What is a trailing stop?
Most explanations stop at “it locks in profit as price rises.” That is true and it is the least useful half. The half that decides whether you make money is arithmetic: the moment you type a callback rate into the box, you have set a ceiling on your own exit price, and you can calculate exactly where it is.
What does a trailing stop actually do?
It is a ratchet. As price makes new highs, the trigger price is recalculated upward and stays there; when price falls back, the trigger does not follow it down. That one-way property is the whole mechanism, and everything else on this page follows from it.
A plain stop-loss sits where you put it until you move it by hand. A trailing stop moves itself, using a rule you fixed in advance. That is the trade you are making: you give up judgement about where to exit, and in return you get an exit that works while you are asleep.
What do “callback rate” and “activation price” mean?
Two fields do all the work, and exchanges name them differently even within the same exchange. On Binance futures the retreat you tolerate is the Callback Rate; on Binance spot the identical idea is called Trailing Delta. The Activation Price is the level price must touch before any trailing begins at all.
The formula behind the field is worth seeing once. For an exit from a long position, the trailing price is the highest price reached since activation, multiplied by one minus the callback. The exchange then compares the retreat from that high — the rebound rate — against your callback, and fires when it is met.
| Binance spot | Binance USDⓈ-M futures | |
|---|---|---|
| Name of the field | Trailing Delta | Callback Rate |
| Stated range | 0.1% – 20.0% | 0.01% – 20% on one help page, 0.1% – 10.0% on another |
| How it fills | Limit or market, your choice | Always market |
| Trigger reference | Last price | Last price or mark price |
Field names and ranges read from Binance's own help pages on 6 September 2026 (futures page updated 30 July 2026, spot page 9 April 2026). The two pages disagree about the futures range, and Binance's spot page itself advises checking “the actual limit range on the trading page” — so treat the number in the box in front of you as the authority, not any article, including this one.
The choice between limit and market matters more than it looks. A limit exit can simply fail to fill and leave you holding the position; a market exit always fills, but at whatever the book offers — and it triggers at the exact moment price is moving against you, which is when the spread is widest and slippage is worst. Neither is free. See crypto order types for the rest of the ticket.
The quiet failure is the activation price. Set it too far away and price never reaches it, so nothing ever trails and no order exists — but the entry still sits in your open orders list looking like protection. Leaving it blank makes it the current market price, which starts the trail immediately.
What does the callback rate guarantee you will give back?
This is the part that is arithmetic rather than opinion. With a callback of c, your exit can never be higher than the peak multiplied by (1 − c). Not usually, not on average — never. The trigger is defined as a retreat from the high, so the high must be given up before anything happens.
| Callback | Peak reached | Highest possible exit | Handed back, by definition |
|---|---|---|---|
| 2% | $130,000 | $127,400 | $2,600 |
| 5% | $130,000 | $123,500 | $6,500 |
| 10% | $130,000 | $117,000 | $13,000 |
Read the last column as a fee, because that is how it behaves. A hard take-profit at $128,000 pays $128,000 if it is hit. A 5% trailing stop on the same trade cannot pay more than $123,500 no matter how well the trend runs. What you buy for that $6,500 is the chance that the trend runs far past $128,000 — which the fixed target would have cut short.
How do you choose the number?
Here is the uncomfortable answer: the callback that works is the one wider than the deepest pullback the trend has still to produce. That quantity does not exist yet when you type the number in. You are not expressing a risk preference; you are making a forecast about the shape of a move that has not happened.
To see what that costs, take a trend running from $100,000 to $130,000 with ordinary pullbacks along the way — the path drawn above, whose deepest dip is 4.31%. Running every callback against it:
| Callback | Exits at | Profit per 1 BTC | Share of the $30,000 move |
|---|---|---|---|
| 1% | $106,920 | $6,920 | 23.1% |
| 2% | $105,840 | $5,840 | 19.5% |
| 3% | $112,520 | $12,520 | 41.7% |
| 4% | $111,360 | $11,360 | 37.9% |
| 5% | $123,493 | $23,493 | 78.3% |
| 10% | $116,995 | $16,995 | 56.6% |
| 15% | $112,000 | $12,000 | 40.0% |
Two things in that table are worth more than the headline. First, the result is not monotonic — 2% does worse than 1%, and 4% does worse than 3%, because each callback gets caught by a different pullback. Tightening the stop does not reliably keep more of the move. Second, there is a cliff at 4.31%: every callback below the deepest pullback is stopped out mid-trend, and the first one above it captures nearly four times as much. Nothing about the trade changed. Only whether your number happened to clear a bar you could not see.
