The Korean Journal of Applied Statistics (응용통계연구)

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

Power transformation in quasi-likelihood innovations for GARCH volatility

금융 시계열 변동성 추정을 위한 준-우도 이노베이션의 멱변환

  • Sunah, Chung (Department of Statistics, Sookmyung Women's University) ;
  • Sun Young, Hwang (Department of Statistics, Sookmyung Women's University) ;
  • Sung Duck, Lee (Department of Statistics, Chungbuk National University)
  • 정선아 (숙명여자대학교 통계학과) ;
  • 황선영 (숙명여자대학교 통계학과) ;
  • 이성덕 (충북대학교 정보통계학과)
  • Received : 2022.09.09
  • Accepted : 2022.09.26
  • Published : 2022.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

This paper is concerned with power transformations in estimating GARCH volatility. To handle a semi-parametric case for which the exact likelihood is not known, quasi-likelihood (QL) rather than maximum-likelihood method is investigated to best estimate GARCH via maximizing the information criteria. A power transformation is introduced in the innovation generating QL estimating functions and then optimum power is selected by maximizing the profile information. A combination of two different power transformations is also studied in order to increase the parameter estimation efficiency. Nine domestic stock prices data are analyzed to order to illustrate the main idea of the paper. The data span includes Covid-19 pandemic period in which financial time series are really volatile.

본 논문에서는 금융 시계열 변동성 추정을 위한 준-모수(quasi-likelihood) 방법을 다루고 있다. 모형식에서 오차항의 분포를 미지(unknown)로 하여 준-우도 함수를 통한 모수 추정을 하는 경우 이노베이션의 지정을 멱변환을 통해 구성하였다. 고정된 멱변환에 대한 프로파일-정보 행렬을 비교하여 최대값을 제공하는 멱변환을 제안하였다. 이차원 이노베이션으로의 확장을 다루었으며 코로나 펜데믹 기간의 높은 변동성을 보이는 국내 9개 주가 자료 분석을 통해 방법론을 예시하고 있다.

Keywords

Acknowledgement

이 논문은 2021년도 한국연구재단의 지원을 받아 수행된 연구임 (NRF-2021R1F1A1047952)

References

  1. Bollerslev T (1986). Generalized autoregressive conditional heteroscedasticity, Journal of Econometrics, 31, 307-327. https://doi.org/10.1016/0304-4076(86)90063-1
  2. Chung S and Hwang SY (2017). A profile Godambe information of power transformations for ARCH time series, Communications in Statistics-Theory and Methods, 46, 6899-6908. https://doi.org/10.1080/03610926.2016.1139133
  3. Francq C, Lepage G, and Zakoian JM (2011). Two-stage non Gaussian QML estimation of GARCH models and testing the efficiency of the Gaussian QMLE, Journal of Econmetrics, 165, 246-257. https://doi.org/10.1016/j.jeconom.2011.08.001
  4. Francq C and Zakoian JM (2013). Optimal predictions of powers of conditionally heteroscedastic processes, Journal of the Royal Statistical Society B, 75, 345-367. https://doi.org/10.1111/j.1467-9868.2012.01045.x
  5. Godambe VP (1985). The foundation of finite sample estimation in stochastic processes, Biometrika, 72, 419-428. https://doi.org/10.1093/biomet/72.2.419
  6. Hansen PR and Lunde A (2005). A forecast comparison of volatility models : Does anything beat a GARCH (1, 1)?, Journal of Applied Econometrics, 20, 873-889. https://doi.org/10.1002/jae.800
  7. Hwang SY and Basawa IV (2004). Stationarity and moment structure for Box-Cox transformed threshold GARCH (1, 1) processes.
  8. Hwang SY and Basawa IV (2011). Godambe estimating functions and asymptotic optimal inference, Statistics & Probability Letters, 81, 1121-1127. https://doi.org/10.1016/j.spl.2011.03.006
  9. Tsay RS (2010). Analysis of Financial Time Series (third ed), Wiley, New York.
  10. Vasudevay R and Kumari JV (2013). On general error distributions, ProbStat Forum, 6, 89-95.
  11. Yoon JE and Hwang SY (2021). On the threshold innovation in quasi-likelihood for conditionally heteroscedastic time series, Communications in Statistics : Simulation and Computation, 50, 2042-2053. https://doi.org/10.1080/03610918.2019.1593453