Dead Cat Bounce: Definition, Causes and How to Trade It

A dead cat bounce is a brief recovery in the price of a stock, an index or an entire market after a sharp drop, followed by a resumption of the slide to new lows. The name comes from an old Wall Street saying, the idea that even a dead cat will bounce if it falls from a great height.

This page gives a precise definition, the history of the term, the causes, a step-by-step way to identify one, how it differs from a true turnaround, real examples and a risk-first approach for investors. Our bench applies one fixed rule built on the confirmation that defines the pattern, a new low, to full price history and measures what happens after it, against random dates on the same assets.

What is a dead cat bounce?

A dead cat bounce is a temporary upswing inside a larger downtrend. Prices drop hard, win back part of the loss for a few days or weeks, then roll over and break under the prior bottom. Because it fails, the move is also known as a sucker rally: investors who bought it as the start of a new uptrend end up holding losses.

In technical analysis, this phenomenon is a continuation of the trend, not a turnaround. It describes what happened, not what will happen: while the upswing is under way, nobody can know whether it will fail or become the first leg of a lasting uptrend. It is only confirmed once price makes a new low.

• Before: a sharp sell-off, sometimes driven by news.
• During: a short-lived upswing, smaller than the drop.
• After: the slide resumes and price sinks under the bottom that came before the upswing.

The history and origin of the term

The financial use of the phrase goes back to the Financial Times in December 1985, which used it to describe a brief upswing in the stock markets of Singapore and Malaysia after a sharp sell-off, before they dropped again. The report, by correspondents Chris Sherwell and Wong Sulong, quoted a broker describing the markets' short-lived upswing during the regional recession of that year. The image is deliberately grim: the bounce is not a sign of life. The expression spread through Wall Street and is now an established term in financial markets, used for stocks, bonds, commodities and crypto alike.

Why do dead cat bounces happen?

A sharp drop leaves the market stretched, and several factors can push price up for a while without changing the underlying picture.

Technical factors and short covering

After a large sell-off, short sellers sit on gains. Some buy back their shares to lock in profit, and that short covering is real buying demand, even though it is not a bet on a turnaround. Bargain hunters add to it, attracted by a stock that looks cheap against its old high. Classical technical analysis also expects price to stall at obvious floors, round numbers and oversold readings on momentum tools, where many market participants act at the same time; we have not tested that. Once that demand is used up, nothing is left to hold price, and the slide continues.

Market sentiment and psychological bias

Market sentiment plays its part. Investors anchor on the old price and see the drop as an opportunity. Others who missed the exit hope for a better level to sell, and that supply caps the upswing. The causes of the sell-off, such as poor earnings, a policy shock or a credit event, are still there. When the next piece of bad news arrives, sentiment turns again and the bounce fails. These explanations come from market commentary: our bench cannot measure short covering or sentiment, only the prices that follow.

How to identify a dead cat bounce

There is no way to identify one with certainty while it happens. What you can do is check whether the upswing looks weak. Classical technical analysis looks at the following clues:

1. A sell-off came first. A large, fast drop, sometimes driven by news or a change in fundamentals.
2. The upswing is shallow. It wins back a small part of the drop, for example up to the 38.2% or 50% Fibonacci retracement, and stalls under resistance such as the old floor or a sinking moving average.
3. Buying fades. The move up comes on thinner activity than the sell-off, and up days lose momentum.
4. The cause is unchanged. The news, earnings or policy that triggered the drop has not changed.
5. Price breaks the bottom. This is the only real confirmation: a close under the bottom that came before the upswing.

None of these clues alone proves anything. Together they describe a weak upswing, but we have not tested them as a combined filter; the last one decides.

Differentiating a dead cat bounce from a true market reversal

The hardest part is telling a temporary upswing from the start of a real turnaround. Both begin the same way: prices climb from a bottom after a sell-off. The table below sums up the differences classical analysis looks for; we have not tested them as filters.

The closest classic shape we test on our bench is the bear flag: a sinking pole, then a small counter-move that wins back at most half of it before the slide continues. On 831 daily bear flags that broke downward, across all markets, 47.8% were lower 10 days after the breakout, against 47.7% for random days on the same assets: no measurable difference from chance. The opposite case, a V bottom where the recovery closes above the start of the sell-off, did no better: on 4,689 daily occurrences, 54.2% were higher 10 days later, against 54.1% for random days, with an average excess return of −0.45% over the rest of the market, rated worse than chance. Neither shape, on its own, told investors which way price would go next. The dead cat bounce itself, tested in the next section, did no better.

Is the dead cat bounce reliable? What the data says

A dead cat bounce can only be named once price makes a new low, but that confirmation is itself a rule a computer can apply. Our bench turned the clues above into one fixed reading and measured what happens after that close, on every market. The main results come first; the rule follows them.

