Let me start with the conclusion: in this market, leverage is the thing retail traders most need to avoid. Picking the wrong coin or buying at a bad time hurts, but a spot holder can still wait. Hold 10x leveraged Bitcoin over the same stretch and the path-dependent volatility tends to liquidate the position long before the directional thesis can play out. This piece walks through the math first, then three historic blow-up events in which huge numbers of real accounts were force-closed.
§1 · What is leverage in crypto trading?
Leverage is borrowed capital used to amplify your position size. It amplifies not only your returns, but the cost of every mistake you make. Concrete example. You have $1,000 of margin. At 5x leverage, you can open a position worth $5,000. Gains scale to the $5,000 position size. Losses scale to the $5,000 position size. Money is made faster. Money is lost faster.
For context, retail stock margin in traditional finance is kept low by regulators — in the US, Regulation T caps initial margin at 2x — with strict initial and maintenance margin rules on top. The offshore crypto perpetual futures market pushed leverage much further — Binance offers up to 125x on BTC perpetuals, Bybit up to 150x (first tier only; the cap drops as the position grows), and BitMEX's perpetual also offered 100x. That is far beyond what regulators let retail traders use on stock margin, and it exists in crypto largely because the offshore venues sit outside the reach of the regulators that would normally cap it.
§2 · How far does the price have to move against you to liquidate a 5x position?
Suppose you have $1,000 and BTC is trading at $50,000. You open a 5x leveraged long: you post $1,000 as margin, and you control a position equivalent to 0.1 BTC (notional value $5,000).
BTC drops 20% to $40,000. Your 0.1 BTC is now worth $4,000. The position is sitting on a $1,000 unrealized loss — equal to your entire margin balance. The exchange's liquidation engine force-closes the position ("rekt," "liquidated," "blown out"), and your $1,000 of margin is gone. Note that the underlying asset only moved 20%. You did not get the direction wildly wrong. You got it slightly wrong, and the leverage took the rest.
In practice it happens sooner: the exchange liquidates you once your margin falls to the maintenance level, not at zero, and it also charges a liquidation fee. Don't expect anything back from that position's margin. With isolated margin the most you lose is the margin on that position; with cross margin, the limit is the margin in your futures account. If price gaps past the bankruptcy price, the shortfall is generally absorbed by the insurance fund as far as it can, and auto-deleveraging (ADL) kicks in when the fund isn't enough. More on that in Can You Owe Money After a Futures Liquidation?, and on the two margin modes in Cross or Isolated Margin?
| Leverage | Reverse move to liquidate (illustrative) | Room for error |
|---|---|---|
| 2x | 50% reverse | Time to think, time to recover |
| 3x | 33% reverse | Wide tolerance |
| 5x | 20% reverse | One decent pullback can reach it |
| 10x | 10% reverse | A big-move day can cover it |
| 20x | 5% reverse | One sharp drop is enough |
| 50x | 2% reverse | A single liquidity wick kills you |
| 100x | 1% reverse | Effectively dead before the order is filled |
The table uses the simplified rule: reverse move ≈ 1 ÷ leverage, ignoring maintenance margin and fees, so real liquidation comes earlier. BTC's intraday swings are not small, and on big-move days a move of several percent is nothing unusual. That means any leverage above 10x exposes you to liquidation on ordinary volatility — you don't need a crash. Above 50x, a single wick is enough to take the position out.
§3 · Spot vs futures — these are different products
Spot · you actually own the asset
You spend $1,000 to buy 0.02 BTC on the exchange's spot market. Those 0.02 BTC are yours. You can withdraw them to a self-custodial wallet. You can hold them for thirty years. You can do nothing. Spot's structural advantage is that there is no liquidation. BTC drops 50%? You still hold 0.02 BTC; the position is now worth $500 instead of $1,000. BTC drops 99%? You still hold 0.02 BTC. The asset has to go literally to zero for the position to disappear. As long as Bitcoin the network exists, you still own your fraction of the supply.
Futures · you own a contract, not the asset
You open a long position on 1 BTC of perpetual futures. You do not actually hold Bitcoin. What you hold is a contract: a derivative instrument whose value tracks the spot price, settled in USD or USDT. On major exchanges the other side of your long is another trader's short — the exchange matches the two of you, keeps the books, and runs liquidations. When the position closes, your account is credited or debited by the dollar difference. The futures market's structural risk is that the wrong direction wipes the position to zero in a single move. Spot can be held through a crash because the asset itself doesn't disappear. Futures cannot, because the contract terminates the moment your margin is exhausted.
