Decentralised News · 2027 planning edition · Evidence reviewed 22 November 2026
What Matters
An entry reset can change where P&L appears without changing the position’s quantity or total gross result. Preserve original fills, session marks and settled components. Reconcile them over the same period before adding costs. Verify the contract’s actual settlement interval; current documentation does not support a universal eight-hour assumption.
Evidence and method
Evidence checked on 22 November 2026 for this 2027 planning edition. This is a documentary analysis of Bybit’s USDC perpetual mechanism, supported by its settlement guide, September 2026 FAQ, entry-price and P&L explanations, and API field definitions. The FAQ and dedicated mechanism describe contract-specific intervals, while some other pages retain eight-hour wording; we flag that conflict and require instrument-level verification. The original examples and local calculator use synthetic quantities and marks, with explicit cost signs and a constant-size assumption. We did not fund an account, place trades or observe settlement on a live position. The arithmetic checks attribution, not platform solvency, available withdrawals or margin adequacy.
DN Alpha Thesis
When a USDC perpetual entry changes without a matching fill, reconcile the session boundary before treating it as a trading event. The useful control is a bridge from original price P&L through each settled component to the current unrealized component. For an unchanged linear quantity, those components must match the direct endpoint calculation before costs. A report that keeps only the latest session loses history; one that adds the original-entry gain to all settlements counts history twice. Map the source fields and funding signs, check the actual contract schedule and leave unknown costs unresolved. Use a separate event ledger whenever quantity changes.
USDC Session P&L Bridge
Explore an original two-settlement example for one unchanged linear position. All prices are USDC per base unit. The model reconciles gross price P&L and adds only the costs you declare; it does not query an account or verify an actual event.
Scenario method and limits
The initial inputs and results are an original, synthetic DN example. They are provided to make the method reproducible, not as live prices, measured performance or a prediction for 2027. Change the inputs and download your own scenario above.
- Two settlement boundaries, constant quantity and direction, linear base-asset units only. Additions, partial exits, inverse products and contract multipliers require a different ledger.
- Synthetic values are not live marks or current venue charges. Arithmetic agreement does not verify the inputs or actual account.
- Fees and funding use paid-positive signs; known costs must cover the same period. The tool deliberately leaves net P&L unknown when costs are incomplete.
- No forecast closing fees, account equity, withdrawable amount, collateral availability or liquidation calculation. Rounded display uses two decimal places; retain raw venue precision in a real reconciliation.
The trade can stay open while its accounting basis moves
A trader opens a USDC perpetual long, leaves its quantity untouched and returns to find a different average entry price. The unrealized profit has also fallen sharply, perhaps to zero. Those two changes can look like an unauthorized trade or missing profit. The first question is whether a scheduled session settlement occurred between the snapshots.
Bybit provides a concrete documented example of this mechanism. Its USDC session settlement guide says the position’s session P&L is realized and its entry basis becomes the settlement mark. That is an accounting event; it does not, by itself, require an opening or closing order. This article examines that design, not a universal rule for every exchange using USDC.
There are two useful questions after a reset. What has the position earned across the entire period being evaluated? What has it earned since the latest settlement? Both answers can be correct and different. A dashboard becomes misleading when it displays the second answer under a label the reader interprets as the first.
Keep exposure separate from those reporting periods. In a simple linear position that remains 0.5 BTC long, another 1,000 USDC increase in the BTC mark produces another 500 USDC of gross P&L. Moving the accounting entry from one mark to another does not change that sensitivity. Quantity, direction and the contract’s units determine it.
The settlement clock is contract specific
Do not build a 2027 monitoring rule around an assumed eight-hour reset. Bybit’s USDC contract FAQ, updated 8 September 2026, describes intervals of one, two, four or eight hours and says settlement follows the contract’s funding interval. The dedicated mechanism page, updated 19 March 2026, gives the same variable-interval explanation.
The documentation is not fully consistent. The P&L calculation page, updated 5 August 2026, still uses an eight-hour statement, and some API descriptions retain eight-hour wording. We give precedence to the specific interval explanation for planning, then require verification against the actual instrument and recorded events. An example using eight hours does not establish the schedule of every contract.
