Finding Stochastic Volatility Models for Simulation and Estimation

Summary

The document describes a researcher’s search for continuous-time stochastic volatility models suitable for simulation and estimation through indirect inference. The stated preferences are models that are reasonably easy to simulate and have relatively few parameters. Heston-style square-root processes and CEV-style processes, with or without mean reversion, are already under consideration; multifactor variants remain possible.

The reply recommends a book on stochastic volatility, local volatility, and local-stochastic-volatility models, noting its application to equity derivatives and potential relevance to foreign exchange. It does not name particular models from the book, provide simulation procedures, compare parameter counts, or evaluate estimation performance. Consequently, the recommendation may help broaden the researcher’s reading, but it does not directly answer which parsimonious models are easiest to simulate or best suited to indirect inference. The document contains no empirical results or model-selection guidance.

Key ideas

• The research goal is to compare continuous-time volatility models using indirect inference.
• Ease of simulation and a small parameter set are key model-selection preferences.
• The candidate families mentioned include square-root and CEV-type volatility processes.
• The reply recommends a broad reference covering local, stochastic, and combined volatility models.
• No specific simulation method or comparative evidence is supplied.

Tags

# Reference request about stochastic volatility model


# Reference request about stochastic volatility model












I'm fiddling with estimation of stochastic volatility models and have build up a somewhat flexible framework using indirect inference.

I would like to try and throw a lot of different continuous time models into it and see how it performs. I only need it to be reasonable easy to simulate and not have to many parameters.

I know of Heston type models (i.e. square root processes) and the CEV type processes with and without mean reversion. Multifactor is an option but generally the fewer parameters the easier for me (as always).

Hope to get your input, thanks.

## Answer by Lost1 (score 1)

https://quant.stackexchange.com/a/76735

In my opinion, the best book on this area is Lorenzo Bergomi's "Stochastic Volatility Models" which covers the local volatility models, stochastic volatility models and local stochastic volatility models in great details with their properties and appropriateness.

His application is equities derivatives, but this would also be useful for FX. He was head of SocGen equity derivatives research for many years. SocGen is one of the most technical shop on the street.

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.