Is a fixed percentage better than trailing on structure?
The alternative is to move the stop by hand to sit under levels the market has actually made — the swing lows of a trend rather than a percentage of the high. To compare them fairly, run both against two trends that begin and end at the same prices and differ only in how deep the middle pullback goes.
The 5% callback earns $23,493 on one path and $12,097 on the other — an $11,396 swing between two trends a chart-reader would describe the same way. Trailing on structure earns $20,500 and $21,000: lower than the best case, far above the worst, and it never had to guess a number. Averaged over the two, structure comes out ahead ($20,750 against $17,795) with a twentieth of the variability.
That is the honest shape of the comparison, and it is not a clean win. On the first path the fixed 5% beat structure outright. What structure buys is not a higher ceiling but a narrower range of outcomes, because its exit level is derived from the trend in front of it instead of from a number chosen before the trend existed.
Structure has a cost of its own: it needs you present, and it needs a rule for when a level counts. The usual discipline is to wait until a pullback has completed and price has turned back up before moving anything — while price is still falling there is no confirmed low to sit under, and moving early is a guess wearing the costume of risk management. Choosing which chart to trail on matters just as much; the stop-loss entry works through what trailing on too small a timeframe does to a position's room, and Lesson 15 covers reading the swing sequence that gives you the levels.
When is a trailing stop the wrong tool?
In a range. A trailing stop assumes a trend to follow. Inside a sideways band, price sets a marginal high and retreats by more than any sensible callback several times a week, so the order fires again and again with nothing to show for it. Establish whether you are in a trend or a range first — Lesson 20 is about exactly that decision.
In thin markets. A futures trailing stop always exits at market. On an illiquid pair the book below you is shallow, and the fill can land well under the trailing price you were looking at.
When your order slots are full. Binance caps conditional orders at ten per symbol on USDⓈ-M futures, and a trailing stop counts toward that limit alongside your stop-loss and take-profit orders. Binance's own documentation describes the case where a take-profit is accepted and the stop-loss is rejected for hitting the cap — leaving the profit target in place and the protection missing.
When the trade thesis, not the price, has changed. A trailing stop only ever reacts to price. If the reason you took the position has gone, waiting for a percentage retreat is a slower and more expensive way to reach a decision you have already made.
Common mistakes
- Treating the callback as a risk setting. It does not shrink the position, so it does not change what you can lose if price gaps against you. It only governs how a winner ends.
- Reading a tight callback as “safer”. Tighter means stopped out sooner, which usually means a smaller profit, not less risk — see the non-monotonic table above.
- Setting an activation price you never reach. The most common way to hold an order that does nothing while believing you are protected.
- Attaching one to an entry, then forgetting it exists. A trailing stop placed at entry starts trailing from a price with no profit in it, and can close the trade for a loss on the first ordinary wobble.
- Assuming the fill equals the trailing price. On futures it is a market order, submitted at the least liquid moment of the move.
- Using one because it feels like discipline. Automation replaces one judgement with another — the number you chose in advance. It does not remove the judgement.
FAQ
What is a good callback rate for crypto? There is no correct number, and Binance's own documentation says so. A callback only survives the trend if it is wider than the deepest pullback still to come — a quantity you cannot know when you type it in. Too tight and ordinary noise closes the trade; too wide and the guaranteed give-back swallows the profit. Measure the pullbacks the current trend has actually produced and treat that as a floor, not a forecast.
Is a trailing stop better than a normal stop-loss? They do different jobs. A stop-loss caps the loss on a position you may be wrong about; a trailing stop converts an open profit into a smaller but locked one. A trailing stop does not reduce your position size, so it does not reduce the risk you took on entry — it only decides how the trade ends.
Why did my trailing stop never trigger? Most often the activation price was never reached, so the order sat dormant and nothing trailed at all. Both conditions must be met: price must touch the activation price, and the retreat from the high must reach the callback rate. An activation price set too far away is the quiet failure — you believe you are protected and you are not.
Does a trailing stop protect me in a crash? Only partially. On futures it fills as a market order, so it executes at whatever price is available once triggered — and it triggers precisely when price is moving against you and the spread is widest. In a fast move the fill can be well below the trailing price you saw.