The dominant futures instrument in crypto is the perpetual (or "perp") contract — invented by BitMEX in 2016, now standard on every major exchange. Perps have no expiration date; theoretically a position can be held forever. The mechanism that keeps the perp price tethered to the spot price is the funding rate: longs pay shorts (or vice versa) depending on whether the rate is positive or negative. The exchange doesn't collect it. Most contracts settle every eight hours, some more often — check the contract page. The cost works out as: funding ≈ notional value × rate × settlements per day × days. Illustration: at the common 0.01% per eight hours, a year comes to about 10.95% of notional (0.01% × 3 × 365); at 10x leverage that is roughly 109.5% of your margin. When the market runs hot, the rate can be higher still. Anyone leaning long on a perp instead of spot needs to model this drag into the trade. For how perps compare with dated contracts, see Perpetual vs Quarterly Futures.
§4 · Historic liquidation 1 · What happened to leveraged traders in the May 19, 2021 crash?
May 19, 2021. BTC collapsed in a matter of hours. The causes stacked up: tighter regulation out of China, Elon Musk's announcement days earlier that Tesla would suspend BTC payments, and large on-chain transfers that fed the panic.
In a single day, the futures market saw a chain of liquidations on a scale rarely seen at the time. A huge number of trader accounts were force-closed. Binance's perpetual order book briefly printed a deep wick, and many high-leverage longs were liquidated on it.
An unleveraged spot position had a very ugly day, but nobody could force-close it. BTC later bounced back a long way, but the liquidated positions were no longer there to catch the rebound.
§5 · Historic liquidation 2 · How did the LUNA collapse liquidate people who never traded futures?
The LUNA implosion produced a different category of leveraged blow-ups. Besides straight futures liquidations, there was another route: borrowers who had taken LUNA-collateralized loans on Terra-ecosystem money markets such as Anchor and got liquidated as the collateral collapsed. An illustrative example of how that works: at $60 per LUNA, someone posts 100 LUNA (100 × $60 = $6,000) and borrows 4,000 USDC, a collateral ratio of 6,000 ÷ 4,000 = 150%. Assume the protocol liquidates at 120%. When LUNA hits $48, the collateral is worth $4,800, and 4,800 ÷ 4,000 = 120% — liquidation starts: a liquidator repays part of the debt and takes a matching amount of LUNA at a discount. As LUNA slides toward zero, the 100 LUNA gets liquidated round after round, and whatever is left is worthless.
The end state: the borrower still has the 4,000 USDC, and the LUNA collateral is basically gone — the loss is the collateral. DeFi loans are over-collateralized and non-recourse. If the collateral falls too fast to cover the debt, the gap becomes the protocol's bad debt; the protocol has no way to come after the borrower for it.
People who had LUNA staked were in a worse spot. Unstaking takes a waiting period, so while LUNA was in free-fall they watched the position's value drop without being able to sell. Three Arrows Capital, Celsius and Voyager followed LUNA into bankruptcy within months, taking customer funds with them.
§6 · Historic liquidation 3 · August 5, 2024 — Black Monday
August 5, 2024 was Black Monday in Japanese equities: the Nikkei 225 fell about 12% in a single session as the yen carry trade unwound after the Bank of Japan raised rates. Crypto wasn't spared. BTC fell hard over a few days, with the deepest drop on August 5 itself. ETH fell harder than BTC, and the larger alts fell harder still.
That night the derivatives market saw another chain of liquidations — smaller than 5/19, but still a big one.
What 8/5 showed again: experience does not protect a leveraged position from an extreme move. Three big blowups in a few years, all driven by the same structural pattern: leverage stack + sudden volatility event + thin liquidity in the cascade window. As long as offshore venues offer 100x perpetuals to retail with no real cap, this category of event is going to keep happening on roughly the same cadence.
§7 · "I'll just use a stop loss" — why this almost never saves you
"I'll set a stop loss; if it drops 5% I'm out." In practice this works almost never. The reasons:
- Wicks blow through your stop. During liquidity events, the price can transit your stop level in milliseconds. Your stop fills at the next available price, which can be a long way from your intended exit.
- Your psychology won't cooperate. When the position is down 5%, your brain says "let me give it another candle." Five turns into ten, ten into twenty, and you eventually get liquidated at the absolute worst price. Set your stop before you enter and don't move it once the trade is on.
- Stop orders slip in thin markets. Liquidity disappears precisely when you most need it. When the order book is thin, even a disciplined stop can fill well below the trigger level.
- Discipline produces stop fatigue. Illustration: ten stops in a row, each losing 5% of the account at the time, leaves you down about 40% (1 − 0.9510 ≈ 0.40). "I cut my losses" and "I lost a lot of money cutting losses" are not mutually exclusive outcomes.
