hrcak mascot   Srce   HID

Izvorni znanstveni članak
https://doi.org/10.17535/crorr.2014.0007

A comparison of four different block bootstrap methods

Boris Radovanov ; Faculty of Economics Subotica, University of Novi Sad
Aleksandra Marcikić ; Faculty of Economics Subotica, University of Novi Sad

Puni tekst: engleski, pdf (209 KB) str. 189-202 preuzimanja: 5.315* citiraj
APA 6th Edition
Radovanov, B. i Marcikić, A. (2014). A comparison of four different block bootstrap methods. Croatian Operational Research Review, 5 (2), 189-202. https://doi.org/10.17535/crorr.2014.0007
MLA 8th Edition
Radovanov, Boris i Aleksandra Marcikić. "A comparison of four different block bootstrap methods." Croatian Operational Research Review, vol. 5, br. 2, 2014, str. 189-202. https://doi.org/10.17535/crorr.2014.0007. Citirano 17.06.2019.
Chicago 17th Edition
Radovanov, Boris i Aleksandra Marcikić. "A comparison of four different block bootstrap methods." Croatian Operational Research Review 5, br. 2 (2014): 189-202. https://doi.org/10.17535/crorr.2014.0007
Harvard
Radovanov, B., i Marcikić, A. (2014). 'A comparison of four different block bootstrap methods', Croatian Operational Research Review, 5(2), str. 189-202. doi: https://doi.org/10.17535/crorr.2014.0007
Vancouver
Radovanov B, Marcikić A. A comparison of four different block bootstrap methods. Croatian Operational Research Review [Internet]. 2014 [pristupljeno 17.06.2019.];5(2):189-202. doi: https://doi.org/10.17535/crorr.2014.0007
IEEE
B. Radovanov i A. Marcikić, "A comparison of four different block bootstrap methods", Croatian Operational Research Review, vol.5, br. 2, str. 189-202, 2014. [Online]. doi: https://doi.org/10.17535/crorr.2014.0007

Sažetak
The paper contains a description of four different block bootstrap methods, i.e., non-overlapping block bootstrap, overlapping block bootstrap (moving block bootstrap), stationary block bootstrap and subsampling. Furthermore, the basic goal of this paper is to quantify relative efficiency of each mentioned block bootstrap procedure and then to compare those methods. To achieve the goal, we measure mean square errors of estimation variance returns. The returns are calculated from 1250 daily observations of Serbian stock market index values BELEX15 from April 2009 to April 2014. Thereby, considering the effects of potential changes in decisions according to variations in the sample length and purposes of the use, this paper introduces stability analysis which contains robustness testing of the different sample size and the different block length. Testing results indicate some changes in bootstrap method efficiencies when altering the sample size or the block length.

Ključne riječi
Block bootstrap; returns; stability analysis

Hrčak ID: 133696

URI
https://hrcak.srce.hr/133696

Posjeta: 5.530 *