Trading concepts

Catching a Falling Knife: What It Means and What Stocks Did After a 30% Monthly Fall

The meaning of the phrase, then a count on U.S. stocks from 2000 to 2026: how 8,875 falls of 30% in a month played out over the next month, quarter and year, against ordinary days, and why crisis years dominate.

· Founder of PatternsWizard
Checked 8 min read
Catching the knife against waiting for a higher low: a price rises, then falls steeply. A red dot marks a purchase during the fall, catching the knife; the price keeps falling well below it to a low. A base forms, the next low is higher than the first, and a green dot after the higher low marks the cautious purchase, made later and at a higher price than the low.
On this page
  1. What happened after stocks fell 30% in a month
  2. Most of the rebound came from three crisis years
  3. Falling knife vs. buying the dip
  4. If you buy a falling stock anyway
  5. Frequently asked questions

Catching a falling knife means buying an asset while its price is dropping fast, in the hope of buying close to the bottom. The phrase is a warning: like a real knife, a falling price is easy to grab at the wrong moment, and nothing on the chart says where it will stop. There is no official size or speed that makes a drop a falling knife. To give the question an answer in numbers, this page uses one stated rule, a fall of 30% or more in a month, and shows what U.S. stocks did after 8,875 such falls from 2000 to 2026, next to what they did on ordinary days.

Catching the knife against waiting for a higher low

Catching the knife against waiting for a higher low: a price rises, then falls steeply. A red dot marks a purchase during the fall, catching the knife; the price keeps falling well below it to a low. A base forms, the next low is higher than the first, and a green dot after the higher low marks the cautious purchase, made later and at a higher price than the low.
Schematic, not market data.The knife catcher buys while the price is still falling and can sit through the rest of the fall; the cautious buyer waits for a higher low and pays more than the low.

What happened after stocks fell 30% in a month

The rule: an event is a close at least 30% below the close 21 sessions earlier, about one month. Only the first such day counts, and the same stock cannot give a new event for the next 63 sessions. The data are the site's daily bars of U.S. stocks, adjusted for splits and dividends, from January 3, 2000 to October 28, 2026; three stocks whose price histories have a break were left out (CHRD and TDS, whose adjusted prices jump more than fortyfold in one day, and PFBC, which has an eleven-month hole). That gives 8,875 events in 1,582 stocks, with a median fall of −32.2% at the event. The control is every trading day of the same stocks over the same years that has 21 sessions of history: it says what an ordinary day was followed by. "Down 20% or more" compares the close at the horizon with the close on the day of the event.

HorizonAfter a 30% fall (events)All days (control)
21 sessions laterMedian +2.8%; 56.0% higher; 13.5% down 20% or more (8,830)Median +1.2%; 55.8% higher; 3.1% down 20% or more
63 sessions laterMedian +9.2%; 62.0% higher; 17.5% down 20% or more (8,763)Median +3.2%; 58.9% higher; 8.3% down 20% or more
252 sessions laterMedian +30.4%; 68.2% higher; 21.7% down 20% or more (8,300)Median +11.3%; 64.7% higher; 15.8% down 20% or more

The event counts are lower at longer horizons because recent events have not had 252 sessions yet: 575 events had no one-year outcome on October 28, 2026, 112 no three-month outcome and 45 no one-month outcome. Those are open, not counted as gains or losses.

Median return after a 30% monthly fall, against all days

Median return after a 30% monthly fall, against all days. 21 sessions later: After a 30% fall +2.8%, All days +1.2%; 63 sessions later: After a 30% fall +9.2%, All days +3.2%; 252 sessions later: After a 30% fall +30.4%, All days +11.3%. Our computation; the inputs and the method are in the text.
  • After a 30% fall
  • All days
Our computation; the inputs and the method are in the text.U.S. stocks in the site's list, January 3, 2000 to October 28, 2026, 8,875 events; survivors only, so the event medians are too high.

Read on its own, the median looks like an argument for catching the knife. The other columns say what it costs. After an event, a stock was more than four times as likely as on an ordinary day to be down another 20% a month later, 13.5% against 3.1%. Within the next 63 sessions, 44.3% of the events saw the price close at least 20% below the event's close at some point, against 16.3% of ordinary days, and the median deepest close in that window was 17.3% below the event's close, against 7.3%. Only 57.0% of the events, of the 8,478 that had a full year or had already recovered, got back to the price of a month before the fall within 252 sessions; 397 had not yet recovered and were too recent to judge.

How often the price fell another 20% or more

How often the price fell another 20% or more. At some close within 63 sessions: After a 30% fall 44.3%, All days 16.3%; Down 20% or more, 21 sessions later: After a 30% fall 13.5%, All days 3.1%; Down 20% or more, 63 sessions later: After a 30% fall 17.5%, All days 8.3%; Down 20% or more, 252 sessions later: After a 30% fall 21.7%, All days 15.8%. Our computation; the inputs and the method are in the text.
  • After a 30% fall
  • All days
Our computation; the inputs and the method are in the text.Each move is measured from the close on the day of the event, or of the ordinary day for the control.

