Leverage lets you control a big position with a small deposit. At 10x leverage, $100 of your own money controls $1,000 of Bitcoin. So if Bitcoin goes up 5%, you don't make 5%. You make 50%.
That is the entire pitch, and the pitch is true. This is why every exchange puts the leverage slider front and centre and lets you drag it up to 50x or 100x. It looks like a shortcut: same trade, ten times the profit.
Here is the sentence the slider never shows you. It works in exactly the same way on the way down. If that position moves 5% against you, you don't lose 5%. You lose 50%. And a few more percent after that, you don't lose the rest. You lose everything.
Liquidation means the exchange force-closes your position because your deposit can no longer cover the loss. This is the part beginners underestimate. You are not "down a lot" and waiting for a bounce. Your money is gone. The position no longer exists. If the price rockets back the next hour, it comes back without you.
How far does the price have to move against you before that happens? Roughly 100 divided by your leverage, as a percentage:
| Leverage | Move against you that wipes it out | In plain terms |
|---|---|---|
| 2x | about 50% | very hard to hit |
| 3x | about 33% | a bad crash |
| 5x | about 20% | a rough week |
| 10x | about 10% | an ordinary Tuesday in crypto |
| 20x | about 5% | a quiet day can do it |
| 50x | about 2% | lunch |
| 100x | about 1% | a rounding error |
These are the friendly numbers. In reality liquidation hits slightly sooner than this, once you count trading fees and the exchange's maintenance margin, and on perpetual futures a funding fee quietly bleeds your deposit the whole time you hold. So treat the table as the best case.
Now put that next to how crypto actually moves. A 10% down day in Bitcoin is not rare. It is a normal Tuesday. Over the last six years Bitcoin has had 47 single days that would have liquidated a 10x position, and 257 days that would have liquidated a 20x one. You do not need to be wrong about the market. You need one bad day while you happen to be holding.
Talk is cheap, so here is real data. We took a strategy that actually passed our testing — the 9/21 EMA crossover, one of the only two on this whole site that beat simply holding Bitcoin (its full autopsy is here). Then we ran the identical trades on Bitcoin, over the same six years, and changed one thing: the leverage.
| Leverage | $1,000 became | Return | Worst drawdown | Outcome |
|---|---|---|---|---|
| 1x (no leverage) | $11,328 | +1,033% | −61% | survived |
| 2x | $41,117 | +4,012% | −87% | survived, barely bearable |
| 3x | $46,571 | +4,557% | −97% | survived on paper only |
| 5x | $0 | −100% | −100% | liquidated, Jan 2021 |
| 10x | $0 | −100% | −100% | liquidated, May 2020 |
| 20x | $0 | −100% | −100% | liquidated, Nov 2019 |
Read that honestly, because there are two lessons and the first one is the trap.
One: a little leverage did make a real edge more money. 3x turned a +1,033% result into +4,557%. That is real, and it is exactly why leverage is so hard to resist. If the story ended there, everyone would use it.
Two: now look at the drawdown column. To earn that 3x result you had to sit through a 97% loss along the way. Ninety-seven percent. Your $46,000 would spend long stretches showing $1,400 on the screen. Almost nobody holds through that. They sell at the bottom, swear off trading, and never see the recovery. The paper result and the human result are two different people.
And cross one more line — 5x — and it does not just make less. It makes nothing. Zero, permanently, on a single day in January 2021. The strategy still had a real edge the entire time. The leverage killed it anyway. And you could not have known in advance where that line was.
Most people never run a tested edge. They just buy and hold with leverage because they feel bullish. So here is plain leveraged buy-and-hold on Bitcoin, no strategy at all, over the same six years:
| Leverage | $1,000 became | Worst drawdown | Outcome |
|---|---|---|---|
| 1x (just holding) | $7,533 | −77% | fine |
| 2x | $3,005 | −97% | less than 1x, with double the risk |
| 3x | $0 | −100% | liquidated Mar 2020, never came back |
Look at the 2x row. Double the exposure, and you ended up with less money than doing it with no leverage at all ($3,005 versus $7,533). That is not a typo, and it is the single most important idea on this page.
| If you lose | You need this just to break even |
|---|---|
| 25% | +33% |
| 50% | +100% |
| 75% | +300% |
| 90% | +900% |
This is why the market can finish far higher than where it started and the leveraged holder can still finish lower, or at zero. And the 3x holder did not even get the chance. One day — the March 2020 crash — wiped the account, and it never recovered, no matter how high Bitcoin later went.
This page is not "never touch leverage." It is "know what you are actually buying." Leverage sells you a bigger number on the gains, and quietly hands you two bills the advert never mentions:
First, a drawdown so deep you will not emotionally survive it, even in the cases where the maths technically does. Second, a cliff edge to zero whose exact location you cannot see until you have already gone over it.
Across the dozens of strategies we have put through this same testing, the honest answer kept landing in the low single digits at most, and even there the drawdowns were brutal. The people who blow up rarely do it because the strategy was bad. They do it because the leverage turned a survivable bad month into a permanent zero.
If you take one thing from this page, take this: the return is not the number that matters. The drawdown is. Leverage flatters the first and quietly wrecks the second, and the account is settled by the second.
We put the famous ones through the same honest test — multi-year, out-of-sample, fees included, measured against simply doing nothing. Most of them died. Read the graveyard →
Soon you will be able to run your own strategy through the exact same gauntlet, at the leverage you actually trade, and see where it lands before a single real dollar is on the line.