The instruments-info API documentation identifies a fundingInterval field measured in minutes. Check the exact symbol, settlement currency and current field, then preserve its retrieval time. That field describes the interval; it does not retrospectively prove what schedule applied to an older ledger entry. Historical reconciliation needs the historical event timestamps as well.
A practical consequence follows. If a monitor expects three resets per day while the relevant instrument moves to a shorter interval, it may misclassify legitimate resets as unexplained entry changes. Fix the schedule assumption before inventing compensating trades. Conversely, a time that looks plausible is only a clue: confirm the matching accounting event and unchanged quantity.
An original two-session example: 1,500 USDC did not disappear
Assume a synthetic linear long of 0.5 BTC, initially entered at 100,000 USDC per BTC. There are no additions, reductions, transfers or other positions in this calculation. The first settlement mark is 102,000; the second is 101,000; the current mark is 103,000. These values are illustrative inputs, not market observations or a trade we placed.
| Point in the example | Calculation | Gross P&L attributed | Basis for the next interval |
|---|---|---|---|
| First settlement | 0.5 × (102,000 − 100,000) | +1,000 USDC settled | 102,000 |
| Second settlement | 0.5 × (101,000 − 102,000) | −500 USDC settled | 101,000 |
| Current open session | 0.5 × (103,000 − 101,000) | +1,000 USDC unrealized | 101,000 remains the session basis |
| Entire illustrated period | 1,000 − 500 + 1,000 | +1,500 USDC gross | Original basis retained for attribution |
The direct check is 0.5 × (103,000 − 100,000) = 1,500 USDC. The settled total of 500 plus the current unrealized 1,000 reaches the same result. Looking only at the last session misses 500. Adding the full original-entry gain to the settled 500 counts that earlier gain twice.
The cancellation of intermediate prices explains the reconciliation: q × [(S1 − E0) + (S2 − S1) + (P − S2)] = q × (P − E0). The settlement marks cancel algebraically. They determine when P&L is recorded and how it is divided, while the unchanged quantity and endpoints determine this simplified gross total.
Immediately after the second settlement, imagine the mark is still exactly 101,000. The new session has zero unrealized P&L, yet the illustrated holding has earned 500 gross. A zero on the session display is therefore consistent with an earlier gain. It is also consistent with an earlier loss in a different price path. Zero alone says nothing about the full holding period.
For a 0.5 BTC short on the same path, every price component reverses: −1,000, +500 and −1,000, totaling −1,500. Trading costs do not automatically reverse with direction. A short that paid the same 30 USDC of fees and 20 of funding would finish at −1,550, while the long would show +1,450 after those assumed costs.
Why the website and an API can show different entries
A field called average entry is not enough to identify its accounting purpose. Bybit’s position API distinguishes avgPrice from sessionAvgPrice. Its documentation says the former is not reset by session settlement for USDC contracts, while the latter corresponds to the entry shown in the web interface. Its average-entry guide explicitly directs USDC users to the session field when matching that display.
In the unchanged-position example, an exporter retaining the original basis and a screen displaying 101,000 can therefore describe the same exposure. The disagreement is meaningful only after comparing field definitions, timestamps and position changes. Replacing one value with the other merely to make a spreadsheet agree can destroy the evidence needed to explain the difference.
| Question being answered | Record to retain | Common analytical mistake |
|---|---|---|
| Where did this holding period begin? | Opening fills and original cost attribution | Overwrite the history at every reset |
| What basis starts the current session? | Session entry and settlement mark | Treat the reset as a new discretionary trade |
| What happened to prior session P&L? | Dated settlement entries | Omit them because the position is still open |
| Did actual exposure change? | Quantity, side and executions | Infer size changes from entry-price changes |
| What is the result after known costs? | Scoped gross components and signed costs | Subtract costs already included in a net field |
Store explicit names in your own report: original-fill basis, current-session basis, cumulative settled price P&L and current-session unrealized P&L. Those labels remain intelligible when software or column headings change. Also scope every total to the same start time. A lifetime realized total cannot be added to the unrealized value of one newly selected reporting period without adjustment.
Funding can arrive in the same row without being the same mechanism
Bybit’s transaction-log documentation says USDC funding and session settlement can share a record. It separates price-related cashFlow from funding and trading fee. Its balance-change relation is change = cashFlow + funding − fee. In that endpoint, positive funding means received, while a positive fee means paid; negative fees represent rebates.
