the largest public catalogue of chart patterns. Both things are true at once. Here is how to read the shape, what the measured data actually says, why so many wedges fail, and why we score contraction with rules instead of drawing wedges by hand.

The short version. A falling wedge is a price pattern bounded by two down-sloping trendlines that converge, with the upper line (the lower highs) falling faster than the lower line (the lower lows). It is usually read as bullish: in Thomas Bulkowski’s bull-market sample of over 800 wedges, price broke out upward 68% of the time. But he also ranks it a poor performer among bullish patterns, with a 26% break-even failure rate on upward breakouts.

What is a falling wedge pattern?

A falling wedge is a narrowing, downward-sloping range: price makes lower highs and lower lows, but the highs fall faster than the lows, so the two trendlines converge toward a point. The shape describes selling pressure that is still present but losing force, since each push lower travels less distance than the drop in the highs would suggest.

The geometry is specific, and it is worth getting right because it is what separates a wedge from its neighbours:

• Both lines slope down. The upper (resistance) line joins the reaction highs; the lower (support) line joins the reaction lows. StockCharts’ ChartSchool describes the upper line as having the steeper negative slope, so the lows become comparatively shallower as the pattern matures.
• The lines converge. If they run parallel, the shape is a descending channel (or, if short and after a sharp advance, a bull flag), not a wedge.
• Enough touches to be a pattern. Bulkowski’s identification guideline asks for at least five trendline touches in total, three on one line and two on the other.
• Enough time to be a wedge. Bulkowski sets a minimum of three weeks; shorter converging shapes he classes as pennants.

Is a falling wedge bullish?

Mostly, yes, in the sense that it breaks upward more often than downward. It is not reliably bullish in the sense of producing strong, dependable gains. Bulkowski’s statistics (bull market, updated August 2020 and still the figures on his site as of October 2026) show upward breakouts 68% of the time, which leaves roughly one wedge in three breaking down instead.

The pattern earns its bullish reputation from what the shape implies: sellers are still in control of direction, yet each new low is harder won, and the range is tightening as volume fades. That is a story about supply drying up. Whether the story pays is an empirical question, and the numbers below answer it more soberly than most pattern guides do.

Reversal or continuation: where the wedge appears

A falling wedge can mark either a reversal or a continuation, and the prior trend decides which. Bulkowski’s guideline for the prior trend is simply “any direction”, and ChartSchool regards falling wedges as bullish in both contexts.

• As a reversal, the wedge forms at the end of a decline. ChartSchool’s description of the reversal version calls for an extended downtrend in front of it (at least three months old) and a wedge that develops over roughly three to six months. Here the pattern is read as the decline losing momentum.
• As a continuation, the wedge forms as a pullback inside an uptrend, sloping against the trend before the trend resumes. In this context it behaves like a larger, looser cousin of the bull flag.

The continuation reading is the one that fits a momentum framework best, because it puts the pattern in the direction of a trend that already exists. A reversal wedge asks the trader to call the end of a downtrend, which is a harder and less reliable bet.

Volume: the contraction that matters

Volume usually falls as a falling wedge forms. Bulkowski reports a downward volume trend in 72% to 75% of his samples until the breakout. Fading volume during the decline is consistent with sellers losing urgency, the same supply-exhaustion idea that underlies the volatility contraction pattern.

The breakout is where volume needs to reverse. ChartSchool is blunt that without volume expansion the breakout lacks conviction and is vulnerable to failure, and Bulkowski’s trading tips note that wedges with heavy breakout-day volume, an upward gap on the breakout day, and taller patterns tend to perform better. Quiet inside, loud on the exit: that is the signature.

What counts as a falling wedge breakout?

A breakout is a close above the upper trendline. Bulkowski advises waiting for a close outside the trendline rather than reacting to an intraday poke, because intraday breaches often reverse by the close. ChartSchool adds a stricter confirmation: a break above the most recent reaction high inside the wedge, not just the sloping line.

The distinction matters because a down-sloping line is easy to “break”. As the line falls, price can cross it simply by going sideways. A break of a falling line on quiet volume, without clearing a prior swing high, is weak evidence. Many traders therefore treat the trendline cross as the alert and the swing-high break as the confirmation, a trade-off we cover in anticipation versus confirmation.

Expect a retest. Bulkowski records a throwback (price returning to the breakout area after an upward breakout) in 62% of upward-breakout wedges. A throwback is normal; a throwback that closes back inside the wedge and keeps going is a failed breakout.

How do you calculate a falling wedge price target?

There are three common conventions, and they give different answers. Bulkowski’s measure rule offers two: use the highest peak in the wedge as the target, or take the wedge’s height (highest peak to lowest valley), multiply it by the share of patterns that historically met their target, and add it to the breakout price. The third, widespread in trading education, adds the full height to the breakout price.

A worked example with round, illustrative numbers. The wedge’s highest peak is $50, its lowest valley is $40, so the height is $10. Price closes above the upper trendline at $44.

The 0.62 is Bulkowski’s measured rate of upward-breakout wedges meeting their target. Put differently, even the conservative “return to the top of the wedge” goal was missed in about four upward breakouts in ten in his sample. A target is a reference point, not a forecast.

Entry and stop conventions

The literature describes a few standard placements. We list them as conventions so the vocabulary is clear, not as a method to follow.

• Aggressive entry: on the first close above the upper trendline.
• Confirmed entry: on a close above the last reaction high inside the wedge, accepting a worse price for better evidence.
• Retest entry: on a throwback to the broken trendline that holds, which only happens in some breakouts.
• Stop placement: conventionally below the most recent swing low inside the wedge, or below the lower trendline. A fixed percentage ignores how volatile the stock is; an ATR-based stop scales the distance to the stock’s own range (see volatility-aware stops).

