Quant Memo
Core

Bond Futures and the Cheapest to Deliver

A Treasury bond future doesn't settle into one specific bond — the short can deliver any bond from a basket, and rationally chooses whichever one is cheapest for them to deliver, which is the bond that actually drives the future's price.

Prerequisites: Bond Pricing and Accrued Interest, Futures vs Forwards

Most futures contracts settle against one crystal-clear thing — one stock, one index level. Treasury bond futures are different: the exchange defines a basket of eligible bonds that the seller (the short) is allowed to deliver at expiry, spanning a range of maturities and coupons. That flexibility is deliberate — it prevents any single bond from being squeezed into scarcity right before delivery — but it also means the future's price behaves like whichever bond the short will actually choose to hand over.

Because deliverable bonds differ in coupon and maturity, the exchange assigns each one a conversion factor to roughly equalize what the short receives regardless of which bond they deliver. In practice the equalization is imperfect, so one bond is always the cheapest for the short to deliver — the cheapest-to-deliver, or CTD — and it is that bond, not some average of the basket, that the future effectively tracks.

Why one bond always wins

At delivery, the short receives the futures settlement price times the delivered bond's conversion factor, plus accrued interest. The short's profit from delivering any given bond is:

Profit=(Futures price×CF)Bond’s market price\text{Profit} = (\text{Futures price} \times \text{CF}) - \text{Bond's market price}

In words: what the short is paid for that bond, minus what it actually cost them to buy it. Conversion factors are fixed by the exchange using a standardized yield assumption, but real bonds don't all trade at that assumed yield, so the profit differs bond by bond. The bond with the highest profit (or lowest net cost) is the CTD, and rational shorts always deliver it.

bond A bond B (CTD) bond C net cost to deliver
Bond B costs the short the least to source and deliver against the future — it's the CTD, and its price behavior is what the futures price actually follows.

Worked example

The futures price is 118-00 (118.00). Bond A has a conversion factor of 0.92 and trades at 108.75. Bond B has a conversion factor of 0.98 and trades at 115.60.

Implied delivery value of A: 118.00×0.92=108.56118.00 \times 0.92 = 108.56; cost to buy A is 108.75, so delivering A loses 108.56108.75=0.19108.56 - 108.75 = -0.19.

Implied delivery value of B: 118.00×0.98=115.64118.00 \times 0.98 = 115.64; cost to buy B is 115.60, so delivering B gains 115.64115.60=0.04115.64 - 115.60 = 0.04.

Bond B is cheaper to deliver — it's the one shorts will source and hand over, so the futures price will move in line with Bond B's price, not Bond A's.

What this means in practice

The identity of the CTD can change as yields move, because conversion factors only match bonds correctly at one specific yield level; away from that level, a different bond in the basket becomes the cheapest. Relative-value desks track which bond is CTD and how close a "flip" to a different CTD is, because that shift changes the future's effective duration overnight.

Beginners often assume a Treasury future tracks "the 10-year bond" generically. It tracks one specific bond at a time — the CTD — and that identity can and does change, sometimes abruptly near contract expiry.

Related concepts

Practice in interviews

Further reading

  • Burghardt et al., The Treasury Bond Basis
ShareTwitterLinkedIn