Consider an illustrative settlement row with 1,000 USDC of price P&L, funding of −12 and zero trading fee. Its change is +988. Calling the full 988 a funding receipt misidentifies the economic driver. Calling the whole 1,000 net profit ignores the funding cost. A shared timestamp is a reason to separate components carefully, not a reason to merge their meanings.
The calculator uses paid-positive signs for both costs. A funding receipt of 12 maps to −12; a trading rebate is also negative. Map the source convention before copying values.
For the two-session example, 1,500 gross minus 30 net trading fees minus 20 net funding paid equals 1,450 USDC. With 20 received instead, the input is −20 and the result is 1,490. These are assumed costs over the illustrated period. They are not current fee quotes, a predicted funding bill or a provision for closing an order that has not happened.
Bybit’s P&L guide includes trading and funding fees in its realized-P&L description. Therefore, identify whether an imported figure is gross or already net before subtracting anything. The calculator builds gross settlement components from prices. Feeding it an already-net realized total as though it were a gross price component would mix definitions. For a detailed funding bill, use our funding reconciliation guide.
Adding or closing quantity breaks the constant-size shortcut
The telescoping example is intentionally narrow. A position that changes size needs an event ledger. The average-entry documentation describes the session basis as weighted by position additions. Applying the final size to the original price difference would attribute earlier gains to units that did not yet exist.
Here is a separate synthetic illustration. Carry 0.5 BTC from an original 100,000 entry through a settlement at 102,000, realizing 1,000. Then add 0.25 BTC at 104,000. The current-session weighted basis becomes [(0.5 × 102,000) + (0.25 × 104,000)] ÷ 0.75 = 102,666.6667. At a mark of 103,000, the current-session gain is approximately 250. Add the earlier 1,000 and the gross holding result is 1,250.
Independently, the original 0.5 BTC gained 1,500 from 100,000 to 103,000, while the added 0.25 BTC lost 250 from 104,000 to 103,000. Their sum is the same 1,250. Using the final 0.75 BTC against the original 100,000 entry would instead produce 2,250 and overstate the result by 1,000.
Partial exits require the same care. Return to the original 0.5 BTC example after its first 102,000 settlement. Close 0.2 BTC at 101,500, creating −100 of post-settlement price P&L. If the remaining 0.3 BTC is marked at 103,000, its current-session gain is 300. Earlier settled P&L of 1,000 plus −100 plus 300 equals 1,200 before costs. The previous settlement must remain in the history even though some units have closed.
These calculations assume one linear contract and a consistently applied session cost basis. They do not settle a platform-specific ledger dispute or reconstruct every position-mode rule. When size changes, partition the history at actual executions and settlement events, retaining quantities at each boundary. The local tool in this guide deliberately accepts a constant quantity; it cannot validate this variable-size extension.
A reset does not promise more spendable collateral
The P&L identity is not a margin engine. At an unchanged mark, moving a price gain between unrealized and settled components conserves their sum in the simplified example. It does not establish the account’s transferable amount, margin requirement or liquidation distance. Those depend on account mode and other balances and exposures that the example does not contain.
Bybit’s liquidation documentation, updated 15 October 2026, separately discusses USDC isolated-margin adjustments after settlement, including the basis used for maintenance margin and closing-fee calculations. That is a concrete reason to avoid treating an entry reset as a fresh position with freshly deposited margin.
A useful diagnostic separates three observations. The position remains open; the accounting basis changes; the risk engine may update its own figures. None is a substitute for the others. A trader who sees realized profit and increases exposure without checking the current account state has made a new risk decision. The settlement calculation alone cannot justify it.
The same applies to a hedge spanning two venues. If one venue restates session P&L and the other keeps a lifetime position display, compare total scoped components before concluding the hedge lost money at the boundary. Then check actual size, prices, basis differences and costs. Reporting conventions can explain a display mismatch; they cannot prove the exposures offset.
A practical reconciliation when the entry changes unexpectedly
- Pin down the product. Record the exact symbol, settlement asset, account mode, side and units. A ticker containing USDC is insufficient evidence of this settlement design.
