Answers are grounded in this paper and cite their sources.
This paper investigates accuracy of volatility estimation methods for asset prices characterized by several volatility regimes. Two criteria for evaluation of estimator accuracy are estimation volatility error and option pricing error in the context of the Black-Scholes option-pricing model. We performed a Monte Carlo experiment to obtain average option pricing errors for various patterns of underlying asset volatility. In the case of an ergodic in variance process it is acceptable to use volatility estimates derived either from sample standard deviation of continuously compounded returns or from GARCH-fitted volatility of continuously compounded returns. In the case of a non-ergodic process all estimation methods always yield biased results, in terms of both volatility errors and option pricing errors. Directions of the biases are case-specific.
No takes yet. Share an insight, caveat, or question.
Rakhmayil et al. (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: