Why a profitable Bitcoin trade can still lose money, why the liquidation price is not a stop-loss, and how to measure the real cost of keeping a leveraged position open.
A perpetual future lets a trader take long or short exposure to a cryptocurrency without buying the underlying coin and without a conventional expiry date. That apparently simple proposition hides three interacting systems: a price-linking mechanism that periodically transfers funding between traders, a margin engine that can close positions before collateral reaches zero, and an execution market where fees and slippage are charged on exposure rather than the trader’s deposit. A fivefold leveraged position is not a fivefold better prediction; it is a fivefold exposure to errors in price, timing and market liquidity.
The central question is not whether perps track Bitcoin closely. It is whether the anticipated price move will be large enough, and occur soon enough, to cover funding, trading costs and the probability of forced exit. An illustrative $10,000 notional long backed by $1,000 has a theoretical $1,000 gross loss after a 10% adverse move, before fees or maintenance margin. In practice it may be liquidated earlier. A relatively small funding bill can also consume a visible share of the same $1,000 deposit.
The Contract Is a Claim on Price Changes, Not Bitcoin
An ordinary spot purchase gives the investor crypto exposure through an asset held directly or through a custodian. A USDT-margined BTC perpetual is instead a derivative position: the trader posts collateral, receives a mark-to-market profit or loss, and never needs to take delivery of Bitcoin. Contracts denominated or settled in the underlying coin are different again. Inverse BTC contracts introduce an additional source of risk because the value of BTC collateral changes as Bitcoin moves, even apart from the contract’s profit and loss. Product jurisdiction, eligibility and clearing structure also matter: the rules of an offshore crypto exchange need not match those of a regulated US futures exchange.
Perpetuals resemble expiring futures in leverage and margin, but they solve the expiry problem using a recurring funding transfer. The mechanism is designed to discourage persistent differences between the futures price and a spot reference price. It does not guarantee equality at each moment, especially in a volatile market. On Coinbase Derivatives’ US perpetual-style futures, funding is calculated hourly; Binance’s published default for many contracts is an eight-hour interval, subject to change. Therefore a quoted rate without its interval, contract and venue is incomplete information. [1][2][3]
A long position gains when the price moves up and loses when it falls. A short position does the reverse. Both can be liquidated; unlike an unleveraged spot holding, neither has the practical luxury of waiting indefinitely through an adverse price movement. That is the cost of borrowing exposure from a platform’s risk engine rather than simply holding an asset.
Funding: The Cost That Changes Direction
Funding is usually a transfer between participants on opposing sides of the contract, not a fixed interest charge retained by the exchange. When a rate is positive, longs normally pay shorts. When negative, shorts normally pay longs. The payment depends on position notional at the funding observation: funding payment = position value multiplied by the rate. Exchanges calculate the rate using different variations of the contract/spot premium, reference indexes, interest assumptions, clamps and caps. Binance explicitly says it may change intervals and limits; Bybit publishes contract-specific details. [1][4]
Suppose an account has $1,000 and opens a $10,000 long, giving 10x exposure. At a positive 0.01% rate every eight hours, the long pays $1 per interval, or $3 per day if the size and rate remain unchanged. A seven-day holding period then costs $21, equal to 2.1% of initial margin. At 0.05% per interval, the same holding period costs $105, or 10.5% of margin. These are illustrations, not observed or forecast rates. They exclude the effect of price changes on notional, exchange-specific settlement and any changes to the rate.
Figure 1. Funding sensitivity | Original fixed-notional scenario.
The fact that funding is paid to the other side creates a tempting trade: short a rich perpetual while holding the underlying spot asset. But ‘cash and carry’ is not risk-free. The hedge can fail through exchange default, temporary basis expansion, collateral mismatch, borrowing and custody costs, liquidation of one leg, or a sudden negative funding reversal. And a highly positive current funding rate is not a promise that the next 21 payments will be identical.
