Skip to content
All library documents

Bond Forward Pricing and Treasury Futures DV01

Article Quant Q&A · Author: forwardPrice

Summary

The discussion compares two ways to value a bond for a future date: projecting a spot bond price using financing carry, or discounting its cash flows to obtain a price settled on that date. In theory, the approaches should agree when their financing assumptions and market conventions match. In practice, a curve built from short-term Treasury rates can imply different financing from a calculation based on repo rates, and repo markets may be illiquid at longer terms. Carry conventions and the choice of discount curve can also create differences.

For Treasury futures DV01, the responses describe shifting yields across the delivery basket while holding repo rates fixed, then repricing the future with a model that accounts for delivery-option value. Yield shifts may be applied to spot or forward yields according to desk preference. Holding yields fixed while shifting repo rates measures financing sensitivity, often called rho. The material is conceptual guidance rather than a worked calculation; it notes that bond-market discount curves are not uniquely determined.

Key ideas

  • Forward valuation and future-settlement cash-flow pricing should agree when financing assumptions and conventions align.
  • A Treasury curve and a repo-based carry calculation can produce different forward prices.
  • Futures DV01 can be estimated by shifting basket yields, keeping repo rates fixed, and repricing with a futures model.
  • Shifting repo rates while holding yields steady measures financing sensitivity.

Tags

Full text
# "Forward price of bond" VS "Price of a bond with a future settlement date"


# "Forward price of bond" VS "Price of a bond with a future settlement date"












What is the difference between

1) computing the 'forward price' of a bond at a future time T. ( spot price - carry, involving repo rates)

2) computing the price of a bond (discounting all cash flows) with a settlement date on T.

And if I were to compute the DV01 of a Treasury future, are both of these acceptable:

a) Compute the change in forward price as defined in 1) when tweaking par yields and repo rates, with the forward date being hte delivery date of the future contract.

b) Compute the change in price of a bond as defined 2), with the settlement being the delivery date of the contract, a conversion factor applied to the result.

## Answer by Helin (score 1)

https://quant.stackexchange.com/a/33950

With regard to your first question: theoretically, these two methods should produce the same forward price. In practice, they might not. This is because a lot of (most?) institutions build the front end of the bond curve using short-term Treasuries. The forward price calculated using such a curve (as discounted present value) effectively assumes the forward position is financed at short-term Treasury rates. By contrast, your "spot - carry" calculation mostly likely uses repo financing. Since repo trades at a spread to Treasury rates, the two answers will differ slightly. If your yield curve is built with repo rates at the front end, you'd have no problems. (There's another reason why the two answers might differ, which boils down to what market convention you use to calculate your carry, see formula for forward price of bond for details.)

As to bond futures DV01, it is most often calculated by shifting the yields of the entire delivery basket by $x$ basis points, holding repo rates constant, and then apply a bond futures model to see how much price changes by (a model is needed to account for changes in delivery option value). You have the options of shifting either the spot yield or the forward yield; you should discuss with your traders to see whether they have any preferences.

You can also hold bond yields constant and shift the repo curve, which produces the "rho" (sensitivity to financing rate).

## Answer by InnocentR (score 0)

https://quant.stackexchange.com/a/36619

Theoretically, both methods should give same result. But practically, term repo market is very illiquid (wide bid-offer that become wider further out you go, though you might find quotes for term repos to bond future delivery dates) plus the discount curve is never fully defined for bond markets, traders use their intuition to build one - unique curve does not exist. These two issues will give differences.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.