Skip to content
All library documents

Treasury Futures Conversion Factors and the Cheapest-to-Deliver Bond

Article Quant Q&A · Author: S. Dupree
Read the original

Summary

The document explains why a Treasury futures conversion factor should not be treated as a general estimate of a deliverable bond’s current price. The factor is based on a standardized 6% yield convention, while the question’s proposed alternative would vary with the prevailing yield curve. The response says the convention is an approximation and points to maturity rounding as one source of differences: a bond’s remaining term may be rounded down when its factor is determined.

The response also reframes the futures contract as linked to a basket of eligible bonds rather than to a hypothetical 6% coupon bond. In practice, the cheapest-to-deliver bond can dominate price behavior, and its maturity and coupon may differ substantially from the contract’s label or the standard coupon assumption. The examples are illustrative and tied to particular Treasury contracts; they do not derive a general pricing formula or quantify approximation error. The document recommends specialist references for a fuller treatment of delivery options and the futures cash basis.

Key ideas

  • A Treasury futures conversion factor follows a standardized 6% yield convention and is an approximation.
  • Remaining-maturity rounding can contribute to differences between the factor and a bond’s price.
  • A Treasury futures contract is associated with a basket of eligible deliverable bonds.
  • The cheapest-to-deliver bond can drive futures behavior more strongly than the contract’s nominal maturity suggests.

Tags

FuturesFixed incomeDerivatives pricing
Full text
# Conversion factor for bonds


# Conversion factor for bonds












Hull defines the conversion factor for a bond as the "quoted price the bond would have per dollar of principal on the first day of the delivery month on the assumption that the interest rate for all maturities equals 6% per annum."

My understanding is that the hypothetical bond underlying the futures contract has a 6% coupon, and that the (quoted) price of this bond varies with the zero curve. If this zero curve is flat and equal to 6% for all maturities, then the quoted price of the hypothetical bond should be 100 and the product of this hypothetical price and the conversion factor for a particular bond is exactly the quoted price of the bond in question. However, if the zero curve is flat and not equal to 6%, or more generally is not flat at all, then why should this conversion factor be considered a good approximation of the quoted price of the actual bond?

Suppose for the moment that the zero curve is flat. Let P(N,y,c) be the present value of a bond with semiannual coupon c and time to maturity N when the zero rate is y (for all maturities). The present value of the hypothetical bond is P(N,y,0.06). What naturally seems to be the correct conversion factor to get the quoted price of a bond with coupon c and time to maturity M is

CF = P(M,y,c)/P(N,y,0.06).

When y=0.06, this conversion factor is the same one defined by Hull, but otherwise they need not be the same. Is there some reason why it is assumed that the conversion factor is constant? y need not be close to 0.06, and M can be different from N, so it doesn't seem clear that the (constant) conversion factor gives anything useful.

## Answer by Helin (score 4)

https://quant.stackexchange.com/a/14556

First, the exact computation of conversion factor is actually quite tricky. The "6% yield" rule is really an approximation (although a very good one). CME provides a spreadsheet that you can use to compute the exact conversion factor for each bond and each contract (http://www.cmegroup.com/trading/interest-rates/us-treasury-futures-conversion-factor-calculator.html). The main reason for the discrepancy comes from rounding of term to maturity. For example, if we are looking at the classic bond (US) contract and a deliverable has 15 years 2 months to maturity on the first day of the delivery month, then you need to round it down to 15 years.

Second, it's not a good idea to think of the underlying as a hypothetical 6% coupon bond. What makes bond futures so complex and so interesting is that a full basket of deliverables act as the underlying. For example, for TYZ4 (10-year Treasury note expiring in December of this year), the deliverable basket includes 21 Treasury notes (5 of which haven't even been auctioned!) Right now, TYZ4 behaves very similarly to 2.25% 31-July-2021 (the cheapest-to-deliver into TYZ4). You can see that 1) it behaves nothing like a 10-year bond -- in fact, over past few years, the 10-year Treasury contract has mostly behaved like a 7-year bond; and 2) it behaves nothing like a bond with a 6% coupon.

The definitive guide to bond futures is The Treasury Bond Basis: An in-Depth Analysis for Hedgers, Speculators, and Arbitrageurs. I read it cover to cover a few times and learned new things every time I read it... In addition, there are a few Salomon Brothers research notes that I found tremendously helpful:

- Koenigsberg, Mark. "Understanding Treasury Bond Futures: Questions and Answers." Salomon Brothers Bond Portfolio Analysis Group (1990).

- Koenigsberg, Mark. "The Salomon Brothers Delivery Option Model: Understanding Treasury Bond Futures." Salomon Brothers Bond Portfolio Analysis Group (1991).

- Koenigsberg, Mark. "Deciding Between Futures and Cash: A Closer Look at the Basis." Salomon Brothers Bond Portfolio Analysis Group (1991).

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.