A crypto futures roll moves exposure beyond expiry by closing the dated position and opening a later contract. For a long roll, compare the replacement buy price with the expiring sell price, then report the calendar price gap separately from execution friction and fees. Match the intended exposure and verify both legs; an accepted replacement order does not prove the roll finished.
The important distinction is between paying for execution and changing the price basis of a continuing position. Calling both simply “roll cost” can obscure what left the account, what changed future profit and loss, and whether the intended exposure actually survived.
Define what must continue after expiry
A roll is a new trading instruction, not a change to the original contract's expiry date. Identify the exact outgoing and replacement instruments, last trading times, settlement currencies and quantity definitions. Record deadlines in UTC, including any earlier broker cutoff. Do not infer a deadline from a similar ticker or from the date displayed in a local timezone.
| Existing position | Outgoing leg | Replacement leg | Intended final state |
|---|---|---|---|
| Long the near expiry | Sell to close | Buy the later expiry | No near position; intended later long |
| Short the near expiry | Buy to close | Sell the later expiry | No near position; intended later short |
Match the economic quantity you intend to retain. For comparable linear BTC contracts, preserving 0.50 BTC of price exposure can require different contract counts if their multipliers differ. The contract multiplier and order-size worksheet explains that conversion. Matching dollar notionals instead can change the BTC quantity because the two expiries have different prices.
Final expiry and interim PnL bookings are separate events. Use the settlement versus mark-to-market guide for that accounting distinction; this article follows the exposure transfer and its execution economics.
Read the spread direction before sending the order
Define the calendar price gap as later price − near price, using the same quote unit. CME's cryptocurrency-futures convention makes buying its calendar spread equivalent to buying the deferred contract and selling the nearby contract. Its quoted spread is deferred minus nearby. These are explicit CME conventions, so check the leg preview rather than assuming another venue's symbol order means the same thing. CME: cryptocurrency calendar-spread direction and pricing.
When later contracts trade above nearer ones, that segment of the curve is in contango; the reverse is backwardation. A positive gap means a long roll buys at a higher price than it sells. A short buys back the near contract and sells the later one, so the same gap gives a higher replacement sale price. Explicit costs still reduce its result. Neither observation establishes the eventual trading return.
Synchronize the quote observations. Subtracting an old near-contract last trade from a recent later-contract quote can manufacture an apparent calendar gap. Record whether the comparison uses midpoints, executable bid and ask prices, or actual fills. They answer different questions.
A two-leg roll-cost example
Consider an original hypothetical roll of a 0.50-BTC-equivalent long between two otherwise comparable linear, USD-settled futures. The objective is to preserve BTC quantity, not dollar notional. Assume sufficient margin and complete fills. All prices, fees and instrument terms below are invented, rather than a current trading opportunity or venue fee quotation.
At decision time, the near midpoint is $60,000 per BTC and the later midpoint is $60,600. The near sale subsequently fills at $59,990, while the later purchase fills at $60,620. Assume each separate leg pays a made-up commission of 0.04% of its executed notional in USD.
| Leg | BTC equivalent | Decision midpoint | Fill price | Adverse price difference | Commission |
|---|---|---|---|---|---|
| Sell near to close | 0.50 | $60,000 | $59,990 | $5.00 | $11.998 |
| Buy later to open | 0.50 | $60,600 | $60,620 | $10.00 | $12.124 |
| Total | — | — | — | $15.00 | $24.122 |
The near sale's shortfall is 0.50 × (60,000 − 59,990) = $5. The later purchase's shortfall is 0.50 × (60,620 − 60,600) = $10. Each is measured against its own decision midpoint.
Keep three results visible:
- The benchmark maturity gap is
0.50 × (60,600 − 60,000) = $300. - The executed maturity gap is
0.50 × (60,620 − 59,990) = $315. - Implementation friction is
$15 + $24.122 = $39.122: adverse execution against the benchmark plus explicit commissions.
The executed gap exceeds the benchmark gap by $15, exactly the combined price shortfall. Implementation friction is approximately 13.041 basis points of the $30,000 near-contract decision notional: 39.122 ÷ 30,000 × 10,000. Keep that denominator fixed when comparing alternative execution methods.
The midpoint differences already include the price consequences of spread crossing, depth consumption and price movement during execution. Adding another estimated bid-ask charge to those differences would overlap costs. They also do not identify how much movement the order itself caused. The transaction-cost analysis guide covers those attribution and benchmark limits.
