What is a basis trade (spot–futures basis)?

The basis trade and funding capture look like the same trade and are not. Both are delta-neutral; they run on different clocks and are paid by different mechanisms.

What “the basis” is

The basis is the gap between an asset’s spot price and the price of a dated future on that asset. Future above spot is contango (positive basis); below is backwardation. A future must equal spot on the day it expires, so a positive basis is a gap with a deadline: it must close.

The basis trade, or cash-and-carry, harvests that closure: buy the asset in spot, short an equal amount of the dated future. Price direction cancels — the same delta-neutral construction used throughout this research — and what is left is the convergence of the two prices as expiry approaches.

Basis trade vs funding capture: same shape, different clock

Both are long spot, short a derivative. The difference is what tethers the derivative to spot:

• A dated future is tethered by its expiry. No payment along the way; the price gap you entered at is the gross return, known at entry provided you hold to expiry.
• A perpetual is tethered by a payment — the funding rate, exchanged between longs and shorts every settlement (every eight hours on Binance). Nothing is locked in: the rate is re-quoted continuously and can flip sign.

A basis trade is a fixed-term position with a rate known at entry; funding capture is an open-ended position with a floating rate. One is a deadline, the other a stream.

When each one actually pays

• The basis trade pays when the dated future is at a premium wide enough to clear two-leg trading costs — on the contract’s schedule, not yours. At expiry you stop or roll into the next contract at whatever basis exists then. Entry timing dominates, because the rate is set once.
• Funding capture pays whenever funding is positive — usually, because leverage-hungry long demand is persistent. No expiry, no roll, no forced exit — but nothing guaranteed: when the market flips net-short, funding inverts and the hedged short pays rather than earns.

Both share the same unglamorous killer: two-leg trading costs. The edge is thin per period, so how often you touch the position decides whether anything survives.

Why Fygga’s harness measures funding specifically

A measurement reason, not a claim about which is better. Funding is published, per symbol, every settlement — one continuous public series a harness can read and a reader can re-derive. Fygga’s backtest does exactly that: Binance USDⓂ-M daily closes plus funding rates, five liquid pairs, 1,636 days across 2022–2026. A basis-trade record would instead require modelling a term structure of expiring contracts and every roll between them — assumptions a backtest is worst at being trusted about.

The numbers Fygga does have — for funding capture

Over 2022–2026 on real Binance history (1,636 days, five liquid pairs), gross funding accrual before costs was +19.3% — about ~4%/yr. That proves the edge exists; it is not achievable, because it charges nothing for trading. With realistic costs on both legs, the honest net at a weekly cadence was +2.8% total with a -5.3% maximum drawdown.

The cadence sweep carries to any delta-neutral trade, basis included. Rebalancing daily turned the same strategy into -19.1% — a cost drag of 8.7%/yr against an edge worth roughly ~4%/yr gross. Weekly costs 2.1%/yr; biweekly costs 1.1%/yr and returned +4.6%. Only how often the position was touched changed.

What a basis trade does not remove

• Roll risk. The basis at the next roll is unknown, so a “fixed rate” is really a chain of unknown future rates.
• Margin and mark-to-market risk. Convergence is certain at expiry, not monotone on the way there; a short leg moving against you can require margin long before the trade is right.
• Cost risk. Two legs, opened and closed, plus every roll — exactly what the daily-cadence run above did to a positive gross edge.
• Exchange and counterparty risk. The venue can fail, freeze withdrawals, change margin rules or auto-deleverage a position.

A basis trade is a lower-drama position than a directional bet, not a risk-free one. This is educational information, not financial advice.

Frequently asked questions

What is a basis trade in crypto?

A basis trade captures the price gap — the basis — between an asset's spot price and the price of a dated future on that same asset. When the future trades above spot (contango), you buy the asset in spot and short the dated future in equal size. The position is delta-neutral, so price direction is hedged away. As the future approaches expiry its price must converge to spot, and that convergence is the profit. The classic name for it is cash-and-carry. It is educational information, not a recommendation.

How is a basis trade different from funding-rate capture?

They share the same delta-neutral shape — long spot, short a derivative — but run on different clocks. A basis trade uses a dated future, and the return is set by the price gap you enter at: hold to expiry and the gross payoff is essentially known when you open. Funding capture uses a perpetual future, which never expires; instead a funding payment is exchanged between longs and shorts every settlement (every eight hours on Binance). That stream is not locked in — it is re-priced every settlement and can turn negative and cost you money.

When does each one actually pay?

A basis trade pays when a dated future trades at a premium to spot wide enough to survive the trading costs on both legs. The rate is fixed at entry and realised through convergence, so your capital is committed to a fixed term and you must either hold to expiry or roll into the next contract. Funding capture pays whenever the funding rate is positive — most of the time, because the crowd is usually net-long — and it pays continuously. There is no expiry, no roll and no forced exit, but no locked-in rate either: when the market flips net-short, funding inverts and the hedged short pays instead of earns.

Why does Fygga measure funding rather than the basis?

Because funding is directly and continuously observable, which makes it honestly measurable. Every perpetual publishes a funding rate every settlement, so the research harness reads one public series per symbol — Fygga's backtest uses Binance USDⓂ-M daily closes plus funding rates across five liquid pairs over 1,636 days. Reconstructing a basis-trade record would mean modelling a term structure of expiring contracts and every roll between them, where the result depends heavily on which contract you picked and when you rolled. That is a much more assumption-heavy research problem, and Fygga has not done it.

Does Fygga publish basis-trade returns?

No, and that is deliberate. Fygga has never backtested a basis trade, so no basis-trade performance figure exists in its data and none is shown anywhere on this site. Every published number describes delta-neutral funding capture. In the committed backtest, gross funding accrual over 2022–2026 was +19.3% (about 4% a year) before costs, and the honest net at a weekly rebalance cadence was +2.8% total with a -5.3% maximum drawdown. Those are historical funding-capture backtest figures, not basis-trade figures and not a promise of future returns. Past performance is not indicative of future results.