One-Factor Hull–White Model and Bermudan Hedge Robustness

Summary

The document asks how to interpret a claim that single-factor interest-rate models can make the vanilla-option hedge for a Bermudan option sensitive to changes in the yield curve. The author observes that hedges in any model may need rebalancing when rates move, and asks why a one-factor model would imply more rebalancing or less robustness.

No answer, derivation, hedge comparison, or empirical evidence is included. The useful point is the distinction the question raises: ordinary hedge rebalancing after market moves is not itself evidence that a model is less robust. Assessing the claim would require specifying what robustness means, how the hedge portfolio is chosen, and how curve shifts alter the hedge instruments or residual risk. The document provides no basis for concluding that the one-factor Hull–White model requires more frequent rebalancing than other models.

Key ideas

• The document questions whether a one-factor rate model makes Bermudan hedges unusually sensitive to curve shifts.
• Hedge rebalancing after rate moves can occur in models with any number of factors.
• The claim cannot be assessed without defining hedge robustness and comparing residual risk under specified curve changes.
• The document presents the question but supplies no answer or supporting evidence.

Tags

# Hedge robustness of the one factor Hull White model


# Hedge robustness of the one factor Hull White model












I recently came across a quote in a book:

"All single factor models share the limitation that shifts in curve levels cause shifts in the package of vanilla options that are a good hedge for the Bermudan option".

As far as I can see, all models would require hedge rebalancing for the Bermudan when rates change.

I fail to see why in particular, a one factor model would imply a higher amount of rebalancing, which is what I think this statement implies. Can anyone interpret this statement better for me, or reason why in particular a one factor model is less robust?

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.