The rule, never tuned on results:

• The sell-off: a swing from a peak to a low, at least 4 ATR deep, that falls at least 0.5 ATR per bar, the speed our bench requires of a V reversal.
• The upswing: it wins back at most 50% of the drop, the limit our bench sets for a textbook flag, and lasts at most 3 times as long.
• The signal: a close under the bottom that came before the upswing. The stop goes at the high of the upswing, and the target at twice the risk. A close above that high first cancels the setup.

Turning points are swings of 2, 4 or 8 times the average true range.

The table below details the short trade for the main markets and timeframes: every occurrence our rule finds on full price history, in crypto, forex, US stocks, ETFs, indices and commodities, compared with the same trade placed on random dates on the same assets.

How to read it: each row is the short trade above, compared with the same entry, stop and target placed on random bars of the same assets. Expectancy is the average result per trade in units of risk (R), before fees.

On hourly, 4-hour and daily charts, the new low told little about what came next: the average result per trade never beat chance measurably. Hourly crypto shows one of the widest gaps in direction on a large sample: 40.2% of signals were lower 10 bars later, against 50.7% of random bars. Yet the excess return over the rest of the market is about zero (+0.01%): other coins moved the same way, so the gap says nothing about the pattern. On 4-hour crypto charts, the average excess return was −0.46%, slightly against the short, a gap that passes a simple test but not our correction for multiple testing. On daily ETFs, the short did worse than on random dates (−0.46 R against −0.12 R), again passing a simple test but not our correction; on daily indices, the gap (−0.38 R against −0.14 R) is within the margin of error.

On weekly charts, price was lower 10 weeks after the signal in 39.3% of cases, against 46.3% for random weeks, but the 95% interval of the average excess return over the rest of the market includes zero. The weekly short trade showed no clear edge, all markets combined; on US stocks alone, the short reached its 2:1 target more often than on random dates (20.1% against 13.4%), and lost a little less per trade: −0.30 R against −0.39 R on random dates, a gap that is not statistically significant, so the row reads No clear edge. Stock data also misses most delisted companies, a survivorship bias we have not sized. Not every weekly US stock setup became a trade: when the 2R target would have sat at or below zero, the bench opens no short, and a few shorts were still open when the data ended. That leaves 319 closed trades out of 424 setups. On weekly crypto charts, the upswing is often so tall that a target at twice the risk would sit below zero, and the bench opens no trade in that case: not one of the 47 setups gave a closed short, so the trade row reads Not enough data, and the 10-week move of those 47 occurrences is shown without a verdict (Not enough data).

Price reached the measured move, the height of the upswing projected down from the bottom, before the stop and within 100 bars in 53.7% of daily setups; that rate has no random baseline, so it describes the shape rather than an edge.

The dead cat bounce can share its turning points with the bear flag, which has its own rule and its own page; the shapes overlap, so their figures should not be added together. What this means for investors: a close under the prior bottom is a description of what happened, not a signal that the slide will go on.

Real-world examples of the pattern

History offers well-known examples of a dead cat bounce. They are easy to spot afterwards, difficult to call at the time. The gallery adds real dead cat bounces detected by our bench, with their actual outcome. Trade results are in R, the initial risk of the trade.

• The dot-com bust, 2000 to 2002. After the Nasdaq peaked in March 2000, technology stocks went through several sharp rallies that each failed and led to new lows.
• The 2008 financial crisis. US stock indexes climbed several times in late 2008 before the final bottom of March 2009.
• Bitcoin in 2018. During the crypto bear market after the December 2017 peak, Bitcoin had repeated upswings that each ended under the prior high.

A counterexample is just as useful. In March 2020, after the Covid crash, the first upswing in the stock market looked like a textbook candidate for a dead cat bounce. It turned into a lasting recovery instead. That case shows the limit of the concept.

Trading a dead cat bounce: a step-by-step approach

Some investors use the concept to bet against a failing upswing, others simply to avoid buying too early. Either way, the strategy is the same: let the market prove the bounce has failed before you act.

Our bench tests the conservative entry: a short at the close under the prior bottom, a stop above the high of the upswing and a target at twice the risk, the same 2:1 rule it uses for every chart pattern so that results can be compared.

Managing risk and setting stop losses

Frequently asked questions about dead cat bounces

Summary of the main points

• A dead cat bounce is a brief recovery after a sharp drop, followed by a new low.
• It is a continuation of the downtrend that can only be confirmed in hindsight.
• Market commentary explains the upswing by short covering, bargain hunting and hope, and its failure by unchanged causes; our bench does not measure these causes.
• Tested with one fixed rule on 4,614 daily cases, it did not beat random days, and neither did the two shapes we compare it with, the bear flag and the V bottom, its opposite.
• Act on the break of the bottom, put the stop above the upswing high and keep the risk small.

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 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. The dead cat bounce rule on this page was set from chart reviews, never tuned on results. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.