The outgoing contract's profit since its original entry belongs in its position ledger. It cannot be inferred from this $315 gap between two different contracts. For inverse or differently settled instruments, use their actual payoff formulas and convert fees under a recorded currency policy; this linear worksheet cannot be copied unchanged.
The calendar gap is not an upfront purchase bill
Futures do not generally require paying the entire quoted notional to acquire the position. The CFTC explains that futures margin is a performance bond rather than a purchase down payment. CFTC: how futures margin works. The contract's rules govern subsequent PnL cash flows. Therefore, the example's $315 executed gap is not, by itself, an immediate $315 commission or cash debit. It records the difference between the old exit price and the new entry price.
That difference still matters economically. Suppose, as a separate hypothetical endpoint, the replacement contract later closes at $60,000. Its gross price PnL is 0.50 × (60,000 − 60,620) = −$310, before its closing fee. A higher replacement entry creates a less favorable long outcome for that fixed exit.
Do not deduct the $315 gap again after calculating PnL from actual entries, exits and settlement records. That would mix a price-basis comparison with the actual position result. Nor does contango guarantee a loss: the eventual underlying price, remaining basis and exit timing can change the result.
Perpetual funding is another ledger altogether. A calendar gap between dated contracts is not a periodic funding payment; the basis versus funding comparison explains why a price relationship and a cash-flow rate need separate treatment.
Choose an execution method with a completion rule
Separate outright orders create an interval during which only one leg may have traded. Closing first can leave less exposure than intended; opening first can temporarily increase exposure and margin demand. Neither sequence is universally preferable. Set a maximum unmatched quantity, a time limit and a response to a failed second leg before execution.
A venue-native spread order can bind the executions together. Deribit documents a single futures-spread order that buys one future and sells another simultaneously. Its individual leg prices are allocated using the spread execution and mark prices; they need not look like independently chosen outright fills. Its IOC setting can cancel unfilled spread quantity. Deribit: futures-spread execution and order instructions.
This can address legging risk for matched executions, but an accepted order can still remain incomplete. In a hypothetical matched 0.30-BTC roll out of the intended 0.50, the account retains 0.20 near and gains 0.30 later. Combined linear BTC quantity remains 0.50, yet 0.20 still faces the original expiry. “Exposure unchanged” is therefore insufficient evidence of a completed roll.
Fees also depend on the route. Deribit's published futures-combo schedule applies a 100% fee discount to the cheaper taker leg and reduces maker rebates to zero, while block-traded spreads use a different model. This is a venue-specific reason to compare the applicable spread fee with two outright orders, rather than doubling a familiar rate. Deribit: futures-spread fee treatment.
Validate the roll before the deadline
Use a record that follows quantities as well as prices:
| Checkpoint | Evidence required | Failure it catches |
|---|---|---|
| Before submission | Exact symbols, exposure target, deadlines, quantity steps and margin state | Wrong expiry or multiplier; insufficient temporary capacity |
| During execution | Distinct fills for both legs, cumulative quantities and active remainders | Partial transfer hidden by an accepted status |
| After cancellation or timeout | Authoritative final orders and positions | A late fill or replacement that duplicates exposure |
| Before outgoing expiry | Near position zero, later position at target, residual orders resolved | Unrolled exposure or a stale order reopening it |
| Cost review | Synchronized benchmarks, actual fees and native currency amounts | A misleading gap or double-counted cost |
Treat “near position zero” as a roll success only when the replacement is also verified. Expiry can remove the near position without creating any later exposure. An auto-roll label likewise needs documented scope: supported instruments, schedule, sizing and failure behavior.
Choose an internal deadline early enough to investigate rejected or incomplete orders. There is no universal best number of days before expiry: compare live depth for the intended size, the spread market, account constraints and time remaining. A cheaper displayed gap is not useful if the roll cannot complete within its limits.
Margin risk continues through the process. The insurance-fund and bankruptcy-price explanation describes what sits behind a liquidation loss; an insurance fund does not reserve capacity for the second leg or guarantee uninterrupted exposure.
Finish with a receipt stating the outgoing quantity closed, replacement quantity opened, unresolved remainders, benchmark and executed gaps, and explicit fees. If any quantity is unknown, resolve that state before calling the roll complete. That receipt makes the next expiry decision reproducible.