One-way vs hedge mode: what does your futures order change?

One-way mode keeps a single net position in a futures contract, so an opposite trade reduces, closes or reverses that position. Hedge mode can keep separate long and short sides in the same contract. Choose the mode that matches how your strategy owns and closes exposure, then verify the exact account, product and order-side rules before enabling automation.

A sell instruction does not, by itself, mean “open a short.” Its effect depends on the existing position, the selected position side and the venue's rules. That distinction matters when two strategies share a contract and each expects its own position to remain intact.

One-way versus hedge mode changes the position book

In one-way mode, a contract has one net direction. Buy more of an existing long and its size increases. Sell against it and the long becomes smaller; an ordinary sell larger than that long can reverse the remaining exposure if the venue accepts it. A valid reduction constraint changes that last behavior.

In hedge mode, the same contract can contain both a long side and a short side. A buy can add to the long or close the short. A sell can add to the short or close the long. You must identify the intended side as well as the trade direction. Binance's position-mode explanation documents these distinct position structures.

Neither mode identifies which bot owns a fill. Two strategies buying the same long side can still contribute to one aggregated long position. Hedge mode separates directions; it does not automatically create a separate position for every strategy, entry or time horizon.

Follow identical orders through two books

Consider two hypothetical linear BTC futures books. One uses one-way mode; the other uses hedge mode. Both start flat. Assume complete fills, quantities measured in BTC and no commission, funding, slippage or liquidation. These assumptions isolate position accounting rather than forecast a trading result.

Quantity units remain a separate choice from position mode. The linear-versus-inverse contract comparison explains why an inverse face amount cannot simply replace the BTC quantity in this ledger.

A longer-horizon strategy buys 0.060 BTC at 60,000 USDT/BTC. A shorter-horizon strategy later wants a separate short and sells 0.025 BTC at 61,000. In the hedge book, that second instruction is explicitly routed to open the short side. In the one-way book, it is an ordinary opposite-direction sell.

The one-way sell reduces the original long by 0.025 BTC. Its hypothetical realized price PnL is 0.025 × (61,000 − 60,000) = 25 USDT. There is no separate short for the second strategy to manage.

Now value both books at a hypothetical common price of 62,000:

• One-way: the remaining long has 0.035 × (62,000 − 60,000) = 70 USDT unrealized price PnL. Add 25 realized and the combined result is 95 USDT.
• Hedge: the long has 0.060 × 2,000 = 120 USDT unrealized price PnL, while the short has 0.025 × (61,000 − 62,000) = −25 USDT. The combined result is also 95 USDT.

The matching total is a consequence of the identical executions and valuation assumptions. Hedge mode did not create an extra source of profit. It preserved two sides and their separate entry records. Gross open quantity also differs; neither that quantity nor net exposure alone is the exchange's margin or liquidation calculation.

The operational question is who may close what next. If the short strategy sends a buy to close its supposed 0.025 BTC short in the one-way book, that ordinary buy adds to the remaining long. Its internal “short closed” label describes a position the exchange never held separately.

Write the four intents before mapping API fields

This conceptual matrix helps distinguish direction from purpose. It is not a complete order request or a promise that every account supports each action.

The field names differ. Binance USDⓈ-M Futures documents positionSide=BOTH for one-way mode and LONG or SHORT for hedge mode, with the side required in hedge mode. Its new-order API also says reduceOnly cannot be sent in hedge mode. Those requirements must be interpreted together; a closing intention is not implemented by copying a flag from another venue. Binance USDⓈ-M Futures trade API.

Bybit's order API uses positionIdx: 0 for one-way, 1 for the hedge buy side and 2 for the hedge sell side. It requires that index in hedge mode and documents reduceOnly=true for reducing or closing a position. Bybit place-order documentation.

Read the reduce-only order guide for stale exit quantities and competing closes. Position mode answers how positions are represented; an exit constraint answers which exposure changes an order may make.

Verify the scope of a mode change

A realistic setup failure is changing a default and assuming every symbol changed. Bybit documents configuration priority as symbol, then settlement coin, then system default. A coin-level change applies to eligible symbols without positions or open orders; a symbol with existing activity can retain its previous setting. Bybit switch-position-mode documentation.

Binance's cited change-position endpoint operates across symbols. Its current documentation also describes a shared USDⓈ-M/COIN-M setting after the relevant migration and rejects changes when either side has open orders or positions. Treat that as the scope of the documented account configuration, not a universal rule for other exchanges.

Before a planned change, stop new strategy submissions, inventory positions and all relevant open or conditional orders, and decide how existing commitments will be managed. Do not cancel protection or close exposure solely to make a settings request succeed. After any permitted change, read back the actual setting and position records for each instrument that automation will use.

Use a preflight worksheet for each strategy

Test each of the four intents in a suitable non-live environment, including a close after a partial fill. Record resulting positions and fills, rather than accepting an order acknowledgement as proof. The testnet-versus-live guide explains what that test can and cannot establish.

Give those observations timestamps too. The trading latency budget distinguishes decision, submission, acknowledgement and observed execution; a fast response cannot make an old position snapshot current.

If strategies need independent capital ownership, investigate an actual account boundary. The subaccount-isolation guide explains why separate credentials or strategy labels alone do not establish it.

Equal long and short quantities still need management

Equal opposite quantities in the same linear contract can offset price sensitivity under a simplified common-price model. Entry costs, exit costs, funding treatment, collateral and liquidation rules still require review. Closing one side also restores the other side's directional exposure. Use the perpetual-funding ledger to evaluate payments at their actual settlement times instead of assuming two sides cancel every cost.

KLINEO Editorial prepared this educational guide with AI assistance and reviewed the cited primary documentation on 23 October 2026. The examples are original hypothetical accounting exercises, not live tests or performance evidence. Venue capabilities described here do not establish support for those settings or order combinations in KLINEO.