Calling a rebound and keeping a position until it arrives are different achievements. A hypothetical 10× long starting with 1,000 USDT closes with 2,000 USDT if the price rises straight by 10%. If it first falls 10% before reaching the same endpoint, our model below leaves zero.
Before putting more money behind a rebound, separate where the price might go from the quantity that will still be open then. Compare unborrowed spot, a 2× long and a 10× long on the same starting money, then compare the price of adding collateral with the price of reducing exposure.
Same final price, different money
Set Bitcoin’s initial price at 100,000 USDT and its final price at 110,000 USDT. These are fictional prices. Path A rises directly to the endpoint; path B first falls to 90,000 USDT and then rises. Assume continuous price movement within each path.
Every choice starts with 1,000 USDT. Unborrowed spot buys 0.01 BTC. A USDT-settled linear contract gives the 2× long exposure to 0.02 BTC of price movement and the 10× long exposure to 0.1 BTC. The futures quantity measures price exposure; it is not that quantity of coins purchased into a wallet.
Each futures case is a single isolated position at an assumed lowest risk tier, with a fixed 0.5% maintenance margin rate. There are no other orders, positions or collateral, automatic margin replenishment or re-entry. Full liquidation is assumed to consume all allocated margin through bankruptcy-price settlement. Surviving futures and spot are fully closed or sold at the final price. The maintenance margin deduction, fees, funding, taxes, slippage and collateral haircuts are zero. Spot sale, futures mark and futures execution all use the same fictional price. This is not a current product risk tier or a reconstruction of an actual account. All money is measured in USDT units.
| Start: 1,000 USDT | Direct rise | Dip then rise |
|---|---|---|
| Unborrowed spot | 1,100 | 1,100 |
| 2× long | 1,200 | 1,200 |
| 10× long | 2,000 | 0 |
The 10× long on path B is fully liquidated before the final 110,000 arrives. Its open quantity is zero when the later rebound occurs. Entering again with new money is a separate trade, not recovery of the original position in this table.
On path A, however, 10× leaves the most money. Reducing exposure to 2× on the same 1,000 USDT cuts the gain from 1,000 to 200 USDT. Lower leverage gives up money in the rising path in exchange for room to withstand an intervening decline.
Why the liquidation line arrives before a 10% fall
Let C be starting collateral, q the futures quantity in BTC, P₀ the entry price, P the mark price and m the maintenance rate. Equity in this zero-cost model is C + q × (P − P₀); required maintenance margin is q × P × m. Equating the two gives the liquidation trigger (q × P₀ − C) ÷ [q × (1 − m)]. This comparison removes fees and deductions from the mark-based structure described in Bybit’s margin calculation explanation.
With C = 1,000, q = 0.1, P₀ = 100,000 and m = 0.005, the 10× trigger is about 90,452.26 USDT, roughly 9.55% below entry. Equity and required maintenance at that line are each about 45.23 USDT. That is a trigger-time calculation, not money promised back for withdrawal.
The separate bankruptcy reference, where equity reaches zero, is P₀ − C ÷ q = 90,000 USDT. Bybit’s bankruptcy-price explanation describes takeover at that reference, with any margin remaining from a better subsequent execution going to the insurance fund. The liquidation trigger, takeover settlement and market execution have different roles.
The current liquidation rules describe full closing at bankruptcy for a lowest-tier position if canceling orders cannot restore margin, and attempts at partial closing for higher tiers. Do not turn every liquidation into zero for every asset: inspect partial versus full takeover, remaining quantity and settlement.
Adding collateral and cutting quantity are different 2× choices
Compare two ways to create more room before the price moves: keep the original 0.1 BTC exposure and raise collateral to 5,000 USDT, or keep 1,000 USDT collateral and reduce exposure to 0.02 BTC. These are separate initial setups, not deposits made after liquidation.
Both have initial effective leverage of 2×, calculated as initial notional divided by collateral. Under the same assumptions, both trigger at about 50,251.26 USDT. Their capital commitments and absolute gains or losses differ.
| Initial choice | Collateral | At 90,000 | Final cash |
|---|---|---|---|
| 0.1 BTC; add collateral | 5,000 | 4,000 | 6,000 |
| 0.02 BTC; cut exposure | 1,000 | 800 | 1,200 |
Amounts are USDT. The first setup loses 1,000 during the dip; the second loses 200. Final gains are also 1,000 and 200, respectively, both 20% of their own starting collateral. The first ties up five times as much money to obtain the same liquidation distance. Adding collateral does not reduce the cash loss on unchanged exposure. The newly allocated money also bears that position’s risk.
After changing a leverage setting, inspect what actually happened to quantity and allocated collateral. A smaller number in a menu does not by itself show that the existing price exposure became smaller.
Can one stop-loss price remove the difference?
Read the liquidation reference alongside the stop trigger. Bybit’s order execution and liquidation FAQ explains that mark-based liquidation can occur before a stop triggered by the last traded price. Triggering and execution are also different events: liquidity and slippage affect the price at which a position actually closes.
Inspect the stop’s trigger reference, order type and filled quantity. A number before liquidation does not secure an exit at that price.
Write down four things before the next rebound
Open quantity now. A fully closed position and an open position carrying a loss participate differently in the next price rise. Spot prices are not guaranteed to recover either, but unborrowed spot in this model has no margin-shortfall condition forcing its quantity closed.
Your money committed to this exposure. Compare extra collateral and reduced quantity on consistent capital denominators. A 1,000 gain on 5,000 and a 200 gain on 1,000 cannot be ranked by absolute profit alone.
Isolated or cross margin. This model is isolated. Cross-margin liquidation depends on the account maintenance margin ratio, other positions and available collateral; the displayed price is a reference.
Your chosen price and the actual termination conditions. Read the current risk tier, maintenance rate, collateral, liquidation reference and order execution together. If they change, recalculate the fixed-condition comparison. Seeing how much more money becomes exposed to buy time for a rebound makes the next deposit a clearer choice.
For the recovery problem in a fund that resets leverage daily, read why a leveraged ETF can rebound without breaking even. For the difference between submitting an order and the money left after a fill, compare signal price and actual execution. In Whalehunter’s Korean-language community, begin with quantity, collateral and termination conditions in the same table before debating the next price forecast.