Why Callable Bond Spreads May Not Determine a Unique Price
Summary
The document explains why a callable bond’s Z-spread may not map cleanly to a unique price. The call option makes the bond’s value depend on whether and when the issuer exercises, and that decision is affected by both risk-free rates and credit spreads. Consequently, a spread or yield can correspond to multiple prices under different exercise assumptions.
A practical approximation is to assess whether the call is likely to be exercised across plausible prices. If exercise is clearly unlikely or likely, price the bond under that assumption. When exercise is uncertain, one can assume an exercise outcome, derive a price, recalculate yield to worst, and check whether the assumption remains consistent. The response cautions that this is a crude approach that omits the call’s value; option-adjusted spread and models incorporating rates and credit spreads may be more suitable when available.
Key ideas
- Callable bond spread-to-price conversion can be non-unique because the call exercise assumption affects valuation.
- Call moneyness depends on both risk-free rates and credit spreads.
- A simple approximation prices the bond under an assumed call exercise outcome and checks consistency with yield to worst.
- The approximation can miss the call option’s value and should be treated cautiously.
Tags
Full text
# Single Callable Bond - from Option Price to Z-Spread # Single Callable Bond - from Option Price to Z-Spread The pricing software I use requires the specification of a z-spread for the valuation of a single callable bond. According to many sources, however, the Z-spread is determined based on the bond price, which is a catch-22. I am now trying to price the callable bond as a plain vanilla bond minus a receiver swaption. The subsequent question would then be: how can I derive the Z-spread from the swaption price? ## Answer by Dimitri Vulis (score 1) https://quant.stackexchange.com/a/85206 Very generally, for an arbitrary callable bond, you simply can't always convert a spread (or yield) to price because the solution may not be unique. Often you can, but it's possible to contrive examples where the same yield could correspond to more than one prices / exercise assumptions. If your tools support Option Adjusted Spread (OAS), rather than Z-spread, then you're a little better off converting that to price, keeping its limitations in mind. It would be better than trying to price an interest rate swaption separately. The moneyness of the call has 2 main drivers: risk-free interest rates, and the credit spread. Some people price callables off of 2-dimensional trees (rates and credit spread) with some confidence in the result. Practically, start by looking at the moneyness of the call. A lot of times, by design, the call is so far away from the money for all reasonable prices, that you can assume that you know for sure whether it'll be exercised. Then convert the Z spread to price under this exercise assumption. This isn't great, because we've seen plenty of times how bonds with out of the money make-whole calls sometimes get called. If the bond may or may not get called, then you can somehow guess the exercise assumption for yield to worst, and then under this assumption convert the Z spread to price, recalculate yield to worst from the resulting price, and check that the exercise assumptions are the same. But this is really crude, ignores the value of the call, and you shouldn't rely too much on such a price.
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.