Using HJM to Simulate Bond Yield Scenarios and Curve Shocks

Summary

The document asks how to model future yield scenarios for a basket of bonds and measure how an initial yield-curve shock affects those scenarios and bond futures prices. It proposes using the Heath–Jarrow–Morton framework to simulate forward rates for zero-coupon bonds, then price the basket bonds at the future horizon and infer their yields.

The questions focus on implementation and calibration: whether the forward-rate model should be calibrated to a zero-coupon curve or directly to the basket bonds, and whether an initial parallel shift or steepening can be incorporated to compare scenario outcomes. The post offers a proposed workflow but supplies no model specification, calibration procedure, results, or answer. Practical choices such as curve construction, bond cash flows, and shock treatment therefore remain unresolved in the document.

Key ideas

• The proposed workflow simulates forward-rate scenarios with an HJM model and values basket bonds at a future horizon.
• Yields at that horizon would be inferred from the simulated bond values.
• The author asks whether calibration should use the zero-coupon curve or bonds in the portfolio.
• Initial curve shocks such as parallel shifts and steepening are intended to affect future scenarios and bond futures prices.
• No implementation details or empirical results are provided.

Tags

# Modeling Yield Scenarios and Curve Shocks for Bonds


# Modeling Yield Scenarios and Curve Shocks for Bonds












I would like to do the following:

Given a basket of bonds I want to generate different yield scenarios at a future time $T$ for the different bonds in my basket. I also want to see how I can shock the yield curve initially (using for example a parallel shift), and see how it impact my different yield scenarios at time $T$. In this way I want to be able to see how a shock of the yield curve affects the bond future price.

The first model I considered for this purpose is the HJM framework. However, I am uncertain about how to implement it in practice. Using the HJM framework, I would generate different forward rate scenarios for zero-coupon bonds and then price the bonds in my basket using these forward rate scenarios at time $T$, subsequently determining the yields.

Is this a correct method to achieve my objective? Should the forward rates be calibrated on the zero-coupon curve, or can they be directly calibrated on the bonds in my basket? Additionally, will this method allow me to incorporate initial yield curve shocks such as steepening or parallel shifts to observe their effects on my future scenarios?

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.