Most of the rebound came from three crisis years

Falls of 30% in a month cluster in crashes: 1,146 of the events are in 2008, 628 in 2009 and 1,467 in 2020. Those falls were followed, within a year, by sharp market recoveries, which lifts the averages. Splitting them out changes the picture:

One year after the eventEventsControl, same years
2008, 2009 and 2020 onlyMedian +59.0%; 85.2% higher; 8.9% down 20% or more (3,241)Median +22.9%; 70.0% higher; 16.7% down 20% or more
All other yearsMedian +10.6%; 57.2% higher; 29.8% down 20% or more (5,059)Median +10.3%; 64.0% higher; 15.7% down 20% or more

Outside the three crisis years, a stock bought after a 30% monthly fall did about as well as one bought on any day at the median, was less often higher a year later, 57.2% against 64.0%, and was almost twice as often down another 20% or more, 29.8% against 15.7%. In the crisis years the events beat the control too, +59.0% against +22.9% at the median, yet 53.5% of them first fell at least another 20% within 63 sessions. Catching knives paid when the whole market was about to recover, and that was not known on the day.

One year after the event: crisis years against the rest

One year after the event: crisis years against the rest. 2008, 2009, 2020: median return: After a 30% fall 59.0%, All days, same years 22.9%; Other years: median return: After a 30% fall 10.6%, All days, same years 10.3%; 2008, 2009, 2020: down 20% or more: After a 30% fall 8.9%, All days, same years 16.7%; Other years: down 20% or more: After a 30% fall 29.8%, All days, same years 15.7%. Our computation; the inputs and the method are in the text.
  • After a 30% fall
  • All days, same years
Our computation; the inputs and the method are in the text.3,241 events in the three crisis years and 5,059 in the others, with a one-year outcome.

A scaled-down rule cannot be tested usefully on the index itself: a 10% fall in 21 sessions on the S&P 500, counted with the same first-day and 63-session rules, happened only 17 times from 2000 to 2026, too few to conclude anything.

Falling knife vs. buying the dip

Buying a dip is buying a small pullback in an asset that is still rising; catching a knife is buying while the fall is still fast and the trend is down. The buying on weakness page covers the first. A sharp rebound inside a continuing decline is called a dead cat bounce, and strategies that bet on a return to an average are covered in mean reversion.

If you buy a falling stock anyway

  • Size the position for the next fall. In the data above, more than four in ten events (44.3%) fell another 20% within three months. A position that cannot survive that is too large.
  • Decide in advance what proves you wrong. A price level, or a fact about the company, written down before buying.
  • Wait for evidence that the selling has stopped, such as a higher low after the fall. That costs part of the rebound and avoids part of the further decline; the site has not tested which rule balances the two best. Some candlestick patterns try to show the same moment, such as the three stars in the south, where the selling continues but each bearish candle is smaller than the last. One setup built on that wait is the first green day pattern, the first green close after a run of red days.
  • Know why it fell. A company can fall to zero; the stock price drop to zero page explains what happens to shareholders then. A low interest coverage ratio, with operating profit that barely covers the interest bill, is one of the facts worth checking.

More trading terms are on the trading concepts page.

Every pattern's results on one sheet

The only pattern cheat sheet with tested stats: 152 patterns, what the textbook says and what really happened on 6 markets. Free PDF, sent to your inbox.

The PDF by email, plus our newsletter: new tests and results that change. Unsubscribe anytime. Privacy.

Frequently asked questions

What does the phrase catch a falling knife mean?

Buying something while its price is dropping fast, in the hope of buying at the bottom. It is used as a warning, because the fall can continue after you buy.

Is catching a falling knife a good strategy?

On U.S. stocks from 2000 to 2026, buying after a 30% monthly fall gave a higher median result than buying on any day, mostly thanks to 2008, 2009 and 2020. Outside those years the median was about the same, and further falls of 20% or more were almost twice as common.

How do you catch a falling knife stock?

The site has not tested an entry rule that does it reliably. If you try, size the position so that another 20% fall is survivable, decide in advance what would prove you wrong, and know why the stock fell.

How long does a falling knife keep falling?

The data do not give a typical duration. What they show is depth: in the median event, the lowest close of the next 63 sessions was 17.3% below the event's close, against 7.3% after an ordinary day.

What is the difference between buying the dip and catching a falling knife?

Buying the dip means buying a small pullback in a rising asset. Catching a knife means buying during a fast fall, with the trend down.

Sources

  1. Elton, Gruber and Blake, Survivor Bias and Mutual Fund Performance, Review of Financial Studies 9(4), 1996, on survivorship bias
, Founder of PatternsWizard

Passionate about finance, quantitative trading and data analysis. Amaury built the PatternsWizard test bench, which detects each pattern on full price history and measures what happened next against random entries.

PatternsWizard tests chart and candlestick patterns on full market history, against random entries. How we test patterns.

179 patterns testedRandom-walk self-test passed

All trading concepts articles