Abstract
In this paper we develop a weighted forward search (WFS) approach to the correction of outliers in GARCH(1,1) models relying on the foward search (FS) method introduced by Atkinson and Riani (Robust diagnostic regression analysis. Springer, New York, 2000). The WFS is based on a weighting system of each unit and is an extension of the FS from independent to dependent observations. We propose a WFS test for the detection of outliers in GARCH(1,1) models and a WFS estimator of GARCH(1,1) coefficients which automatically corrects outliers. Extensive Monte Carlo simulations show the good performance of the WFS test with respect to other methods of outlier detection for the same models. The marked similarity between the distribution of MLE before strong contamination of the time series and after decontamination through the WFS proves the reliability of the WFS estimator. Finally, the application of the WFS procedure to several financial time series of the NYSE reveals the effectiveness of the method when extreme returns are observed.
Similar content being viewed by others
Notes
This point was suggested by a referee.
We are very grateful to a referee who brought our attention to this point.
Weekly frequency makes it possible to analyze long time periods with not too long time series. On the other hand, the presence of extreme returns is more probable than in the case of daily returns.
The same plots are available for the remaining companies analyzed and can be obtained from the authors upon request.
Abbreviations
- FS:
-
Forward search
- WFS:
-
Weighted forward search
- CDS:
-
Clean data set
References
Atkinson A, Riani M (2000) Robust diagnostic regression analysis. Springer, New York
Atkinson AC (1994) Fast very robust methods for the detection of multiple outliers. J Am Stat Assoc 89(428):1329–1339
Bahamonde N, Veiga H (2016) A robust closed-form estimator for the garch(1,1) model. J Stat Comput Simul 86(8):1605–1619
Bali R, Guirguis H (2007) Extreme observations and non-normality in ARCH and GARCH. Int Rev Econ Financ 16(3):332–346
Bilen C, Huzurbazar S (2002) Wavelet-based detection of outliers in time series. J Comput Gr Stat 11(2):311–327
Bollerslev T (1986) Generalized autoregressive conditional heteroskedasticity. J Econ 31(3):307–327
Bollerslev T (1987) A conditionally heteroskedastic time series model for speculative prices and rates of return. Rev Econ Stat 69(3):542–547
Bondon P, Bahamonde N (2012) Least squares estimation of ARCH models with missing observations. J Time Ser Anal 33(6):880–891
Carnero M, Peña D, Ruiz E (2007) Effects of outliers on the identification and estimation of GARCH models. J Time Ser Anal 28(4):471–497
Carstensen K (2003) The finite-sample performance of robust unit root tests. Stat Pap 44(4):469
Catalan B, Trivez FJ (2007) Forecasting volatility in GARCH models with additive outliers. Quant Financ 7(6):591–596
Cerioli A, Riani M (2002) Robust methods for the analysis of spatially autocorrelated data. Stat Methods Appl 11(3):335–358
Cerioli A, Farcomeni A, Riani M (2014) Strong consistency and robustness of the forward search estimator of multivariate location and scatter. J Multivar Anal 126:167–183
Charles A (2008) Forecasting volatility with outliers in GARCH models. J Forecast 27(7):551–565
Charles A, Darnè O (2005) Outliers and GARCH models in financial data. Econ Lett 86(3):347–352
Chen C, Liu LM (1993a) Forecasting time series with outliers. J Forecast 12(1):13–35
Chen C, Liu LM (1993b) Joint estimation of model parameters and outlier effects in time series. J Am Stat Assoc 88(421):284–297
Doornik JA, Ooms M (2005) Outlier detection in GARCH models, economics papers 2005-W24. Economics Group, Nuffield College, University of Oxford
Engle RF (1982) Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Econometrica 50(4):987–1007
Franses PH, Ghijsels H (1999) Additive outliers, GARCH and forecasting volatility. Int J Forecast 15(1):1–9
Franses PH, Van Dijk D (2000) Non-linear time series models in empirical finance. Cambridge University Press, Cambridge
Galeano P, Tsay RS (2010) Shifts in individual parameters of a GARCH model. J Financ Econ 8(1):122–153
Gómez V, Maraval A, Peña D (1999) Missing observations in ARIMA models: skipping approach versus additive outlier approach. J Econom 88(2):341–363
Grané A, Veiga H (2010) Wavelet-based detection of outliers in financial time series. Comput Stat Data Anal 54(11):2580–2593
Grané A, Veiga H (2014) Outliers, GARCH-type models and risk measures: a comparison of several approaches. J Empir Financ 26:26–40
