When to take profit and when to cut losses in crypto

You bought bitcoin at an average cost of 74,903 USDT and it has risen to 149,806 USDT. How much do you sell to get back the money you put in? Half: share to sell = average cost ÷ current price = 74,903 ÷ 149,806 = 50%. If it has only reached 1.5 times your average cost, you need to sell two-thirds.

Taking profit can be worked out ahead of time like this, and the conditions for cutting a loss can be set ahead of time too. Selling is hard because at the moment you have to act, the market is pushing you around: on the way up you can't bring yourself to sell, and on the way down you can't accept the loss. So both sets of rules get written before you buy, while the market is quiet, and followed when the time comes.

Why write your take-profit and stop-loss rules before you buy

When the market is really moving, you don't judge the way you normally do. On the way up the thought is "wait a little, it could go higher"; on the way down it is "hold a little longer, it will bounce."

Behind this is a well-studied bias called loss aversion. It comes from Kahneman and Tversky's prospect theory; their 1992 paper Advances in Prospect Theory: Cumulative Representation of Uncertainty estimated that a loss weighs roughly twice as much as a gain of the same size. That is why a losing position is so hard to sell: selling makes the loss real.

Writing the rules beforehand means that, mid-move, all you have to do is follow them. Before you buy, answer three questions and write the answers down:

Why am I buying this? At what level do I start selling? What would make me admit I was wrong and sell?

If you can't answer all three, don't buy yet. Another signal worth writing into the rules is urgency: when you feel you must buy or sell right now, and that a minute's delay will be too late, stop and go back to the rules.

How much to sell when taking profit in batches: average cost ÷ current price

Nobody sells at the exact top, so take profit in batches: once the price has risen to a certain level, sell part. The size of the first batch can be calculated. To get back the money you put in, share to sell = average cost ÷ current price:

Current price relative to average cost	Share to sell to recover your stake
1.25×	80%
1.5×	66.7%
2×	50%
3×	33.3%
4×	25%

Average cost has to be the total amount invested ÷ the quantity held. If you DCA, don't substitute the price of your first buy; the method is in how to work out your DCA average cost.

What you get back from this batch is your nominal stake. The remaining position still moves with the market, and the paper profit can be given back entirely. The fee or quote spread you pay on the sale, and the platform risk of leaving coins on an exchange, are still there as well. What this step does is bring the position down to a size you can hold.

When your share goes over the cap, sell the excess

A second take-profit line has nothing to do with price forecasts and looks only at share. Say you set crypto's cap at 10% of your investable money (how to set that line is in How much should I invest in bitcoin?). After a rally, that share grows on its own even if you haven't added a cent.

Run the numbers: your investable money is now worth $120,000 in total, of which $30,000 is crypto, or 25%. To get back to 10%, keep $12,000 in crypto and sell $18,000.

You don't need a view on whether the market has topped to do this. The share is over a line you set when you were calm, and bringing it back is following your own rule. Inside your crypto money, the same applies when altcoins rise past their line; see how often to rebalance when the split drifts.

What to use for a stop-loss if not the chart: two exit conditions

Moving averages and support levels make a poor stop-loss for a beginner, who can't easily tell which line means anything, and after the price crosses it back and forth a few times you end up selling again and again at painful prices. Use two exit conditions instead:

• Does the reason you bought still hold? There was a reason when you bought: you believed in the project, or in the direction. If events have overturned that reason (the project has run into serious trouble, the thesis has been disproved), it is time to admit the mistake, whatever the price is now. Even if it hasn't fallen yet, a changed reason means you should cut back.
• Is the position costing you sleep? If a holding makes you anxious every day, gets you up at night to check prices and spills into the rest of your life, that part of the position is too heavy, whether it is up or down on paper. Cut it until you can sleep.

If neither condition has been met, the size of a drop alone is no reason to sell. Bitcoin's monthly close fell from 114,048.93 USDT in September 2025 to 58,624.71 USDT in June 2026, down 48.6% in nine months (source: Binance public market data API, BTCUSDT monthly candles, UTC). A drop like that has to be built into your position size before you buy. Deciding whether to sell after it has happened is too late.

Admitting a mistake means accepting either that your original judgment was wrong or that your position was too heavy. The two opposite errors, hanging on long after the reason has gone and selling everything at the first drop while the reason is intact, come from the same place: at the moment of selling there were no conditions written in advance to check against.

What to do during a crash: four things first

Your position is within its cap and your reason for buying is intact, yet a crash still hurts and still makes you want to act. Do four things first:

1. Make no major decisions. Accept that you are in no state to judge right now, and put off anything like selling everything, adding leverage or borrowing to buy more.
2. Close the trading app and step away from the screen. Watching the price tick second by second only makes the urge stronger.
3. Go back to the time frame of your plan. Look at how long you meant to hold. For a position you planned to hold for a year or two or longer, one week's drop counts for little on that scale.
4. Pull out the worst case you worked out before buying. If your cap was calculated for a 70% fall or even a total loss (the method is under How to work out your cap), the drop in front of you is still within what you planned for.

Once you are calm, check the exit conditions: if the reason has been overturned or the position is costing you sleep, cut back; if neither has been met, do nothing. If you are in the middle of a DCA plan, whether to pause the buys during a crash is covered in Should you pause DCA in a bear market?

Can you change the rules? Only when the market is quiet

Once the rules are written, every market move will make you want to change them: when it is time to take profit you want to wait a little longer, and when an exit condition is met you want to ride out this one wave.

A test: if the change you want happens to make the present moment feel better, such as raising the take-profit target during a rise or loosening an exit condition during a fall, don't make it yet.

Rules can be changed, but only when the market is quiet. Wait until the move has passed, then look back at whether the rule made sense, and write down the change together with the reason for it.

A sell-rules checklist you can copy

Fill in the lines below before you buy and save them in your notes:

• Reason for buying: one clear sentence.
• First take-profit batch: when the price reaches __ times my average cost, sell __% (share = average cost ÷ current price).
• Share cap: crypto is no more than __% of my investable money; above that, sell down to the line.
• Exit conditions: if ______ happens, or the position is heavy enough to affect my sleep, cut back.
• Review date: ______ (day and month). The rules are changed only on this date.