Recent advances in econometric methodology and newly available sources of data are used to examine empirically the performance of the various extreme-value volatility estimators that have been proposed over the past two decades. Overwhelming support is found for the use of extreme-value estimators when computing daily volatility measures across all assets: Daily extreme-value volatility estimators are both less biased and substantially more efficient than the traditional close-to-close estimator. In the case of weekly and monthly measures, the results still suggest that extreme-value estimators are appropriate, but the evidence is more mixed.
ASJC Scopus subject areas
- Business, Management and Accounting(all)
- Economics and Econometrics