Skip to content
Convergence and basis decay

A futures price converges to spot at delivery — the basis decays to zero by expiry. Only the endpoint is enforced; the path is noisy.

Try this
PriceF = SFutures FSpot SBasis (F − S)00t = 0y1y (delivery)Time toward delivery
Contango. r > q, so F starts 4.08 above spot. The positive basis decays down to zero by delivery.
Futures Fₜ 104.08Spot Sₜ 100.00Basis Fₜ − Sₜ +4.08Time to delivery (T − t) 1y
Sweep time t toward delivery0y
Spot S₀100.00
Financing rate r6%
Carry / dividend yield q2%
Horizon to delivery T1y
Fₜ = Sₜ · e^((r − q)·(T − t))  ⇒  basisₜ = Sₜ · (e^((r − q)·(T − t)) − 1) → 0 as t → TThe factor (T − t) shrinks to zero at delivery, so the exponential collapses to 1 and the basis vanishes — no matter where the spot wandered. No-arbitrage only pins down the endpoint Fₜ = Sₜ at delivery; in between, the basis is whatever carry and noise make it.
Discuss. Turn on the bumpy path and raise σ until the spot swings wildly. The basis still hits exactly zero at delivery. Why is the endpoint guaranteed by no-arbitrage but the path is not? What trade would you put on if the basis stayed wide one day before delivery?
Convergence of Futures to SpotOpen in Dr Phil's Quant Lab ↗