Skip to content
All library documents

Heston Implied Volatility Approximations Across Market Regimes

Article Quant Q&A · Author: jaehyukchoi49

Summary

The discussion asks whether Heston option prices have a convenient approximate Black–Scholes implied volatility formula, comparable to formulas used with other stochastic volatility models. Heston European option prices can be computed accurately through Fourier inversion, then numerically converted to implied volatility, but numerical methods may be unstable for extreme strikes or very short maturities. The answer emphasizes that no single approximation serves every use case; useful formulas depend on the asymptotic regime being studied.

It outlines separate research for extreme strikes, short maturities, and long maturities. Short-time behavior is a particular challenge for diffusion models, prompting work on jump extensions and rough volatility, including rough Heston. For long maturities, implied volatility is described as converging toward an SVI surface parametrization. These are research directions and asymptotic results rather than one universal plug-in formula. The answer also notes that newer work may have appeared, so the cited overview is not necessarily comprehensive.

Key ideas

  • Heston option prices can be converted to Black–Scholes implied volatility numerically after Fourier pricing.
  • A single analytic approximation may not work equally well across all strikes and maturities.
  • Separate asymptotic methods address extreme strikes, short maturities, and long maturities.
  • Diffusion based Heston models can struggle to describe the short end of the volatility surface.
  • Long maturity implied volatility is described as converging toward an SVI parametrization.

Tags

Full text
# Equivalent BS volatility formula under the Heston model?


# Equivalent BS volatility formula under the Heston model?












Is there an equivalent BS volatility formula for the Heston model, something like Hagan's formula for the SABR model? Of course, such a formula will be an approximation as in Hagan's formula.

Under the Heston model, we can price European options with the inverse Fourier transform (or FFT) quite precisely. So it's possible to numerically invert the price to the BS volatility. Nevertheless, an analytic volatility formula (albeit approximation) will be still helpful for many occasions. For example, FFT method seems unstable for extreme inputs (deep out-of-the-money or short time-to-maturity). See this question.

## Answer by Stephan Sturm (score 3, accepted)

https://quant.stackexchange.com/a/71155

Answering the question in a way coming from the motivation at the back: There is not one universal approximation, but different ones depending on which asymptotics you want to look into. Just to state a few:

- For the extreme strike you have the results by Friz, Gerhold, Gulisashvili and myself, https://arxiv.org/abs/1001.3003, see also the book by Archil Gulisasvili, https://link.springer.com/book/10.1007/978-3-642-31214-4

- For the small time regime you have the results by Forde and Jacquier, https://nms.kcl.ac.uk/martin.forde/HestonSmallTime.pdf. That said, to the best of my knowledge Heston has all diffusion based models an issue with the short end of the vola surface. Besides adding jumps, there is a lot recent work on rough volatility models, also rough Heston, to address these issues. I am not very familiar with the recent literature on these topic.

- Finally, in the large time regime Gatheral and Jacquier have proved that the implied volatility converges to the SVI parametrization of the vol-surface, https://arxiv.org/abs/1002.3633

I am happy to admit that I have not worked on related ideas for several years there might be well newer and better results out, I hope others can fill in the gaps.

PS. While quite early in the asymptotics literature, I feel it is always appropriate to point to point to the 2010 Zeliade Heston White Paper by Jacquier and Martini, I think it really helps to understand the intricacies of Heston asymptotics besides any specific expansion, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1769744.

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.