Hull-White Cap and Floor Calibration: Choosing the Curve

Summary

The document describes a Hull-White interest-rate model calibration problem. The questioner calibrates the model to cap and floor prices while using an OIS curve to match the current term structure. Caps are priced correctly, but floors show large errors, concentrated in the earliest caplets and fading at later maturities. The post gives no diagnostic details about the model setup, market conventions, or calibration procedure.

The accepted answer reports resolving the issue by using a zero curve derived from six-month instruments instead of the OIS curve. This is a specific practical observation, not a general derivation of why the curve change fixes the mismatch. The document does not compare the resulting calibration across instruments or discuss curve consistency, so the result should be treated as context-dependent guidance for diagnosing cap and floor pricing differences.

Key ideas

• The question concerns calibrating a Hull-White model to cap and floor prices.
• Using the OIS curve produced acceptable cap prices but large floor mispricings in the earliest caplets.
• The reported fix was to use a zero curve derived from six-month instruments.
• The brief answer does not explain why that curve choice resolved the issue or establish that it applies generally.

Tags

# Hull White Cap/Floor calibration


# Hull White Cap/Floor calibration












I have a problem and I hope someone could help me.

I calibrated the Hull-White model to Caps and Floors from t=0 so the bond prices are equivalent to todays term structure. See: Hull-White zero-coupon bond price does not depend on the volatility?

For Caps it prices them correctly using the OIS curve as the term structure. But I get huge mispricings for Floors.

The mispricing mostly appears for the first 2-5 Caplets and later it disappears.

Do you have an idea what the problem could be, or do I have a fundamental mistake?

I am just calibrating the HW model to Cap and Floor prices.

Thanks in advance Chris

## Answer by Christian M (score 0, accepted)

https://quant.stackexchange.com/a/49749

I solved the problem by just simply using the Zero curve derived from 6 month instruments instead of the OIS.

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.