Even with "strict stop losses," you can still end up losing over the long run. Strict-stop traders die of stop fatigue. Loose-stop traders die of liquidation. The math of high-leverage crypto futures is structurally hostile to retail, and that hostility doesn't depend on your personal discipline.
§8 · So leverage is never appropriate? Not quite.
It can be appropriate — but not for someone new to the market. Before you touch leverage, the following should already be true:
- At least two years of spot trading experience through one full bull-bear cycle, with documented trades.
- A track record of being profitable on spot. If you can't make money on spot, leverage will only let you lose it faster.
- A documented trading system: entry rules, exit rules, position sizing rules, written down before the trade rather than after.
- An amount of capital allocated to leverage that you can afford to lose completely — never more than 5% of total net worth.
- Income from a source independent of trading. Don't count on leverage to cover rent.
If all five are true, these are the rules to trade by:
- Keep leverage at 2x or below. Stay away from anything above 10x.
- Single position size capped at 20% of leverage capital. Spread the book across five or more independent trades.
- Trade only with the dominant trend. Longs in confirmed uptrends, shorts in confirmed downtrends. Counter-trend trades have a worse win rate at every leverage level.
- Hard stops + acceptance that you will get wicked occasionally. The cost of an unlucky wick is the price of admission, not a system failure.
- Weekly post-mortem. Put your win rate and risk-reward ratio together and work out the expected value (the method is in the risk-reward piece). If it comes out negative, stop and rotate back to spot.
Three practical rules if you trade leverage anyway
First make sure derivatives are permitted where you live and on the venue you use — many countries restrict or ban retail crypto leverage. If they are:
- Isolated margin, never cross. If you must trade leverage, use isolated margin on every position — meaning the maximum loss is capped at the margin posted for that specific trade. Cross margin lets a single bad position eat the whole futures account balance. Check which mode is selected before placing any trade; the differences are laid out in Cross or Isolated Margin?
- Hard stop, not a mental stop. Place the stop order on the exchange the moment the entry fills. Don't rely on watching the chart and reacting. Wicks happen at 3 AM. Discord notifications are not a substitute for an exchange-side order.
- 2x or below, period. 5x sounds modest until you do the arithmetic: a 20% move against you (less, once maintenance margin is counted) ends the position. 2x leaves enough headroom for normal volatility not to kill you, however confident you feel about the trade.
§9 · My advice for beginners · don't
My advice is simple: if you're new, stay away from leveraged contracts. This piece won't teach you how to make money on leverage, because at this stage you have no need to learn it. Spot + long-term holding + zero leverage is the path you are far more likely to survive on.
On top of commissions, funding fees and the occasional bad slip, leverage has a less visible cost: it takes over your attention. You open a leveraged position, and from that moment until it closes, you cannot fully sleep, you can't really read, you can't be present at dinner with your family. Some part of your brain is constantly running the calculation: where is price now, where is liquidation, do I need to add margin. Carry that state of mind for six months and something in your life starts breaking — work, relationships, health. It is not sustainable.
The spot-and-hold approach is the opposite. You can go weeks without opening a chart and get on with work, meals and sleep. BTC up, fine. BTC down, there is no liquidation price to calculate — a single weekly candle doesn't change a long-horizon thesis. That mental state can be sustained for years.
Every "I made 10x on a perpetual" story you see on Twitter or in a Telegram group is told by the winner. The people who blew their accounts to zero do not post about it. They quietly close their account, delete the app, and stop tweeting. You see a few winners loudly broadcasting their wins; you don't see the far larger crowd silently exiting. Be especially skeptical of "leverage guru" account threads showing portfolio screenshots and "x to y" trade results. The same screenshots are easy to fake, and the wins are massively overweighted in the public record relative to the losses.
§10 · Closing thoughts
Talking someone out of leverage rarely works. If what you are thinking is "Other people lose money on leverage. I won't," stop there for a moment. Each of them eventually became a data point in someone else's statistics.
If you have read this far and still want to try, please do the following. Use an amount you are unambiguously willing to lose — under 5% of net worth — and cap leverage at 2x. Trade for at least six months, then work out from your win rate and risk-reward ratio whether your expected value is positive before deciding to continue or stop. After six months, weigh whether the cognitive cost of sitting on leveraged positions for that long was worth the marginal return.
Whatever you decide, the natural next read is my piece on calling bull vs bear — even if you never touch leverage, knowing which phase of the cycle you're in determines whether to be adding to spot or taking profits, and that's the trade that actually compounds over time.