摘要
We establish a feasible central limit theorem with convergence rate n1/8 for the estimation of the integrated volatility of volatility (VoV) based on noisy high-frequency data with jumps. This is the first inference theory ever built for VoV estimation under such a general setup. The central limit theorem is applied to provide interval estimates of the VoV and conduct hypothesis tests. Furthermore, when one is interested in the null hypothesis that the VoV is zero, we show that a more powerful test can be established based on a VoV estimator with a convergence rate n1/5 under the null. Empirical results on the S&P 500 and individual stocks show strong evidence of non-zero VoV.
学科与关键词
关键词
Volatility of volatilityCentral limit theoremHigh frequency dataMicrostructure noiseSemimartingale
- Heteroscedasticity test of high-frequency data with jumps and market microstructure noiseLiu, Qiang; Liu, Zhi*; Zhang, ChuanhaiAPPLIED STOCHASTIC MODELS IN BUSINESS AND INDUSTRY, 2022, 38(3): 441-457.被引 1
- Trading information, price discreteness, and volatility estimationTang, Yinfen; Peng, Ye; Zhang, Zhiyuan*JOURNAL OF STATISTICAL PLANNING AND INFERENCE, 2022, 220: 49-70.
- Estimating spot volatility under infinite variation jumps with dependent market microstructure noiseLiu, Qiang*; Liu, ZhiEconometrics Journal, 2024.