An annualized funding figure is especially easy to misuse. The simple arithmetic annualization of 0.01% every eight hours is 10.95% of notional per 365-day year; at 10x leverage it is notionally 109.5% of the $1,000 starting margin. That is a scenario conversion, not a reliable expected annual return or cost. No position can be assumed to survive a year unchanged, and the rates are market-contingent. Traders should review the history, next scheduled payment, cap and actual position value rather than trade from a single eye-catching percentage. [1][3]
The $1,000 Account: A Full-Cost Trade Model
The cheapest-looking trade is not necessarily the one that leaves the trader with the most money. Take a concrete benchmark using published Bybit non-VIP derivatives fees as of the research date: 0.055% for market-taking orders and 0.020% for eligible maker fills. Rates vary by region, VIP tier and product, so they must not be treated as universal. [5][6]
A trader deposits $1,000, takes a $10,000 notional long, holds the position for seven days and closes without any Bitcoin price change. At the quoted taker rate, opening costs $5.50 and closing costs $5.50. Assume a 0.01% eight-hour funding rate throughout: another $21. Assume illustrative execution slippage totaling four basis points across entry and exit: $4. The modeled total cost is $36, or 3.6% of initial capital despite a flat underlying market. No actual platform spread or live slippage quote is assumed here; the four basis points is deliberately an explicit hypothetical input.
Now suppose Bitcoin appreciates 2% and the position keeps approximately $10,000 notional exposure for simplicity. Gross profit is $200; after $36 estimated friction the net gain is $164, equal to 16.4% of the margin before taxes and other costs. A 2% decline instead generates roughly a $200 gross loss and a $236 loss after these same frictions, or 23.6% of the deposit. Actual fee notional and funding can change during the trade, so the values are simplified first-order estimates.
The model’s break-even underlying move is about 0.36% when flat notional and long-positive funding are assumed: $36 divided by $10,000. At 0.05% funding each eight-hour interval, the seven-day bill rises from $21 to $105, taking all-in costs to $120 and the required favorable move to 1.20%. Any trader who estimates the likely BTC move as 0.5% but ignores a persistent expensive funding regime is testing the wrong proposition.
| Seven-day $10,000 notional long | Mild funding | Expensive funding |
| Margin posted | $1,000 | $1,000 |
| Round-trip taker fees | $11 | $11 |
| Funding (21 intervals, fixed hypothetical rate) | $21 | $105 |
| Round-trip slippage assumption | $4 | $4 |
| Total modeled cost | $36 | $120 |
| Break-even positive price move | 0.36% | 1.20% |
| Cost as % of initial margin | 3.6% | 12.0% |
Figure 2. Seven-day carrying and execution costs | Bybit base rates; other inputs illustrative.
Leverage Is About Exposure, Not Just a Setting
A displayed ’10x’ means $1,000 in initial margin supports about $10,000 of initial notional, before fee and platform adjustments. A 1% move in the underlying is approximately a 10% change in the starting margin, not a 10% gain on the entire account automatically. If the trader deposits $10,000 but uses just $1,000 for an isolated position, the position’s risk is not the same as if the full account is available as cross margin. Dollar position size, effective leverage against total equity and collateral allocation are separate inputs.
Traders can choose the same dollar exposure at different leverage settings by allocating different amounts of collateral. Increasing leverage without increasing notional reduces the loss buffer and brings the theoretical liquidation region closer. Increasing both leverage and notional compounds the effect. In portfolio margin, offsets and risk-based models add further complexity; a supposedly diversified account can see its margin relief contract when assets become correlated in a crash.
For a simplified isolated long, ignore fees and funding, assume stable USDT collateral and maintenance margin equal to 0.5% of fixed notional. Starting equity is notional/leverage. Liquidation roughly begins when price loss consumes initial margin less required maintenance. The corresponding adverse price movement is approximately (1/leverage – 0.005). This is an educational boundary, not a venue liquidation-price formula. Real systems include maintenance tiers, mark price, fee reserves, position updates, collateral haircuts and account rules. At 5x the simplistic threshold is 19.5%; at 10x, 9.5%; at 20x, 4.5%; at 50x, 1.5%. At higher leverage, tiny differences between exchange-specific calculation details dominate.
