DOI QR코드

DOI QR Code

Sparse vector heterogeneous autoregressive model with nonconvex penalties

  • Received : 2021.07.02
  • Accepted : 2021.09.23
  • Published : 2022.01.31

Abstract

High dimensional time series is gaining considerable attention in recent years. The sparse vector heterogeneous autoregressive (VHAR) model proposed by Baek and Park (2020) uses adaptive lasso and debiasing procedure in estimation, and showed superb forecasting performance in realized volatilities. This paper extends the sparse VHAR model by considering non-convex penalties such as SCAD and MCP for possible bias reduction from their penalty design. Finite sample performances of three estimation methods are compared through Monte Carlo simulation. Our study shows first that taking into cross-sectional correlations reduces bias. Second, nonconvex penalties performs better when the sample size is small. On the other hand, the adaptive lasso with debiasing performs well as sample size increases. Also, empirical analysis based on 20 multinational realized volatilities is provided.

Keywords

Acknowledgement

We would like to thank an Associate Editor and anonymous referees for many useful comments and suggestions.

References

  1. Andersen TG, Bollerslev T, Diebold FX, and Labys P (2003). Modeling and forecasting realized volatility. Econometrica, 71, 579-625. https://doi.org/10.1111/1468-0262.00418
  2. Baek CR, Davis RA, and Pipiras V (2018). Periodic dynamic factor models: estimation approaches and applications. Electronic Journal of Statistics, 12, 4377-4411. https://doi.org/10.1214/18-ejs1518
  3. Baek CR and Park MS (2020). Sparse vector heterogeneous autoregressive modeling for realized volatility. Journal of the Korean Statistical Society, 50, 1-16.
  4. Breheny P and Huang J (2011). Coordinate descent algorithms for nonconvex penalized regression with applications to biological feature selection. The Annals of Applied Statistics, 5, 232. https://doi.org/10.1214/10-AOAS388
  5. Chen J and Chen Z (2008). Extended Bayesian information criteria for model selection with large model spaces. Biometrika, 95, 759-771. https://doi.org/10.1093/biomet/asn034
  6. Corsi F (2009). A simple approximate long-memory model of realized volatility. Journal of Financial Econometrics, 7, 174-196. https://doi.org/10.1093/jjfinec/nbp001
  7. Davis RA, Zang PF, and Zheng T (2016). Sparse vector autoregressive modeling. Journal of Computational and Graphical Statistics, 25, 1077-1096. https://doi.org/10.1080/10618600.2015.1092978
  8. Fan Jianqing and Li Runze (2001). Variable selection via nonconcave penalized likelihood and its oracle properties. Journal of the American statistical Association, 96, 1348-1360. https://doi.org/10.1198/016214501753382273
  9. Kim DW and Baek CR (2020). Factor-augmented HAR model improves realized volatility forecasting. Applied Economics Letters, 27, 1002-1009. https://doi.org/10.1080/13504851.2019.1657554
  10. Lutkepohl H (2005). New Introduction to Multiple Time Series Analysis, Springer Science & Business Media, Springer, New York.
  11. Zhang CH (2010). Nearly unbiased variable selection under minimax concave penalty. The Annals of Statistics, 38, 894-942. https://doi.org/10.1214/09-AOS729
  12. Zhu X (2020). Nonconcave penalized estimation in sparse vector autoregression model. Electronic Journal of Statistics, 14, 1413-1448.
  13. Zou H (2006). The adaptive lasso and its oracle properties. Journal of the American Statistical Association, 101, 1418-1429. https://doi.org/10.1198/016214506000000735