Volatility Regimes and the Curvature of FX Option Smiles

Summary

The document raises an empirical observation about short-term foreign-exchange options: out-of-the-money implied volatility relative to at-the-money volatility appears to show greater smile curvature in low-volatility environments. It notes that related findings have been reported for equities, then asks whether the pattern can arise mathematically from stochastic-volatility models when jumps and non-normal returns are excluded. Mean reversion in volatility is proposed as a possible ingredient, along with correlation between spot and volatility.

No model derivation, data analysis, or answer is supplied, so the mechanism remains unresolved in the document. The observation is framed as an empirical regularity rather than a demonstrated general law, and the suggested explanations are questions rather than conclusions. The text does not specify the smile-curvature measure, the volatility regime definition, the option maturities beyond being short-term, or the model parameters needed to assess whether mean reversion or spot-volatility correlation can account for the effect.

Key ideas

• The author reports higher relative smile curvature for short-term FX options in low-volatility environments.
• The document mentions similar findings in equities but gives no supporting analysis.
• It asks whether mean-reverting stochastic volatility can explain the observation without jumps or non-normal returns.
• Spot-volatility correlation is raised as a possible mechanism, not established as the cause.
• The proposed relationship remains an open question in this document.

Tags

# Effect of Volatility Regime on Volatility Smile


# Effect of Volatility Regime on Volatility Smile












For short-term FX options, I find empirically that the degree of curvature of the smile (OTM/ATM in %) is higher in low volatility environments. Similar results are found by Pena et al. ("Why do we smile", 1999) for stock markets.

My question is: If we rule out jumps and non-normal returns, can this effect mathematically follow from stochastic volatility models - as long as we believe in mean reversion of volatility?

Maybe due to a correlation with spot and volatility?

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.