Risk

Catching a Falling Knife: Why the Stock Kept Falling After You Bought the Bottom

Level: Beginner · 8 min read · Updated November 6, 2026

If you thought you bought the bottom and the stock kept falling, the usual reason is simple: the size of a drop says nothing about where it ends. A stock down 50% can fall another 50%, which makes −75% from the high and needs +300% to get back.

This lesson covers why falling stocks keep falling, the “it’s cheap now” illusion, the arithmetic of a second fall, how to tell a falling stock from a base, and three rules for buying after a drop without predicting the bottom: wait for a higher low or a confirmed reversal, write the stop before the buy, and use a size that survives being wrong. Every example uses XYZ, a hypothetical stock. This is general education, not investment advice, and trading carries a real risk of loss.

A line for a hypothetical stock XYZ falling from an old high of 100.00 through 50.00 to 25.00. A note near the old high says “it was 100, so 50 is cheap”. A circle at 50.00 is labeled “I bought the bottom”, and the line goes on falling to 25.00. A card on the right gives the computed falls: high to my buy −50%, my buy to the low −50%, high to the low −75%, and the gain needed from the low to get back to the high, +300%.

Why does a stock keep falling after it has already dropped a lot?

A fall does not use anything up. No rule says a stock has fallen “enough”: price stops falling when buyers at that price outweigh sellers, and the distance from the old high plays no part in that.

Several kinds of seller can stay active long after a drop looks extreme. Holders who bought higher sell into each bounce to get out closer to breakeven. Funds with large positions can take weeks to reduce them. Traders on margin can be forced out. And if the fall came from real news, such as weaker earnings, the price may still be adjusting to a smaller business. Buying the first big red day means standing in front of all of them.

Buying a stock because it has fallen is called bottom fishing; doing it during a fast drop is catching a falling knife. The rises on the way down are covered in dead cat bounce, and buying more as it falls in averaging down. This lesson is about the decision before the first buy.

Bottom fishing: Buying a stock that has fallen a lot in the hope of getting close to its lowest price. Nobody can see that price until well after it has happened.
Falling knife: A stock in a fast, ongoing fall. Trying to catch one means buying before anything on the chart shows that the fall has ended.

Why does a stock look cheap after a big drop? The anchoring illusion

The eye measures the new price against the old high: “it was $100.00, so $50.00 is half price.” That is anchoring, where the first number you saw becomes your ruler, as the cognitive biases lesson explains. But the old high is only the price someone paid in a different situation. Cheap has to mean cheap compared with something the business does, such as what it earns.

Hypothetical: at $100.00, XYZ earns $5.00 a share, a P/E ratio of 20. Then expected earnings are cut to $1.25. At $50.00 the stock looks half price, yet its P/E is 50 ÷ 1.25 = 40, twice what it was at the high. At $25.00 the P/E is back to 20: a 75% fall, and not a cent cheaper on this measure. That is the pattern behind a value trap. Earnings are not the only measure and estimates can be wrong in either direction; the point is that the old high by itself tells you nothing about value.

Three cards for hypothetical XYZ. At the high, labeled the anchor: price 100.00, earnings per share 5.00, P/E 20.0. First fall, labeled “half price!”: price 50.00, earnings per share 1.25, P/E 40.0, drawn in red. Second fall, labeled “a quarter of the price!”: price 25.00, earnings per share 1.25, P/E 20.0. A note says that at half the price but a quarter of the earnings, the stock costs more per dollar of earnings than at its high.

What happens if a stock falls 50% and then another 50%?

Percentages compound from the new, lower price. From $100.00 to $50.00 is −50%. From $50.00 to $25.00 is −50% again, but the total is not −100%: (1 − 0.50) × (1 − 0.50) = 0.25, so a quarter is left, a fall of 75%.

The way back is steeper, because the gain starts from the smaller number. From $25.00 to $100.00 is $75.00 on top of $25.00, or +300%. A buyer at $50.00 needs +100% just to get back to the purchase price: 100 shares bought for $5,000 are worth $2,500 at the low. The general formula is gain needed = 1 ÷ (1 − loss) − 1, and maximum drawdown applies it to a whole account.

Five rows comparing a loss with the gain needed to get back, on the same scale for every row: −10% needs +11.1%, −25% needs +33.3%, −50% needs +100.0%, −75% needs +300%, highlighted, and −90% needs +900%. Below, the sum for two falls of 50%: (1 − 0.50) × (1 − 0.50) = 0.25, and 1 ÷ 0.25 − 1 = +300%.

How can you tell a falling stock from a base?

You cannot see the bottom when it happens; the lowest price only becomes visible afterwards. What you can read is the structure of the swings, as market structure explains. A falling stock keeps printing lower highs and lower lows: each rise stops below the last one, and each fall goes further.

A base looks different. The stock stops making new lows and moves sideways. Hypothetical: XYZ falls to $25.00, rises to $30.00, pulls back only to $27.50, a higher low, and then closes above $30.00, the last high. This does not prove that $25.00 was the final low. It shows that sellers failed to reach a new low and that buyers got through the last high. That evidence has a price: an entry at $30.00 is 20% above the low. You give that up on purpose, in exchange for a place to put a stop.

