Nonparametric spot volatility from options

Viktor Todorov*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

We propose a nonparametric estimator of spot volatility from noisy short-dated option data. The estimator is based on forming portfolios of options with different strikes that replicate the (risk-neutral) conditional characteristic function of the underlying price in a model-free way. The separation of volatility from jumps is done by making use of the dominant role of the volatility in the conditional characteristic function over short time intervals and for large values of the characteristic exponent. The latter is chosen in an adaptive way in order to account for the time-varying volatility. We show that the volatility estimator is near rate-optimal in minimax sense. We further derive a feasible joint central limit theorem for the proposed option-based volatility estimator and existing high-frequency return-based volatility estimators. The limit distribution is mixed Gaussian reflecting the time-varying precision in the volatility recovery.

Original languageEnglish (US)
Pages (from-to)3590-3636
Number of pages47
JournalAnnals of Applied Probability
Volume29
Issue number6
DOIs
StatePublished - Jan 1 2019

Keywords

  • Itô semimartingale
  • Jumps
  • Nonparametric inference
  • Options
  • Stable convergence
  • Stochastic volatility

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Fingerprint Dive into the research topics of 'Nonparametric spot volatility from options'. Together they form a unique fingerprint.

Cite this