Falling wedge vs descending triangle vs bull flag vs rising wedge

These four shapes are easy to confuse. The quickest test is the slope of each boundary line and what came before the pattern.

The falling wedge versus descending triangle question comes down to the lower line. In a triangle the lows hold a flat floor while the highs press down on it, which is why it is traditionally read as bearish. In a wedge the lows keep falling too, just more slowly than the highs. The rising wedge is the mirror image of the falling one: higher highs and higher lows, with the lows rising faster, and Bulkowski describes rising wedges, especially downward breakouts, as among the worst-performing patterns he tracks, with a 51% break-even failure rate on downward breakouts.

What the evidence says

The best public dataset on falling wedges is Bulkowski’s, and it is less flattering than the pattern’s reputation. His bull-market figures, based on over 800 patterns:

Two cautions on reading that table. First, Bulkowski’s averages are built from “perfect trades”, measured from the breakout to the ultimate high or low with no costs, so they are a ceiling on what the shape offered, not what a real trader with real exits captured. Second, a rank of 31 out of 39 for upward breakouts places the falling wedge in the bottom quarter of the bullish patterns he ranks. Bulkowski himself calls it a poor performer as far as bullish chart patterns go.

The academic evidence is thinner, and it does not cover wedges directly. Andrew Lo, Harry Mamaysky and Jiang Wang (“Foundations of Technical Analysis”, Journal of Finance, 2000) replaced hand-drawing with an automatic method based on kernel regression, applied it to US stocks from 1962 to 1996, and tested ten classic patterns: head-and-shoulders and its inverse, broadening, triangle and rectangle tops and bottoms, and double tops and bottoms. They found that several patterns did carry incremental information about the distribution of subsequent returns. The wedge was not on the list, so the paper says nothing specific about it. Its lasting contribution is methodological: it showed that patterns can only be tested honestly once a rule, not an eye, decides where they are.

Why falling wedges fail

Most failures trace back to a handful of causes, and they overlap with why breakouts fail in general (see why breakouts fail).

• The trendline was drawn to fit. Two people given the same chart will draw different lines. Move one anchor point and the “breakout” moves with it. A pattern found by fitting lines after the fact will always look cleaner than it was in real time.
• A falling line is easy to break. Price can cross a descending trendline by drifting sideways. That is not demand; it is the line moving out of the way.
• No volume on the exit. A break on fading volume is the textbook false breakout. We cover how to tell it from a healthy retest in false breakouts and shakeouts.
• Fighting the trend. A reversal wedge inside a market-wide downtrend asks the stock to turn while the tape is still falling. Many simply resume the decline.
• Hindsight in the examples. Pattern guides show the wedges that worked. The one-in-three that broke down rarely make the illustration.

How Shishin approaches wedges

We do not detect wedges, and we do not draw trendlines. The idea inside a good falling wedge, a range that is tightening while volume dries up, is real and measurable, so we measure that idea directly instead of the shape that sometimes contains it.

Each day the board ranks roughly 1,000 US stocks on eight scoring components built from price, volume and trend. Among the things those components capture are whether a stock’s range and average true range are contracting, whether volume is quieting, and whether the longer-term trend is intact. The setup-state classifier then labels where a name sits in its cycle, from a basing, tightening state through a breakout to an extended one. The same arithmetic runs on every stock every session, so two readers looking at the same data get the same answer. The broader logic is described in how breakout setups work.

This is why rules beat eyeballed patterns, in our view. A rule can be backtested on data it was not tuned on, run live without discretion, and checked afterwards. A hand-drawn wedge can be redrawn after the fact until it agrees with the outcome. Shishin’s entries fill at the market close, which also sidesteps the intraday trendline pokes that reverse before the bell, and its stops are volatility-aware rather than tied to a line someone chose. Rules are no guarantee of an edge. They are what make an edge testable. Our live, paper-traded record, with the misses shown, is on the track record, and every signal is committed to the tamper-evident log on the verification page.

Honest limits and common mistakes

• Calling every downward slope a wedge. Without converging lines, enough touches and enough time, it is just a decline.
• Confusing it with a descending triangle. If the lows are flat, it is a triangle, and the historical odds are much closer to even.
• Treating the target as a promise. In Bulkowski’s data, roughly four upward breakouts in ten fell short of even the conservative target.
• Ignoring the market. Bulkowski’s headline figures are bull-market numbers. The same shape in a falling market is a different bet.
• Trusting old statistics blindly. The pattern statistics are dated 2020, come from one researcher’s identification of patterns, and use best-case exits. They are the best public reference, not a law.

Our own approach has limits too. Scoring contraction objectively removes the drawing bias, but it does not remove false breakouts; a tight, quiet setup can still break and fail. A rules-based process makes those failures visible and countable. It does not make them disappear.

Sources & further reading

• Bulkowski, T. N. “Falling Wedges”, statistics and identification guidelines. thepatternsite.com (statistics updated 2020). Also Encyclopedia of Chart Patterns, Wiley.
• Bulkowski, T. N. “Rising Wedges” and “Descending Triangles”, thepatternsite.com/risewedge and thepatternsite.com/dt.
• StockCharts ChartSchool, “Falling Wedge”, chartschool.stockcharts.com.
• Lo, A. W., Mamaysky, H. & Wang, J. (2000). “Foundations of Technical Analysis: Computational Algorithms, Statistical Inference, and Empirical Implementation.” Journal of Finance, 55(4), 1705 to 1765. NBER working paper 7613.
• The volatility contraction pattern on measuring a tightening base rather than drawing it.
• False breakout vs shakeout and why breakouts fail for the failure modes in more depth.
• The setup-state classifier for how the system labels where a name sits in its cycle.

Shishin Research