<i>L<sub>p </sub></i> quantile regression with realized measure

TANG Li; CHEN Yu

doi:10.3969/j.issn.0253-2778.2020.12.005

PDF( 1251 KB)

Open Access JUSTC Research Article

L_p quantile regression with realized measure

TANG Li,
CHEN Yu

Department of Statistics and Finance, School of Management, University of Science and Technology of China, Hefei 230026, China

Funds: The work is supported by the National Key Research and Development Plan (2016YFC0800100), the NNSF of China (71771203).

Cite this:

https://doi.org/10.3969/j.issn.0253-2778.2020.12.005

More Information

Author Bio:
Tang Li received his master degree from University of Science and Technology of China in 2020. His research interests focus on risk management.
Corresponding author: Chen Yu received her Ph. D. degree in probability and statistics from University of Science and Technology of China in 2006. She is an associate professor of Department of Statistics and Finance, School of Management, University of Science and Technology of China. Her research interests include network risk analysis, extreme value theory, and high-frequency data analysis.
Publish Date: 30 December 2020

Abstract Full text PDF

Abstract

Abstract

A new financial risk model named L_pquantile regression with a realized measure (realized L_p quantile) was proposed. The realized measure and L_p quantiles were combined and L_p quantile were added to the measurement equation. The realized Lp quantile model is a generic model that includes realized quantile model and expectile model. An asymmetric exponential power distribution (AExpPow) was proposed to derive the formula of log-likelihood. And a simulation was conducted to justify the validity of the log-likelihood. Finally an empirical study was conducted to justify the validity of the realized L_p quantile. And some conclusions were drawn as follows: differfent power indices suit different data and different time-frequencies suit different realized measures, and higher frequency is not always better.

Abstract

A new financial risk model named L_pquantile regression with a realized measure (realized L_p quantile) was proposed. The realized measure and L_p quantiles were combined and L_p quantile were added to the measurement equation. The realized Lp quantile model is a generic model that includes realized quantile model and expectile model. An asymmetric exponential power distribution (AExpPow) was proposed to derive the formula of log-likelihood. And a simulation was conducted to justify the validity of the log-likelihood. Finally an empirical study was conducted to justify the validity of the realized L_p quantile. And some conclusions were drawn as follows: differfent power indices suit different data and different time-frequencies suit different realized measures, and higher frequency is not always better.