Abstract Being more and more popular in the past 10 years, Bitcoin has drawn extensive attention from the press, scholars, and practitioners. The aim of this paper is to investigate which predictor is more predictive for Bitcoin volatility from the aspects of in-sample and out-of-sample in a high-speed changing world. We utilise the GARCH-MIDAS model to examine the predictive power of five crucial predictors, including VIX, GVZ, Google Trends, GEPU, and GPR. Our findings provide strong evidence that GVZ exhibits strongest predictability for Bitcoin volatility over other competing predictors. Other empirical results based on different out-of-sample forecasting periods, alternative loss functions and combination methods further ensure our major conclusions are robust.
Beschreibung
Which predictor is more predictive for Bitcoin volatility? And why? - Liang - - International Journal of Finance & Economics - Wiley Online Library
%0 Journal Article
%1 https://doi.org/10.1002/ijfe.2252
%A Liang, Chao
%A Zhang, Yaojie
%A Li, Xiafei
%A Ma, Feng
%D 2020
%J International Journal of Finance & Economics
%K ba bitcoin volatility
%N n/a
%R https://doi.org/10.1002/ijfe.2252
%T Which predictor is more predictive for Bitcoin volatility? And why?
%U https://onlinelibrary.wiley.com/doi/abs/10.1002/ijfe.2252
%V n/a
%X Abstract Being more and more popular in the past 10 years, Bitcoin has drawn extensive attention from the press, scholars, and practitioners. The aim of this paper is to investigate which predictor is more predictive for Bitcoin volatility from the aspects of in-sample and out-of-sample in a high-speed changing world. We utilise the GARCH-MIDAS model to examine the predictive power of five crucial predictors, including VIX, GVZ, Google Trends, GEPU, and GPR. Our findings provide strong evidence that GVZ exhibits strongest predictability for Bitcoin volatility over other competing predictors. Other empirical results based on different out-of-sample forecasting periods, alternative loss functions and combination methods further ensure our major conclusions are robust.
@article{https://doi.org/10.1002/ijfe.2252,
abstract = {Abstract Being more and more popular in the past 10 years, Bitcoin has drawn extensive attention from the press, scholars, and practitioners. The aim of this paper is to investigate which predictor is more predictive for Bitcoin volatility from the aspects of in-sample and out-of-sample in a high-speed changing world. We utilise the GARCH-MIDAS model to examine the predictive power of five crucial predictors, including VIX, GVZ, Google Trends, GEPU, and GPR. Our findings provide strong evidence that GVZ exhibits strongest predictability for Bitcoin volatility over other competing predictors. Other empirical results based on different out-of-sample forecasting periods, alternative loss functions and combination methods further ensure our major conclusions are robust.},
added-at = {2021-12-18T16:25:16.000+0100},
author = {Liang, Chao and Zhang, Yaojie and Li, Xiafei and Ma, Feng},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/206089515cbdbaa0a1351d846f98510c4/georglender},
description = {Which predictor is more predictive for Bitcoin volatility? And why? - Liang - - International Journal of Finance & Economics - Wiley Online Library},
doi = {https://doi.org/10.1002/ijfe.2252},
eprint = {https://onlinelibrary.wiley.com/doi/pdf/10.1002/ijfe.2252},
interhash = {278de75f170a1fe700bf5e057ce08b72},
intrahash = {06089515cbdbaa0a1351d846f98510c4},
journal = {International Journal of Finance \& Economics},
keywords = {ba bitcoin volatility},
number = {n/a},
timestamp = {2021-12-18T15:25:16.000+0100},
title = {Which predictor is more predictive for Bitcoin volatility? And why?},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/ijfe.2252},
volume = {n/a},
year = 2020
}
