The Drawdown Playbook

The one number: maximum drawdown

Maximum drawdown is the deepest fall an account has taken from a peak to a later low, as a percentage of the peak. It answers the question that matters before you trust anyone with money, including yourself: how bad did it get?

Return tells you where an account ended up. Drawdown tells you what it put you through on the way. A record that made 15% while falling 30% along the way is a very different thing from one that made 15% while never falling more than 5%.

• Check the period: a small drawdown over a calm few months hasn't been tested yet.
• Check how often it's measured: month-end figures can hide a deeper fall in mid-month. Daily is more honest.
• Leave out deposits and withdrawals: moving money in or out can make a drawdown look better or worse than it was.

Why losses hurt more than gains help

After a loss, you need a bigger gain just to get back to where you started. The formula is gain needed = 1 ÷ (1 − loss) − 1.

A trader who makes +60% and then loses 50% isn't up 10%. They're down 20%. That's why protecting the downside comes first: a shallow fall can be climbed out of, and a deep one can take years. Try any figure in the drawdown recovery calculator.

Leverage: how small a move ends a trade

With leverage, the exchange lends you money and closes the trade (liquidates it) once losses have used up your margin. For isolated margin, the move against you that liquidates is roughly 1 ÷ leverage − maintenance margin:

With a 0.5% maintenance margin rate, before fees.

Those moves aren't rare. From 1 January 2021 to 12 October 2026, Bitcoin fell far enough within a single day to liquidate a 10× long on 40 of 2,111 days, and Ethereum on 89 (Binance daily prices). Falls spread over several days liquidate trades too. A spot position, bought without borrowing, can't be liquidated at all. Check your own numbers in the liquidation price calculator and the leverage risk check.

Coins fall together

Holding several coins feels like spreading risk. In a crash, crypto coins tend to fall at the same time, and smaller coins usually fall further. Real falls from the day's open to its low:

Binance spot daily prices, UTC days.

So count your risk as if every position pointing the same way could go wrong on the same day. Five leveraged longs on five coins are, on a bad day, closer to one big bet than five small ones.

Size every trade for the bad day

Most blown-up accounts weren't wrong about direction. They were too big for a normal bad day. Three questions to ask before any trade:

1. If this goes wrong, how much of the whole account do I lose? Not "of the trade": of the account.
2. If everything I hold goes wrong on the same day, how much do I lose? Add up chapter 4's worst case.
3. Could I recover from that? Check the chapter 2 table. Above a certain depth, recovery stops being realistic.

A common textbook approach is to decide the most you'll lose on any one idea, say 1% of the account, and size the trade from that. With a $9,000 account and a planned exit 9% below the entry, a 1% risk ($90) means a position of about $1,000. The point isn't the exact figures. It's that the size comes from the loss you can accept, not from how confident you feel.

An illustration of a common method, not a recommendation or our own sizing.

Write your own drawdown brake

Willpower fails exactly when it's needed: after a run of losses, when the urge to win it back is strongest. A drawdown brake is a rule, written down in advance, that cuts risk automatically as the account falls from its peak.

• Measure from the peak: the account's highest value, with deposits and withdrawals left out.
• Use steps, not one cliff: for example, a warning line, a line where you slow down, and a line where you stop opening new trades until the account recovers. What each step does is up to you.
• Put every line well before the point you couldn't recover from: chapter 2 tells you where that is for you.
• Decide what happens to open trades: whether they keep their planned exits or close.
• Make it automatic if you can: alerts, or software, so the rule doesn't depend on your mood on the worst day.

CriptoFocus runs a three-step brake that slows, then stops, new trades as the account falls, all before Bybit Pro's 20% limit. Its exact levels are private. Yours should fit your own account and nerves, so choose them with the plan below.

Protect the account itself

• Turn on two-factor authentication with an authenticator app, not SMS, on every exchange account and on your email.
• Give API keys the least they need: read-only for trackers and dashboards. Never allow withdrawals on a key you give to an app. Restrict keys to IP addresses where the exchange allows it.
• Don't keep everything on an exchange: long-term holdings are usually safer in a wallet you control (a hardware wallet), with the recovery phrase written down offline and never typed into a website.
• Check the whole address every time: malware can swap a copied address, and scammers send tiny payments from lookalike addresses that share the first and last characters. Check the whole address, middle included, never copy one from your transaction history, and send a small test amount first to any new address.
• No one legitimate will ask for your recovery phrase or password, including "support" staff who message you first.

One-page risk plan

Fill this in before your next trade, then print it or keep it beside your screen.