The fundamental misunderstanding most new traders bring to leverage is that "5× leverage" simply means "5× upside or 5× downside." Mathematically that is correct — but in crypto, a 5× loss is only a 20% adverse price move, and Bitcoin can cover that much ground inside a single sharp sell-off. The rest of this article is the math: one formula, four tables, one conclusion. No filler. Just the numbers that show how thin the buffer gets as leverage goes up in this asset class.
§1 · How do you calculate the liquidation price on an isolated long?
For an isolated-margin long position, the simplified liquidation price is:
In plain English: price moves against you by roughly (1/leverage), you are liquidated. The maintenance margin ratio (MMR — usually under 1% for a small BTC position on major venues, rising in tiers as the position grows) is the small adjustment that liquidates you slightly earlier than the pure math would suggest, so the exchange has time to close the position before going negative.
Worked example · 5× long BTC at $70,000
Illustration, with MMR taken as 0.5%. Plug into the formula: Liq ≈ 70,000 × (1 − 1/5 + 0.005) ≈ 70,000 × 0.805 = $56,350. Meaning: BTC dropping from $70,000 to $56,350 — a 19.5% drawdown — liquidates the position. Put another way, the entire buffer 5× leverage gives you is that 20% band — and in crypto that is not an exotic move, so 5× held over an extended period means putting the whole band on the table.
The 10× version of the same entry: Liq ≈ 70,000 × (1 − 0.1 + 0.005) = $63,350. A 9.5% drop is the entire liquidation threshold. Double the leverage and the buffer is halved — and 10% is an ordinary-sized move on BTC.
Shorts are the mirror image: price rising by 1/leverage liquidates you
An isolated-margin short works the same way in reverse. Same simplification — no fees, no funding, no maintenance-margin tiers:
Illustration: a 5× short on BTC at $70,000, MMR taken as 0.5%: Liq ≈ 70,000 × (1 + 1/5 − 0.005) = 70,000 × 1.195 = $83,650 — a 19.5% rise ends the position.
§2 · Table 1 · How big a move against you liquidates each leverage level?
| Leverage | Adverse move = liquidation | What that size of move looks like on BTC |
|---|---|---|
| 2× | ~50% | A full bear-market drawdown |
| 3× | ~33% | A major-degree correction |
| 5× | ~20% | One sharp sell-off run to completion |
| 10× | ~10% | Routine size on a violent day |
| 20× | ~5% | An ordinary trading day covers it |
| 50× | ~2% | Can show up at any point intraday |
| 100× | ~1% | The span of a single one-minute candle |
The second column is not a statistic — it falls straight out of the formula: once price moves against you by 1/leverage your margin is gone, and after the maintenance margin ratio is deducted the trigger comes a little earlier still. The third column is only a sense of scale, not a frequency. The one thing worth keeping: the higher the leverage, the thinner the buffer left for the market, until ordinary volatility is enough to reach the liquidation line. The 5% that 20× tolerates does not even count as a big move in crypto.
§3 · Table 2 · Why can a wick liquidate you even when you called the direction right?
BTC and most crypto assets routinely produce wicks — a brief intraday plunge to a low price that holds for a few seconds before snapping back to the prior range. The wicks are caused by a combination of thin order books at extreme prices, cascading liquidations on leveraged longs, and occasionally market-maker games on the spot book. A rough sense of scale, by asset:
| Asset | Intraday wick depth | Historic extreme drops |
|---|---|---|
| BTC | Shallowest of the four | Mar 12, 2020 · "Black Thursday" one-day crash |
| ETH | Deeper than BTC | Mar 12, 2020 · fell harder than BTC the same day |
| SOL | Deeper still | Nov 2022 · crashed during the FTX collapse |
| Major altcoins | Deepest | Aug 5, 2024 · fell much harder than BTC |
Practical implication: take a 10× long on BTC, and the liquidation line sits less than 10% below your entry. Even if your directional call is correct, one deep enough wick is enough to take you out. The price recovers within seconds. Your position is gone. You watch the recovery from the sideline, knowing exactly what your P&L would have been if the wick had not happened. "Correct direction, liquidated anyway" is not an edge case in leveraged crypto.
