Abstract: This paper constructs an insurance industry stress index to measure the systemic risk of China’s insurance industry, and predicts the future risks of the insurance industry by predicting the future trend of the insurance industry’s stress index. Drawing on the EEMD decomposition integration technology in the TEI@I complex system research methodology, combined with the current research results in the field of artificial intelligence and deep learning, the EEMD-LSTM model is constructed to carry out the systemic risk warning of the insurance industry. The empirical results show that EEMD decomposition integration technology has obvious advantages in predicting nonlinear and non-stationary complex time series. At the same time, LSTM model can effectively describe the dependence and long memory characteristics between time series. The EEMD-LSTM model combination prediction accuracy is superior to other models.

Key words: insurance, systemic risk, early warning, ensemble empirical mode decomposition, long short-term memory neural network