Quick Answer: In a simplified, fully matched perpetual-futures book, one trader’s price P&L is offset by another trader’s opposite price P&L. Funding generally transfers value between long and short holders; it does not create a new market-wide profit. Trading fees and other costs leave traders collectively with less than the zero-cost sum. “Zero-sum” therefore describes a carefully defined price-transfer model, not a promise that every real account, venue or trade has exactly one identifiable opposite winner or loser.
What does “zero-sum” measure in a perpetual contract?
It measures the sum of gains and losses inside a defined group, over a defined period, for a defined component of the transaction. That boundary matters. Consider only the change in contract value for all opposite positions in a closed, matched system, and gains and losses offset. Add funding, fees, liquidations, insurance mechanisms, outside hedges or unmatched observation windows, and a single number labeled “the market’s profit” becomes misleading.
A perpetual is a derivative: it gives exposure to the price movement of a reference asset without requiring the position holder to own that asset. A long generally benefits from an increase in the contract price, while a short generally benefits from a decrease. If equally sized opposite exposures are opened and closed at the same two prices, their gross price P&L adds to zero. This is an accounting statement under those assumptions, not an investment forecast.
Phemex’s broad zero-sum explainer introduces the idea across asset classes. The question here is narrower: what happens when an actual perpetual account has funding transfers and transaction costs? The distinction is useful because a trader can be right about direction and still lose money after costs, or be wrong about direction while receiving funding that partly offsets a price loss.
Perpetual contracts usually have no fixed expiry date. The absence of expiration does not, by itself, decide whether opposite price exposures offset. The P&L transfer relationship arises from the matched derivative exposures. Expiration is a feature of dated futures, not a prerequisite for the simplified zero-sum logic. Holding time instead affects how many funding events and fees a trader may incur.
A two-trader ledger with price P&L, funding and fees
Here is a hypothetical teaching example, expressed in USDT. Trader A holds a long and Trader B holds an equally sized short over the same interval. Assume the price move gives A +100 USDT in gross contract P&L and B −100 USDT. At one funding event, assume a positive funding rate results in A paying B 5 USDT. Finally, assume each pays 2 USDT in total trading fees to the venue across the relevant executions. The numbers are invented to show accounting; they are not Phemex’s current fee or funding schedule.
| Component, hypothetical USDT | Long A | Short B | Sum for the two traders |
|---|---|---|---|
| Contract price P&L | +100 | −100 | 0 |
| Funding transfer | −5 | +5 | 0 |
| Trading fees paid | −2 | −2 | −4 |
| Net result | +93 | −97 | −4 |
The −4 USDT aggregate is precisely the two assumed 2 USDT fees. It is not evidence that an exchange “won” B’s 97 USDT loss. In the model, 100 USDT of price P&L moved from B to A, 5 USDT of funding moved from A to B, and 4 USDT of fees left the two-trader group. Real accounts can have different fee rates, partial fills, multiple counterparties, liquidation charges and other costs. The table explains the categories, not a universal settlement trace for every venue.
Next step: Inspect Phemex futures markets and current terms. Confirm the exact contract, funding schedule and fee tier before calculating a possible holding cost. Perpetuals can produce rapid losses and liquidation.
Does funding create profit for the market as a whole?
Funding is generally a transfer between opposite position holders. Phemex’s funding-rate explanation states that a positive rate generally means longs pay shorts; a negative rate generally means shorts pay longs. The rate is designed to help keep a perpetual contract price related to its reference market. A positive displayed rate is not a free coupon: a trader must hold an eligible position at the relevant time and still bears price, fee and liquidation risk.
In the simple table, funding changes who keeps how much, while its total remains zero. A trader who receives 5 USDT in funding can lose 100 USDT on the contract price move. Conversely, a profitable long can pay funding and still finish ahead. Quoting “funding income” without the offsetting hedge or price P&L therefore hides the actual economics of a position.
Contract-specific details matter. Funding interval, rate caps and the notional base used for the charge can differ across products or change under stated conditions. Phemex’s futures-funding guide explains its mechanism and notes that conditions can vary. Read the current market’s funding display and the transaction record; do not take an old article’s example rate as today’s payable amount.
Funding also has a timing boundary. If your reporting window starts just after a funding payment and ends before the next, it can show zero funding despite a longer trade that incurred payments. If a backtest includes one side’s funding but omits the other side’s corresponding receipt, its closed-system total will appear wrong. For an account-level ledger, count only funding actually posted to that account during the selected interval.
Why do trading fees make the combined trader result negative?
Trading fees are a cost paid for execution services. In the two-trader example, each pays 2 USDT, so the pair loses 4 USDT collectively even though their gross price P&L sums to zero. A trader must exceed their own trading fees, funding paid and other costs to finish with a positive net result. There is no logical conflict between a zero-sum gross price transfer and a negative-sum after-cost trader group.