On August 5, 2024, the Bank of Japan rate hike unwound a massive global yen-carry trade. BTC fell hard over a few days, with the deepest drop on August 5 itself. Perpetual positions across crypto were liquidated in a chain reaction, and the bulk of them were high-leverage longs. BTC bounced back afterwards, but the liquidated positions were no longer there to catch it.
Cross-margin vs. isolated-margin · different liquidation surfaces
The formula above is actually the isolated-margin approximation, where each position uses only the margin you explicitly assign to it. Isolated margin liquidates earlier (less collateral buffer), but the maximum loss is bounded — only that specific position's margin is at risk; everything else in the account is untouched.
The alternative is cross-margin mode, where the exchange uses every unused balance in your account as collateral for the position. Cross-margin liquidates later (because there is more buffer available), but the mechanism deserves to be stated precisely: when a single position is force-closed, the exchange deducts that position's margin and losses from the account's total balance, and only when the total balance can no longer cover the losses on all open positions does it start liquidating the others — one liquidation does not instantly zero the account. The real danger is different: as long as there is idle balance sitting there, the position can keep drawing on the whole account as collateral, so once an extreme move punches through, the loss is no longer capped at that position's own margin and can reach the entire account.
For any trader who is not already actively managing a portfolio of correlated positions, the correct answer is isolated margin, with each position sized at no more than 5% of total account capital. Under those settings, a full liquidation costs you 5% of capital — recoverable, and you trade again next week. The pull of cross-margin is "it liquidates later." But when a cross-margin position does blow up, what goes is the whole futures account balance, not just the position. How the two modes differ, and which to pick, is covered in Cross or Isolated Margin? What Beginners Should Pick.
§4 · Table 3 · How much do you need to gain to recover from a trading loss?
Losses are non-linear in their impact on the path back to break-even. The percentage gain required to recover from a percentage loss is always larger than the loss itself, and it grows faster the deeper the loss:
| Loss | Capital remaining | Gain required to recover |
|---|---|---|
| 10% | $9,000 | +11.1% |
| 30% | $7,000 | +42.9% |
| 50% | $5,000 | +100% |
| 70% | $3,000 | +233% |
| 90% | $1,000 | +900% |
| 99% | $100 | +9,900% |
The table starts from $10,000 of capital (illustrative); gain required = loss ÷ capital remaining. An isolated-margin liquidation is functionally the end of that position's margin — the exchange also charges a liquidation fee, so don't expect anything back. If price gaps past the bankruptcy price, the shortfall is generally absorbed by the insurance fund as far as it can, with auto-deleveraging (ADL) behind it; whether you can end up owing money is covered in Can You Owe Money After a Futures Liquidation? Even at a 99% loss, going from $100 back to $10,000 requires a 100× return. Don't count on a move like that to win your capital back. The typical response to a liquidation is the worst possible one: refund the account, take the same trade with the same leverage, and get liquidated again. That is the leveraged trader's death loop.
Funding rates · the silent bleeder
Perpetual contracts settle funding at fixed intervals: most every 8 hours, some more often — check the contract page. The baseline rate is commonly 0.01% per 8h, and it drifts away from that when the market leans hard one way. Longs pay shorts when funding is positive; shorts pay longs when it flips negative. The payment moves between traders — the exchange doesn't collect it.
The cost: funding ≈ notional × rate × settlements per day × days. Illustration: at 0.01% per 8h over a 30-day hold, notional × 0.01% × 3 × 30 = 0.9% of notional. At 10× leverage the notional is ten times your margin, so that is about 9% of your margin. Over a full year it is about 10.95% of notional, or roughly 109.5% of margin at 10×. If you happen to sit on the paying side for weeks, it is a slow, daily drain that adds up. The takeaway: perp contracts are not built to be held indefinitely, and the longer you hold, the bigger this line gets. How perps compare with dated contracts on this point is in Perpetual vs Quarterly Futures.
