TY - JOUR
T1 - Efficient asymptotic variance reduction when estimating volatility in high frequency data
AU - Clinet, Simon
AU - Potiron, Yoann
N1 - Funding Information:
We would like to thank Markus Bibinger, Jia Li, Dacheng Xiu, Jean Jacod, Yacine A?t-Sahalia (the Editor), two anonymous referees and an anonymous Associate Editor, the participants of Keio Econometrics Workshop, the Workshop on Portfolio dynamics and limit order books in Ecole Centrale Paris, The Quantitative Methods in Finance 2016 Conference in Sydney for helpful discussions and advice. The research of Yoann Potiron is supported by Japanese Society for the Promotion of Science Grant-in-Aid for Young Scientists No. 60781119. All financial data are provided by the Chair of Quantitative Finance of the Ecole Centrale Paris. The research of Simon Clinet is supported by CREST Japan Science and Technology Agency (14532201).
Publisher Copyright:
© 2018 Elsevier B.V.
PY - 2018/9
Y1 - 2018/9
N2 - This paper shows how to carry out efficient asymptotic variance reduction when estimating volatility in the presence of stochastic volatility and microstructure noise with the realized kernels (RK) from Barndorff-Nielsen et al. (2008) and the quasi-maximum likelihood estimator (QMLE) studied in Xiu (2010). To obtain such a reduction, we chop the data into B blocks, compute the RK (or QMLE) on each block, and aggregate the block estimates. The ratio of asymptotic variance over the bound of asymptotic efficiency converges as B increases to the ratio in the parametric version of the problem, i.e. 1.0025 in the case of the fastest RK Tukey-Hanning 16 and 1 for the QMLE. The impact of stochastic sampling times and jump in the price process is examined carefully. The finite sample performance of both estimators is investigated in simulations, while empirical work illustrates the gain in practice.
AB - This paper shows how to carry out efficient asymptotic variance reduction when estimating volatility in the presence of stochastic volatility and microstructure noise with the realized kernels (RK) from Barndorff-Nielsen et al. (2008) and the quasi-maximum likelihood estimator (QMLE) studied in Xiu (2010). To obtain such a reduction, we chop the data into B blocks, compute the RK (or QMLE) on each block, and aggregate the block estimates. The ratio of asymptotic variance over the bound of asymptotic efficiency converges as B increases to the ratio in the parametric version of the problem, i.e. 1.0025 in the case of the fastest RK Tukey-Hanning 16 and 1 for the QMLE. The impact of stochastic sampling times and jump in the price process is examined carefully. The finite sample performance of both estimators is investigated in simulations, while empirical work illustrates the gain in practice.
KW - High frequency data
KW - Integrated volatility
KW - Jumps
KW - Market microstructure noise
KW - Quasi-maximum likelihood estimator
KW - Realized kernels
KW - Stochastic sampling times
UR - http://www.scopus.com/inward/record.url?scp=85048806404&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85048806404&partnerID=8YFLogxK
U2 - 10.1016/j.jeconom.2018.05.002
DO - 10.1016/j.jeconom.2018.05.002
M3 - Article
AN - SCOPUS:85048806404
SN - 0304-4076
VL - 206
SP - 103
EP - 142
JO - Journal of Econometrics
JF - Journal of Econometrics
IS - 1
ER -