Fees can vary with maker or taker execution, account tier, promotions and product terms. The public Phemex fees-and-conditions page should be checked together with the market interface and the actual trade record. Its displayed rates can depend on account settings and current conditions; this article therefore uses no live percentage in its example. The relevant arithmetic is: net account change = price P&L + funding received − funding paid − trading fees − other applicable charges.
Spread and slippage are also economic costs, even when they are not posted as separate fee rows. A marketable buy that fills above the price used in a model, and a later sell below it, leave less realized P&L. If a ledger uses the actual fill prices, the spread effect is already reflected in its price P&L. Subtracting a separate estimated spread again would double-count it. If the ledger uses mid-prices instead, it must account for the execution gap once.
The same caution applies to liquidation and insurance-related adjustments. Their exact treatment depends on the product and venue rules. An individual trader’s loss can exceed the planned stop or collateral estimate under adverse conditions. A neat two-row theoretical P&L equality should never be used as a substitute for the product’s risk disclosure or an actual account statement.
Is the exchange always the losing counterparty to a winning trader?
No such conclusion follows from the zero-sum model. An exchange may operate a matching system in which positions are opened and transferred among many participants, with market makers providing liquidity. The simplified two-trader example is not a claim that A and B were directly paired for their entire trade or that Phemex took the opposite side of each user’s position. Counterparty structure, clearing and risk rules require product-specific evidence.
It is also possible for two visible traders to both report gains over an interval because they entered at different prices, traded against other participants, or have different observation windows. Their gains do not falsify the closed-book accounting identity; they show that the two visible accounts are not the complete book. Similarly, two traders can both lose after fees or after separate decisions at different times. Before invoking “zero-sum,” specify who is included and what is being summed.
A hedge adds another boundary. Suppose a short perpetual holder also owns spot BTC. The short may lose on the derivative while the spot holding gains in market value. Looking only at the perpetual tells one story; looking at the entire portfolio tells another. The hedge may also incur borrowing, custody, funding or execution costs. “Delta-neutral” describes a target exposure, not a guaranteed positive return.
Why is buying spot crypto a different accounting question?
A spot holder owns an asset rather than a contract whose P&L is settled against an opposite derivative position. If the market price rises, the holder’s marked-to-market asset value rises. There need not be a contemporaneous, equal realized loss in another current spot holder’s account. When the holder sells, an actual buyer pays the transaction price and takes on future price risk. Value creation, new information and changing demand can affect the asset’s price over time.
This does not mean spot investing is risk-free or that all spot buyers can realize the displayed market capitalization. Selling a large amount can move the price, and an unrealized gain can disappear. The point is conceptual: a spot asset’s valuation process and a matched derivative P&L transfer are different accounting objects. Calling all crypto activity “zero-sum” because someone eventually sells to someone else confuses a trade with a full lifetime wealth ledger.
For an individual investor, the useful question is simpler than the slogan. What asset or contract do you hold? What are the cash flows? Which costs accrue while you hold it? What price can you actually trade at when exiting? Those answers are measurable and can be checked against account statements. The phrase “zero-sum” only helps when it makes the defined cash-flow boundary clearer.
How can you audit your own perpetual result?
Export or record the actual fills, including side, quantity, execution price and fee for each fill. Next, record every funding debit or credit during the holding period. Then include any realized P&L and remaining unrealized exposure at the same cutoff time. Reconcile the change in account equity with transfers, deposits and withdrawals. If the numbers do not reconcile, check whether you mixed mark-price unrealized P&L with realized fill-price P&L or omitted an open position.
Use a single quote currency for the ledger. A coin-margined contract can create additional conversion effects when collateral is valued in a different unit. A USDT-denominated example cannot be pasted unchanged into a coin-margined position. Specify the contract type and unit before comparing two traders or two strategies.
Finally, evaluate a strategy after costs and under adverse scenarios. A backtest that ignores funding or assumes every stop fills at its trigger can show an appealing gross result while the actual execution path is negative. The ledger does not tell you whether to trade. It tells you where to look for the money that a headline gross P&L omits.
FAQ
If a long gains, must one short lose exactly the same amount?
In a simplified matched book over the same prices and period, opposite gross price P&L offsets. An actual account’s net result also depends on funding, fees, fills and the wider set of participants.
Does receiving funding guarantee a profitable trade?
No. The contract price can move against the position by more than the funding received, and trading costs can reduce the result further.
Does a stop-loss make a derivative no longer zero-sum?
No. A stop changes when an individual position exits and at what executable price. It does not create value inside the simplified matched price P&L ledger.
Are exchange fees the same as funding?
No. Funding generally transfers between long and short holders. Trading fees are charged for execution under the venue’s fee schedule. Keep them in separate account rows.