§5 · Table 4 · Risk of ruin · 100 trades, sized differently
Take a simplified, illustrative model: each trade has 50% win probability, symmetric P&L per trade, so the expected value is zero. Compare how much the account swings over 100 trades under four position-sizing approaches:
| Strategy | Planned single-trade loss | 100-trade account swing (≈ single trade × √100) | Adverse move needed to trigger liquidation |
|---|---|---|---|
| 1% per trade · 2× leverage | 1% | ~ ±10% | ~50% |
| 5% per trade · 5× leverage | 5% | ~ ±50% | ~20% |
| 20% per trade · 10× leverage | 20% | ~ ±200% (truncated in practice by liquidation) | ~10% |
| 100% per trade · 20× leverage | 100% (one liquidation = wiped out) | The swing stops meaning anything | ~5% |
The mathematical name for this is risk of ruin. The probability of hitting zero depends on the hit-rate distribution and market regime you assume, and changes completely when the assumption changes; what can be stated is the last column: the larger the position and the higher the leverage, the smaller the adverse move needed to reach the liquidation line — and 5% or 10% moves are not unusual in crypto. Once a single trade gets taken out at the liquidation price, the expected value of the other 99 stops being your problem.
Why "a little gambling for fun" does not work in crypto specifically
In US equities, you can run 2× leverage against index volatility that is comparatively mild, while 2× tolerates a 50% adverse move — normal drawdowns stay a long way from that line. "A little leverage long-term" is genuinely a survivable strategy for an equity portfolio.
Crypto's volatility profile is completely different. BTC is far more volatile than an equity index, ETH more so, and major altcoins more again. Run the same 2× here and the open drawdown can still get deep enough to be hard to sit through — which is why 2× is the ceiling, not a starting point; add a few more turns of leverage and the liquidation line is one ordinary move away. Humans almost never stay disciplined through these episodes — they either close manually at a loss (locking in the drawdown) or stubbornly hold through liquidation (losing everything). Either way, the long-term outcome converges toward the market's average — which, after fees and funding, is negative. "Leverage as a tool" works in equities. In crypto, the volatility regime turns the same tool into a gambling instrument, because the discipline required to use it correctly is very hard to sustain at that level of volatility.
§6 · One conclusion · if you must try it, stay at 2× or below
The entire buffer 5× leverage gives you in crypto is that 20% band.
What is certain is not when you get liquidated, but that once price moves against you by 1/leverage, the liquidation triggers — that step has no luck in it.
And crypto is volatile enough that a one-directional 20% move is not unusual.
5× tolerance sits right on 20%, 10× has only 10% left, 20× only 5%.
The higher the leverage, the less room the market is given to breathe.
A more visceral comparison: a retail trader running 50× leverage versus a casino blackjack player. Each can lose at most what they put on the table for that hand — on isolated margin, the trader's loss is capped at that position's margin. The difference is speed: at 50×, a 2% move against you finishes the hand, and you can lose several of those in a single day.
By how thick the buffer is, the tiers look roughly like this:
- 1-2×: still liquidatable, just with a much thicker buffer. If you are new and must try, stop here.
- 3×: experienced traders use this for short-term setups with hard stops. Never on size.
- 5×+: already high risk. Strictly intraday — do not hold overnight.
- 10×+: the tier retail uses most, and the tier where a single move is most likely to end the account.
- 50×+: indistinguishable from slot-machine play.
For foundations, see what is leverage for the mechanic, then risk-reward ratio math for how to combine RR with leverage in a way that does not blow up. The contract rules themselves — cross vs isolated, whether a liquidation can leave you owing money, perps vs dated futures — are linked at the relevant points above.