Still falling: lower highs and lower lows, closes near the low of the day, rises that fail under the last high.
Forming a base: no new low, a narrower range, a higher low, then a close above the last high, better still on rising volume.
Bases fail too: trend reversal and stage analysis show how to judge what comes next, and the stop handles the rest.

Base: A sideways range that forms after a decline, in which price stops making new lows. It shows that selling has eased, not that a rise must follow.

Two panels. Left, still falling: a zigzag in which each high (orange dots, labeled lower high) and each low (red dots, labeled lower low) sits below the one before; a note says each rise stops lower and each fall goes lower. Right, a base: a fall to a low of 25.00, a rise to a last high at 30.00, a higher low at 27.50 marked in green, and a move above 30.00, with a dashed stop line at 27.00. A note says the low holds, a higher low forms, then a close above the last high. At the bottom: no one calls the bottom; you wait for evidence, and a base can still fail.

How do you buy after a big drop without catching a falling knife?

Three rules replace the guess, and their order matters:

Wait for evidence: a higher low or a confirmed reversal, not a low price. Until then there is no trade, however cheap the stock looks.
Write the stop before the buy. It goes where the idea is proven wrong, here below the higher low. If you cannot name that price, you do not have a trade yet; see where to put a stop loss.
Use a size that survives being wrong. Shares come from the stop, not from how cheap the stock seems: shares = money at risk ÷ (entry − stop), as in the position size calculator.

Falling knife or planned entry: what is the difference in dollars?

Hypothetical, with a $10,000 account. The first buyer takes 100 shares at $50.00 with no stop, because it is “half price”. That is $5,000, half the account, and at $25.00 the position has lost $2,500, or 25% of the account.

The second buyer waits for the base and risks 1% of the account, $100. Entry $30.00, stop $27.00, so $3.00 a share and $100 ÷ $3.00 = 33 shares, a $990 position. If the base fails, the loss is 33 × $3.00 = $99, or 0.99% of the account. This buyer can still be wrong, and can be wrong several times in a row before a base holds, as losing streaks show. The small size is what pays for those attempts.

Some tools follow the same order of events. Our invite-only UpTickFlow indicator draws harmonic buy zones before price arrives, then waits for a reversal candle with RSI confirmation and shows a stop and targets. A zone is a place to watch, not a call on the bottom.

Two cards on a hypothetical $10,000 account. Buying the fall, no stop: entry 50.00, 100 shares, value lost at 25.00 −$2,500, and a bar one quarter red, −25.00% of the account. Waiting for the base, stop first: entry 30.00, stop 27.00, 33 shares, loss if wrong −$99, and a bar with a thin red sliver, −0.99% of the account.

I already bought and it keeps falling: what should I do?

Start with the question the anchor hides: if you held no shares, would you buy this stock here, today, with a stop you can name? If not, the only reason left to hold is the price you paid, and the market does not know that price. This is general guidance, not advice on your position.

If you set a stop, keep it. If you never set one, decide now at which price the idea is wrong, and check that the loss at that price is one you can take.
Do not add to lower your average unless that was a written plan before the first buy; averaging down shows the math.
Separate the trade from the investment. A long-term holder whose reasons still stand decides differently from a trader whose setup has failed; when to sell a stock and lost money in stocks walk through both cases.
If the position keeps you awake, it is too big. Cutting part of it is a decision you are allowed to make without being sure of the next move.

Test yourself

4 quick questions on this lesson. Pick an answer to see why.

Question 1 of 4
Each chart shows hypothetical XYZ after a long fall. Which one shows a higher low followed by a rise above the last high?
Turn your phone sideways to read the image. Tap to close.
Question 2 of 4
Hypothetical XYZ falls 50%, then another 50%. What gain does it need from the low to get back to its old high?
Question 3 of 4
At $100.00, hypothetical XYZ earned $5.00 a share. It now trades at $50.00 and expected earnings are $1.25 a share. What happened to its P/E?
Question 4 of 4
You risk $100 per trade. Hypothetical XYZ has formed a base with an entry at $30.00 and a stop at $27.00. How many shares fit that risk?

Questions

What does catching a falling knife mean?

It means buying a stock during a fast fall because it looks cheap, before anything on the chart shows that the fall has ended. The risk is that the fall continues and the loss grows with no planned exit.

Why did the stock keep falling after I bought the bottom?

Because you could not know it was the bottom when you bought. Holders selling into rises, funds reducing positions, forced selling and a lower view of the business can all continue after a big drop. The price you bought at only looks like the bottom until a lower one prints.

How do I know when a stock has hit bottom?

Not in real time. You can only see evidence after the fact: no new low, a higher low, then a close above the last high. That evidence can still fail, which is why the stop is written first.

Should I buy a stock because it dropped 50%?

The drop alone is not a reason. Ask what the business earns at this price, whether the structure has stopped falling, and where your stop goes. This is general education, not a recommendation on any stock.

How much does a stock need to rise after falling 75%?

It needs +300%: 1 ÷ (1 − 0.75) − 1 = 3. From $25.00 back to $100.00 is $75.00 on top of $25.00.

UpTickFlow is our product. These guides are education, not financial advice. Trading involves a substantial risk of loss.