Grossi L (2004) Analyzing financial time series through robust estimators. Stud Nonlinear Dyn Econ 8(2), article 3
Grossi L, Laurini F (2009) A robust forward weighted Lagrange multiplier test for conditional heteroscedasticity. Comput Stat Data Anal 53(6):2251–2263
Hampel FR, Ronchetti EM, Rousseeuw PJ, Stahel WA (1986) Robust statistics: the approach based on influence functions. Wiley, New York
Heagerty PJ, Lumley T (2000) Window subsampling of estimating functions with application to regression models. J Am Stat Assoc 95(449):197–211
Hill JB (2015) Robust estimation and inference for heavy tailed GARCH. Bernoulli 21(3):1629–1669
Hill JB, Prokhorov A (2016) GEL estimation for heavy-tailed GARCH models with robust empirical likelihood inference. J Econom 190(1):18–45
Hotta LK, Tsay R (2012) Outliers in GARCH processes. In: Bell WR, Holan SH, McElroy T (eds) Economic time series: modeling and seasonality. Chapman and Hall, pp 337–358
Huber PJ, Ronchetti EM (2009) Robust statistics, 2nd edn. Wiley, New York
Hubert M, Rousseeuw PJ, Van Aelst S (2008) High-breakdown robust multivariate methods. Stat Sci 23(1):92–119
Hwang S, Pereira PLV (2006) Small sample properties of GARCH estimates and persistence. Eur J Financ 12(6–7):473–494
Laurent S, Lecourt C, Palm FC (2016) Testing for jumps in conditionally gaussian ARMA-GARCH models, a robust approach. Comput Stat Data Anal 100:383–400
Lu W, Ke R (2016) A generalized least squares estimation method for the autoregressive conditional duration model. Stat Pap 1–24
Maronna RA, Martin DR, Yohai VJ (2006) Robust statistics. Wiley Series in Probability and Statistics, Wiley
Mendes BVDM (2000) Assessing the bias of maximum likelihood estimates of contaminated GARCH models. J Stat Comput Simul 67(4):359–376
Muler N, Yohai VJ (2008) Robust estimates for GARCH models. J Stat Plan Inference 138(10):2918–2940
Ossandón S, Bahamonde N (2011) On the nonlinear estimation of GARCH models using an extended Kalman filter. In: Lecture Notes in Engineering and Computer Science. Proceedings of the world congress on engineering, WCE 2011, 6–8 July, 2011, London, UK, pp 148–151
Pelagatti M, Lisi F (2009) Variance initialisation in GARCH estimation. In: Complex data modeling and computationally intensive statistical methods for estimation and prediction, S.Co 2009, 14–16 September, 2009, Milan, Italy, Maggioli Editore, pp 323–328
Penzer J (2007) Inference for GARCH and related models with incomplete data, Working Paper, LSE
Proietti T (2008) Missing data in time series: a note on the equivalence of the dummy variable and the skipping approaches. Stat Probab Lett 78(3):257–264
Riani M (2004) Extensions of the forward search to time series. Stud Nonlinear Dyn Econom 8(2), article 2
Riani M, Cerioli A, Atkinson AC, Perrotta D (2014) Monitoring robust regression. Electron J Stat 8(1):646–677
Riani M, Perrotta D, Cerioli A (2015) The forward search for very large datasets. J Stat Softw 67(1):1–20
Roelant E, Van Aelst S, Willems G (2009) The minimum weighted covariance determinant estimator. Metrika 70(2):177
Rousseeuw P, Leroy A (1987) Robust regression and outlier detection. Wiley, New York
Sakata S, White H (1998) High breakdown point conditional dispersion estimation with application to S&P 500 daily returns volatility. Econometrica 66(3):529–567
Salini S, Cerioli A, Laurini F, Riani M (2016) Reliable robust regression diagnostics. Int Stat Rev 84(1):99–127
Van Dijk D, Franses PH, Lucas A (1999) Testing for ARCH in the presence of additive outliers. J Appl Econom 14(5):539–562
Acknowledgements
The authors are grateful to two anonymous referees and the Editor who gave valuable suggestions to improve the paper. They also wish to thank all the participants at the final conference of Project MIUR (2012) in Benevento, at Sco Meeting (2013) in Milan, and at the meeting on Robust Statistics at the Joint Research Centre of the European Commission in Ispra (2014) where preliminary versions of the present work were presented. All comments and suggestions have been considered and have contributed to improve the quality of the manuscript. We thank Matteo M. Pelagatti for the use of his R-codes for GARCH estimation. This work was supported by the project MIUR 2012-MISURA and by the Department of Economics (University of Verona), Grant FAR/2015. The usual disclaimer applies.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Crosato, L., Grossi, L. Correcting outliers in GARCH models: a weighted forward approach. Stat Papers 60, 1939–1970 (2019). https://doi.org/10.1007/s00362-017-0903-y
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-017-0903-y