An integrated early warning system for stock market turbulence

作者：

Highlights：

• Stock turbulence is identified by the SWARCH high volatility regime.

• Crisis cutoffs are dynamically determined by the two-peak method.

• The LSTM network predicts stock crises with a test-set accuracy of 96.4%.

• An average of 2.8 days forewarned period is achieved by the proposed EWS.

摘要

•Stock turbulence is identified by the SWARCH high volatility regime.•Crisis cutoffs are dynamically determined by the two-peak method.•The LSTM network predicts stock crises with a test-set accuracy of 96.4%.•An average of 2.8 days forewarned period is achieved by the proposed EWS.

论文关键词：Early warning system,LSTM,SWARCH,Two-peak method,Dynamic prediction

论文评审过程：Received 24 December 2019, Revised 29 March 2020, Accepted 16 April 2020, Available online 18 April 2020, Version of Record 23 April 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113463