When a leveraged position moves against you, adding margin pushes your liquidation price further away — but it also locks more capital into a losing trade. This guide breaks down the mechanics of both options with numbers, so you can make the decision with a framework instead of panic. This is educational content about risk mechanics, not a recommendation to hold or close any position.
What adding margin actually does
When you open a leveraged position, your liquidation price is determined by the ratio of your margin to your position size. A $10,000 BTC perpetual at 10x leverage with $1,000 of margin gets liquidated when losses consume roughly 85–90% of that margin (the exact threshold depends on the exchange's maintenance margin rate).
Adding margin increases the numerator — total collateral — without changing position size. The result: effective leverage drops and liquidation moves further from the current mark price. Nothing else changes. Your entry price, unrealized PnL, and the position's exposure to further moves all stay the same.
This is worth stating plainly: adding margin does not improve the trade. It only gives the trade more room to breathe. Whether that room is useful depends entirely on why the position moved against you in the first place.
The math: how much margin moves liquidation
Consider a concrete example. You open a 10x long on ETH at $3,200 with $2,000 margin — a $20,000 notional position. With a maintenance margin rate of 0.5%, the liquidation price sits at roughly $2,912, or about 9% below entry.
ETH drops to $3,050. You are down $1,500 unrealized (7.5% of notional). Your remaining margin buffer is thin. Now you have a choice.
| Action | New margin | Effective leverage | New liquidation price | Distance from $3,050 |
|---|---|---|---|---|
| Do nothing | $2,000 (with $1,500 unrealized loss) | ~10x on remaining | ~$2,912 | 4.5% |
| Add $1,000 margin | $3,000 (with $1,500 unrealized loss) | ~6.7x | ~$2,785 | 8.7% |
| Add $2,000 margin | $4,000 (with $1,500 unrealized loss) | 5x | ~$2,660 | 12.8% |
| Close at $3,050 | $500 remaining | 0x | N/A | N/A |
The table shows the trade-off clearly. Adding $1,000 margin roughly doubles the distance to liquidation. Adding $2,000 more than triples it. But closing preserves $500 — the capital that would otherwise be at risk if ETH continues lower.
The decision reduces to one question: is the remaining $500 of capital better deployed defending this position, or freed for the next opportunity?
When adding margin is mechanically rational
There is no universal rule, but adding margin is defensible when all three of these conditions hold:
- The original thesis is intact. The position moved against you on noise, not because the structure that justified the entry broke down. A long entered on a support level still has that support unbroken; price just wicked through it.
- Liquidation is the binding risk, not the drawdown. You are not being stopped out by your own risk limit — you are close to the exchange's hard liquidation level, and the distance to the next structural level (support, resistance, a round number where order flow clusters) is meaningful.
- The added margin is capital you can afford to lose in full. If the position eventually gets liquidated even with the extra margin, the total loss — unrealized loss plus the added margin — should not impair your ability to trade again.
If any of these conditions fails, the calculus shifts. Adding margin to a thesis that has broken is not risk management — it is denial with a button on the exchange UI.
When closing is the better option
Closing is mechanically superior when:
- The reason for the trade no longer applies. You entered on a breakout that reversed. You entered on a catalyst that did not materialize. The structure that made the entry price attractive has changed. Adding margin to a dead thesis does not revive it.
- The drawdown has consumed your allocated risk capital. You decided before entering that this trade risked $500. It has lost $450. Close. The fact that liquidation is "only 4% away" is irrelevant — your personal risk threshold was hit first.
- Funding is compounding the bleed. If you are on the wrong side of funding and the rate is elevated, every eight-hour settlement adds to your loss. Adding margin does not stop funding. Closing does. See our funding settlement timing guide for how the clock works.
- Volatility is expanding. In a volatile regime, price can gap through your liquidation level before you have time to add more margin. The exchange's auto-add-margin feature (available on some venues) can help, but it requires pre-funded balance and still does not protect against gaps.
The sunk-cost trap
The most common mistake in this decision is not analytical — it is psychological. The trader thinks: "I have already lost $1,500 on this position. If I close now, that loss is real. If I add margin and price comes back, I can recover." This is the sunk-cost fallacy, and it is the single largest driver of margin-adding behavior that ends in full liquidation.
The $1,500 unrealized loss is already real. It exists whether you close or add margin. The only forward-looking question is: from this point, with this capital, what is the highest-expectancy use of it? Defending a losing position is almost never the answer — unless the three conditions in Section 3 all hold simultaneously, which is rare.
The margin you add to a losing position is not "saving" the trade. It is placing a new bet that the position will recover. Evaluate that bet as you would any new trade — on its own merits, independent of what you have already lost.
Ready to start trading?
Open a free account on OKX →