Why "a 50% loss hurts more than a 50% gain"
The drawdown asymmetry is the single concept most leveraged traders refuse to internalize. Lose 50%, you need +100% to break even. A 100% return takes a lot longer to produce than a 50% loss does — outside of crypto, almost no asset produces +100% inside a year; within crypto, you need at least one full bull-market leg. This is the engine behind compounding decay from drawdowns, and it is why "do not die" matters far more than "make a lot" in the long-term math of trading.
§7 · "But people make money on high leverage"
Every active trader has seen at least one friend post a "$10K to $500K on 100× scalping" screenshot. Those stories are real — in high-volatility regimes, high-leverage traders genuinely produce 50× returns on individual trades. The issue is what happens next.
- In the same regime, many more traders are on the blown-up side — the tolerance band at 100× is 1%, and the winner and the wipeouts are reading the same candle.
- The trader who just made 50× mostly does not stop there. They continue trading the same way, until one liquidation returns the gains and the principal in a single event.
- The traders who cash out after one big win are rare. Most do not stop, because the same neurological pattern that drove them to use 100× in the first place has been positively reinforced by the win and is now harder to override.
The technical name for the bias is survivorship. You see the one winner posting their screenshot. You do not see the traders who blew up in the same session and never post about it. Run the same strategy long enough and you are more likely to be one of the quiet ones.
§8 · The four big cascades · 2020 / 2021 / 2022 / 2024
Worth memorizing in any conversation about leverage:
- March 12, 2020 · "Black Thursday." The first COVID lockdown wave. BTC crashed within a day, one of the deepest one-day drops in its history, and leveraged longs across the perpetual market were liquidated in a cascade.
- May 19, 2021 · the 5/19 crash. BTC collapsed intraday. Perpetual positions across all venues were liquidated in a chain reaction on a scale rarely seen at the time.
- June 2022 · the post-LUNA grind. BTC kept sliding in the weeks after LUNA, and Three Arrows Capital and several CeFi lenders were forced into unwinds, dragging leveraged positions down with them.
- August 5, 2024 · "Black Monday." The Bank of Japan rate hike unwound the yen carry trade, taking global risk assets down hard. BTC dropped sharply intraday. Perpetual positions across crypto exchanges were liquidated in a chain reaction, and the bulk of the liquidated longs were high-leverage positions. The market recovered afterwards, but the liquidated positions did not.
Three practical rules if you insist on leverage
First check that leveraged crypto products are permitted where you live and on the venue you use — many countries restrict them for retail. If they are:
- Isolated margin only. Never cross-margin a perp position. The downside of isolated is that position's margin; the downside of cross is your whole futures account balance.
- Hard stop-loss orders, not mental stops. Set the order at the exchange the moment the position opens. "I'll watch it and exit if it goes against me" rarely works in a leveraged market — the move that liquidates you can easily come while you are asleep or away from the screen.
- Maximum 2× leverage. 5× leaves a 20% buffer, 10× leaves 10%, 20× leaves 5% — and in BTC's volatility regime those are ordinary distances, not exotic ones. If you cannot make money with the position you would take at 2×, more leverage is not the missing variable. Before you size anything, put your win rate and risk-reward ratio together as in the risk-reward piece; if the expected value is negative, stop.
§9 · Closing
Leverage is a tool. It can make capital more efficient, and it can wipe out a position's margin in one ordinary sell-off. "A little leverage to speed up gains" leaves out its own second half: in a market this volatile, anything above 5× leaves a buffer thin enough for one ordinary sell-off to eat. The formula is easy to work out; the hard part is remembering it at the moment you place the order.
The most defensible setup for a new crypto trader is spot, DCA, zero leverage. If you must try, cap leverage at 2×, use isolated margin, set hard stops, and review your win rate, risk-reward and drawdowns on a regular schedule. The "I flipped my account on leverage" stories you hear are told by the winners; the people who blew up don't post screenshots. High leverage mostly costs you the chance to be around for the next opportunity. Spot plus time is slow, but it won't take you out on a single wick.