Test a funding-settlement trade after all costs

Funding eligibility and look-ahead bias from the settled rate

Read the venue's rules for which positions qualify at the settlement boundary. Do not assume a local clock, a submitted order or an acknowledgement proves the position was eligible. Fills, exchange timestamps and the resulting account funding record are the evidence needed to evaluate the event.

Store the rate estimate available when the decision was made, the final settled rate and any interval change. A historical test using the final rate to choose an earlier entry can look profitable because it knows information that the trader did not yet have. The same applies to using the best price during the settlement window after seeing all the trades.

Funding rate farming: net trade result across every eligible event

For an unhedged position, net result equals price P&L plus funding received minus funding paid minus trading and execution costs. An opposing hedge changes the price risk but adds another set of executions, basis exposure and collateral requirements. Evaluate the complete structure rather than comparing a funding percentage with zero.

Test all eligible events under fixed timing, fill and exclusion rules. Include changed rates, missed entries, unfilled exits, outages and negative outcomes. A parameter selected from the same events it is evaluated on is exploratory evidence, not validation. Use a later untouched period and keep the original decision-time data.

Example: 30 USDT funding on a 100,000 USDT short

Assume a hypothetical 100,000 USDT short is eligible to receive a positive 0.03% settlement: 30 USDT. Entry and exit each cost 0.04% in fees, totalling 80 USDT at an unchanged notional. Add 20 USDT of round-trip adverse execution. Even with zero price movement, net result is 30 − 80 − 20 = −70 USDT.

If the underlying rises 0.2% while the base quantity stays fixed, the short loses 200 USDT before execution costs. The reference exit notional becomes 100,200 USDT, so projected fees are 40 + 40.08 = 80.08 USDT. With the same 30 funding receipt and 20 execution allowance, the budgeted result is 30 − 80.08 − 20 − 200 = −270.08 USDT. The funding valuation is assumed to remain 100,000 USDT at settlement; the adverse move occurs afterward. Receiving funding does not establish an edge, and neither would a favourable price move alone.

This is a pre-trade cost budget: percentage fees use the reference entry/exit notionals, while the 20 USDT execution allowance is a separate stress deduction. An actual ledger must use filled quantities and prices, calculate fees on those actual fills and avoid charging their embedded spread/slippage twice. The budget does not include the small fee change caused by the execution allowance itself.

Mistake: entering seconds before settlement and missing eligibility

Trying to minimise exposure by entering extremely close to settlement increases sensitivity to latency, partial fills and the venue's processing window. The order can miss eligibility while still leaving a live position and trading costs. A realistic test must model or observe those failures; assigning every attempt a full receipt and instantaneous exit manufactures an advantage.