Figure 3. Simplified liquidation-distance sensitivity | Not an exchange liquidation formula.
A stop-loss cannot guarantee escape before that threshold. It can trigger on one price measure and execute at another; a fast gap may produce a much worse fill. Bybit says its liquidations are based on the mark price, and it warns that a stop positioned beyond liquidation may never trigger. Some market interfaces emphasize the last traded price instead, so a trader can appear ‘safe’ on the candlestick chart while the mark price has already breached the account’s limit. [7][8]
Mark Price, Last Price and the Liquidation Engine
The last price is an executed trade. An index price is a reference drawn from selected underlying spot markets. A mark price is a venue’s risk-management estimate based on index, basis and other exchange-defined inputs. A trader’s liquidation is often determined by the mark price or an account-level maintenance margin measure, rather than by a single last traded print. Binance’s FAQ and Bybit’s updated September 2026 methodology explain these distinctions in different implementations. [7][9]
Liquidation is not a stop-loss service. The exchange’s objective is to protect its clearing system from a deficit. A liquidation engine can take control, close at unfavorable prices and charge additional fees, with exact rules varying by product. Some venues use insurance funds to absorb shortfalls; if those funds cannot handle certain bankrupt positions, automatic deleveraging can reduce the positions of otherwise profitable counterparties. Bybit describes an insurance-fund-to-ADL sequence, while Binance’s ADL documentation similarly positions it as a backstop. [10][11]
This is why two traders holding ‘the same’ Bitcoin short can face different outcomes. Their maintenance requirements, collateral currency, account balances, position sizes, funding history, execution liquidity and mark-price methodology differ. A market-wide liquidation figure aggregates this mechanically induced closing of positions; it is not identical to fresh cash leaving the crypto ecosystem or a measure of unique trader losses. Liquidations can amplify volatility when hedging and forced trading occur into shallow depth, but the reported nominal liquidation value is not equal to realized wealth destroyed.
Isolated Versus Cross Margin: Whose Money Is at Risk?
In isolated margin, collateral is allocated to a specific position under the venue’s rules. Losses can consume the allocation, although liquidation costs and automatic margin-add features need to be checked. In cross margin, available equity supports positions collectively; an adverse move in one market may draw on assets that the trader thought belonged to a different strategy. Bybit says its unified account’s cross margin calculations use mark-price-based position values; its isolated and portfolio modes have different calculations. [12][13]
For a trader with $5,000 total account equity and a $1,000-margin speculative trade, cross margin can make the other $4,000 economically relevant to that position. This does not mean that a $1,000 isolated allocation is necessarily the maximum possible legal liability on every platform; terms, negative balances, borrowing and mode settings differ. What matters is the pool of collateral against which the platform’s risk engine can act. Traders should find the ‘auto-add margin’ switch and read what collateral can be drawn down before opening orders.
Collateral denomination creates a second asymmetry. Imagine an inverse Bitcoin perpetual whose margin is held in BTC. A BTC price decline harms a BTC long and can simultaneously reduce the dollar value of its margin. A dollar stablecoin might avoid that particular directional collateral spiral but introduces stablecoin issuer, custody and depeg risk. The contract’s leverage is only one number in a wider balance-sheet problem. [14]
Liquidity, Order Types and What the Quote Does Not Say
A deep market may provide narrow spreads for small orders but not for an account trying to close during a cascade. Maker fees are lower because passive orders add liquidity, yet the maker order might never fill or may fill just before the market moves against it. A taker pays a higher explicit fee in exchange for immediate execution, but a stop-market order can still sweep thin liquidity. Bid-ask spread, slippage and fees should be considered together, rather than treating the fee schedule as the execution cost.
Funding and basis can also be venue-specific. A local perpetual price can temporarily disconnect from a global spot index, and weekend liquidity or smaller altcoins can worsen price impact. An apparent funding arbitrage becomes less attractive when the trader must buy an expensive spot hedge, borrow assets, transfer funds across venues or maintain margins on both legs. Cross-venue margin cannot usually be netted; a market-neutral combined portfolio can still have one leg liquidated.
