Most people who buy an asset at 2x leverage do the maths like this: double the profit, double the loss. In dollars, the first half is true. If the price rises 100 percent, you earn twice as much as without leverage, and your stake in dollars has already doubled at +50 percent.
But count in what you are betting on: units of the asset, whether Bitcoin, Ether or any other coin. Holding 1 coin always leaves you 1 coin. Going 2x long with the same money never gets you to 2 coins, however high the price climbs. If the price halves, you have zero. Measured in the asset, the 2x position is capped on the way up and not on the way down. Knowing that before you enter changes how big you make the position.
In dollars you double, in coins you never do
This is how leverage trading works: you put up your own money as collateral, the exchange lends you the rest, and you trade a bigger position than your money alone would buy. Gains and losses on the whole position are yours; the loan stays the same.
The example: $1,000 of your own money, worth exactly 1 coin at entry, plus $1,000 borrowed, 2 coins in total. That is a 2x position with a fixed size, as with a perpetual future or a margin loan. Next to it is someone holding their 1 coin without leverage.
| Price | Holding, in dollars | At 2x, in dollars | At 2x, in coins (holder: always 1) | Actual leverage |
|---|---|---|---|---|
| −50 % | $500 | $0 | 0 | infinite |
| −40 % | $600 | $200 | 0.33 | 6x |
| −25 % | $750 | $500 | 0.67 | 3x |
| ±0 | $1,000 | $1,000 | 1.00 | 2x |
| +50 % | $1,500 | $2,000 | 1.33 | 1.5x |
| +100 % | $2,000 | $3,000 | 1.50 | 1.33x |
| +300 % | $4,000 | $7,000 | 1.75 | 1.14x |
In dollars it is simple. At +100 percent the holder earns $1,000 and the 2x position $2,000; the profit is exactly twice as large. In coins it looks different. One coin now costs $2,000, and the 2x position is worth $3,000, which is 1.5 coins. At +300 percent it is 1.75 coins, and the higher the price goes, the closer the position gets to 2 coins without ever reaching it. At −50 percent it is zero.
The reason is the loan. It runs in dollars and does not grow. The borrowed $1,000 stays $1,000 however high the price climbs, so it costs fewer and fewer coins. On the way down it costs more and more, until at −50 percent it is worth all of your 2 coins.
On the way up, the loan works less and less for you. On the way down, it works more and more against you.
So why does the crash come so suddenly?
The leverage does not stay at 2x
On the way up the leverage unwinds itself, on the way down it builds itself up. That holds for every asset.
Actual leverage is the position divided by your own money. When the price rises, your own money grows faster than the position, and leverage falls: to 1.33x at +100 percent, to 1.14x at +300 percent. When the price falls, your own money shrinks faster than the position, and leverage rises: to 3x at −25 percent, to 6x at −40 percent.
So at −40 percent, every further percent of price decline costs you 6 percent of what you have left. The last ten points down to −50 take everything that remains.
A 2x position does not stay a 2x position. It turns more cautious on the way up and more reckless on the way down, without you doing anything.
Where exactly does the exchange pull the plug?
The exchange does not wait for −50 percent
Exchanges close a leveraged position before your money is completely used up, as soon as it falls below a maintenance margin. With a maintenance margin of 0.4 percent of the position, the line sits at −49.8 percent instead of −50.
On top comes funding, the ongoing fee longs pay shorts on perpetual futures. For Bitcoin perpetuals on Binance it averaged 3.34 percent a year over the last twelve months, charged on the position. At 2x that is 6.7 percent of your own money per year. After a year at that rate, liquidation is no longer at −50 but at about −46 percent.
The costs run against you even when the price does not move.
How often would real prices have hit that line?