- Preserve both snapshots. Keep the pre-event and post-event entry, mark, quantity and timestamps. Do not overwrite the earlier record to force visual agreement.
- Look for the event. Match the settlement record and instrument schedule. Check executions over the same window because a legitimate reset and a separate fill can occur close together.
- Rebuild price attribution. Use quantity actually held through each interval. Add settled price components to the current unrealized component without counting original-entry P&L again.
- Normalize signed costs. Map each source field to paid or received, retain the original signed values and confirm whether a reported total already includes them.
- Check the residual. Compare reconstructed and reported values over identical boundaries. Investigate missing events, different marks, changing sizes, rounding and scope before labeling the difference a loss.
A comparison window needs more than a date. Record whether its first snapshot is before or after settlement, whether its last mark is the same valuation used by the displayed unrealized field, and whether all pages of the exported history were retained. Missing the first settled component can produce an internally tidy spreadsheet with an incorrect total.
Keep known and unknown separate. If prices and quantity reconcile but a funding record is absent, report the gross attribution and leave the net figure unresolved. If the settlement mark itself is missing, do not substitute the nearest candle close and call the result verified. That approximation can be useful as an estimate, provided it stays labeled as one.
For a broader audit of a trading system, our trade-log P&L audit covers the full performance problem. The decision here is narrower: determine whether an entry reset is an accounted-for session boundary, then preserve the economic history across it. A correct reconciliation explains where each component went without inventing a trade or treating a new display basis as a new investment.
Frequently asked questions
Does a session settlement close my USDC perpetual?
In the documented mechanism examined here, the session entry is restated and price P&L is settled while the position can remain open. Confirm actual executions and quantity before concluding a trade occurred. An unchanged 0.5 BTC linear position retains the same price sensitivity after the accounting reset.
Are all USDC perpetuals settled every eight hours?
No universal interval is supported by the evidence here. Bybit’s current specific mechanism and FAQ describe contract-specific schedules linked to funding, including one, two, four and eight hours. Some other documentation retains eight-hour language. Verify the exact instrument and recorded event times, and do not extend this design to every USDC venue.
Why can the API entry differ from the website?
Bybit documents separate avgPrice and sessionAvgPrice fields for USDC contracts. The session field corresponds to the web entry display and resets at settlement. Compare field definitions and timestamps before labeling the difference an error; also check for additions or reductions during the comparison window.
How should I handle a funding payment in the settlement row?
Separate price P&L from costs, then map each sign. In this calculator, positive means paid and negative means received. Consult the ledger definitions before importing values, and avoid deducting costs already included in a reported total.
Can the calculator tell me how much I can withdraw?
No. It models gross P&L attribution and declared costs for one unchanged linear position across two settlements. It does not model the rest of an account, collateral rules, open orders, liquidation or transfer restrictions. If quantity changed or required records are missing, use a fuller event ledger and retain the uncertainty.
Sources and verification
Primary documentation checked on 22 November 2026. Exchange rules, regional availability, fees and API behavior can change; recheck the exact product and account before using this 2027 planning guide.
- Bybit: USDC Perpetual Session Settlement Mechanism — Session realization and reset to settlement mark; contract-specific interval follows funding. Updated 19 March 2026.
- Bybit: FAQ — USDC Perpetual Contracts — Product scope, no-order entry reset, variable settlement intervals. Updated 8 September 2026.
- Bybit: P&L Calculations (USDC Perpetual Contracts) — Linear long/short P&L and realized components; eight-hour wording conflicts with specific mechanism. Updated 5 August 2026.
- Bybit: Average Entry Price (Perpetual and Expiry Contracts) — Weighted session entry and sessionAvgPrice guidance. Updated 19 August 2026.
- Bybit V5 API: Get Position Info — avgPrice and sessionAvgPrice distinction; web UI maps to sessionAvgPrice.
- Bybit V5 API: Get Instruments Info — fundingInterval specified in minutes and symbol-level instrument fields.
- Bybit V5 API: Get Transaction Log — Combined USDC settlement record; cashFlow, funding, fee and change definitions and signs.
- Bybit: Trading Rules — Liquidation Process (Unified Trading Account) — USDC isolated margin settlement adjustments are separate from P&L attribution. Updated 15 October 2026.