The choice between a market order and a limit order is therefore a choice between execution certainty and price control, not simply between 0.055% and 0.020%. The correct question is what fill the trader can reasonably expect at the size and time needed. [5][6]
A Retail Risk Checklist That Survives Real Market Conditions
Before entering, calculate notional divided by total account equity and by the specific margin allocation. Display mark price, liquidation price and the account’s maintenance ratio beside the stop-loss. Look at historical funding across volatile sessions and convert a plausible expensive rate into dollar cost over the planned holding period. Record contract settlement currency, collateral haircut and whether margin can be auto-added. Set a maximum loss in account dollars rather than judging risk by an exchange’s leverage slider alone.
For an illustrative $1,000 risk budget, if the trader intends to lose no more than $50 on a 2% adverse move, an unleveraged first-order position size is about $2,500 before fees and slippage. Setting leverage at 20x does not change the dollar exposure needed; it simply means roughly $125 of isolated initial collateral, leaving an extremely thin liquidation buffer. An ordinary stop at 2% may prove useless if liquidation arrives earlier. To make the intended loss limit robust, the trader needs a larger margin buffer, smaller position, or both.
The falsification test for a funding-based strategy is simple: rerun it assuming the funding rate reverses, slippage doubles and one leg cannot exit at the expected price. If the profit disappears, the advertised opportunity was largely compensation for risks the initial calculation excluded. The same discipline applies to directional speculation: an expected gain of 0.5% is not useful when modeled execution and funding consume 1.2% of notional.
Bottom Line
A crypto perpetual future offers almost immediate exposure, not free optionality. Funding is a variable transfer, leverage converts modest market moves into large changes in margin, and liquidation is a platform risk-control event rather than an investor protection. The market exposure can be easy to understand while the survival mechanics are not. Retail traders should start with the dollar value of the position and its funding and exit scenarios, then decide whether leverage adds any value to the thesis.
Methodology and Assumptions
All numerical scenarios are calculations created for this article, not backtests, forecasts or current market quotes. The base model fixes initial position notional at $10,000, margin at $1,000, assumed funding at +0.01% or +0.05% per eight-hour interval, seven days at 21 intervals, Bybit’s published illustrative non-VIP 0.055% taker fees on each side, and hypothetical combined slippage of four basis points. Funding and fee notionals are held fixed solely for comparability. The approximation of liquidation distance assumes no fee/funding deductions and a constant 0.5% maintenance requirement; it does not reproduce the actual liquidation formulas used by exchanges. Fees and product access depend on account, jurisdiction, instrument and date.
Sources
[1] Binance, Introduction to Binance Futures Funding Rates (updated March 6, 2026)
[2] Coinbase Derivatives, Perpetual Futures Overview
[3] Coinbase Derivatives, Funding Rate Mechanism
[4] Bybit, Funding Fee Calculation (updated May 12, 2026)
[5] Bybit, Futures Contracts: Fees Explained (updated August 5, 2026)
[6] Bybit, Trading Fee Structure
[7] Bybit, Mark Price (updated September 4, 2026)
[8] Bybit, Take Profit and Stop Loss
[9] Binance, Futures Liquidation Protocols (updated January 4, 2026)
[10] Bybit, Bankruptcy Price and Insurance Fund (updated August 11, 2026)
[11] Binance, Auto-Deleveraging (updated January 13, 2026)
[12] Bybit, Margin Calculations Under Different Margin Modes (updated August 7, 2026)
[13] Bybit, Adjustment and Impact of the New Margin Calculation (updated June 18, 2026)
[14] Bybit, Introduction to Inverse Perpetual and Expiry Contracts
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape. You can reach out to him via his social media accounts:
Linkedin: https://www.linkedin.com/in/johan-shamshad-742851262/
X: https://x.com/Yasmine_FX
Investing: https://www.investing.com/members/contributors/279781574