What the history of Bitcoin and Ether shows
The test: every day from 17 August 2017 to 5 October 2025 is an entry day, 2,972 in all. On each one a 2x long opens at the daily close and runs for a year. It counts as liquidated if the daily low within that year falls 49.8 percent below the entry. Calculated for Bitcoin and for Ether, before funding.
| Entry year | Bitcoin: liquidated within a year | Ether: liquidated within a year |
|---|---|---|
| 2017 (from 17 August) | 27.7 % | 41.6 % |
| 2018 | 80.8 % | 88.2 % |
| 2019 | 52.3 % | 52.3 % |
| 2020 | 18.0 % | 12.3 % |
| 2021 | 54.2 % | 58.9 % |
| 2022 | 36.7 % | 43.6 % |
| 2023 | 0.0 % | 0.0 % |
| 2024 | 0.0 % | 53.3 % |
| 2025 (to 5 October) | 18.7 % | 43.5 % |
| all | 32.8 % | 43.9 % |
For Bitcoin the position was gone within a year after one in three entry days, after a median of 170 days. For Ether it was almost one in two, after a median of 135 days. With funding at the rate of the last twelve months, Bitcoin's figure would have been 36.4 percent. The entry days overlap; this is a little over eight independent years, not 2,972 independent trials.
And how did the 2x position compare with holding after a year?
| After one year, all entry days, in dollars | Bitcoin held | Bitcoin at 2x | Ether held | Ether at 2x |
|---|---|---|---|---|
| Average | 1.75 times | 2.13 times | 2.18 times | 2.96 times |
| Median | 1.40 times | 1.33 times | 1.09 times | 0.65 times |
| Entries where 2x ended worse | 47.8 % | 54.1 % |
On average leverage comes out ahead for both; at the median it falls behind. A few entries in strong years carry the average. For Ether, the typical 2x position ended the year with less money than it started with, while the holder was slightly up.
The more an asset swings, the more often it hits the line. The mechanics are the same; only the frequency differs.
The obvious objection
"I'll just add money before it blows up. Or I'll set a stop."
Both change where the loss stops, not the shape of the curve. Adding money puts more of your own capital into the same position and lowers the leverage; the same maths then applies to a larger stake. A stop is a liquidation you set earlier than the exchange would. It protects the rest of your money, but it takes away what the holder still has at −50 percent: all of their holdings and the option to wait.
2x ETFs, which reset their leverage every day, work differently. They do not go to zero, but they lose value in choppy markets. The piece How leveraged and inverse ETFs work explains that.
What this means for you
If you are considering 2x, do not ask how much more you earn on the way up. Ask whether the asset can lose half on the way there. For Bitcoin since 2017, that happened within a year after one in three entry days, for Ether after almost one in two.
If you hold and the price halves, you still have all of your holdings and can wait. Once a 2x position is liquidated, you have none. The recovery then happens without you.
If you want leverage anyway: the line sits at about −50 percent for 2x, about −33 percent for 3x and about −20 percent for 5x, for every asset. How often your asset has had drops like that decides how much leverage it can carry. The leverage backtest shows what that looks like inside a strategy, and the piece on the 5x grid bot explains why a leveraged grid rarely gets liquidated.
Leverage does not double your holdings. It doubles your bet that the price will not halve.
What remains open is how far funding shifts the picture in years with high rates. For Bitcoin, the calculation here applies the rate of the last twelve months to every year.
FAQ
Why count in coins and not in dollars? Going 2x long on an asset means wanting more out of it than the holder gets. The holder always has their 1 coin, so counting in coins shows directly what leverage adds over holding: at most just under double on the way up, zero on the way down. In dollars, your stake can of course double and more at 2x.
Does this hold for every asset? The curves hold for any long position with a fixed size and a loan in dollars, whether Bitcoin, Ether, an altcoin or a tokenised asset. The asset only decides how often the price reaches the line.
What does "fixed position size" mean? The position stays at the amount you buy at the start, and the loan stays the same. That is how perpetual futures and margin loans work as long as you change nothing. Leveraged ETFs, by contrast, reset to 2x every day.
How is the liquidation line calculated? Liquidation happens when your remaining money falls to the maintenance margin. With a 0.4 percent maintenance margin, that is the price at which $1,000 plus the position's profit equals 0.4 % of the position: −49.8 percent. Actual rates depend on the exchange, the asset and the position size.
Why the daily low and not the close? A liquidation happens as soon as the price touches the line, even if it recovers the same day. Daily candles show the low, but not spikes that only happened on single exchanges.
Where does the data come from? BTCUSDT and ETHUSDT daily candles from Binance, 17 August 2017 to 5 October 2026. Funding from the Arena's funding history, Bitcoin perpetuals on Binance, 5 October 2025 to 5 October 2026.
Not investment advice, not a recommendation, not a forecast — historical patterns are no promise.
Study the Past — Improve your Future 🥋