Smart risk management with financial big data

Smart risk management is a new research area to combine artificial intelligence with traditional quantitative risk management model to leverage financial big data in secondary market more effectively. Usually, the construction of traditional quantitative risk management model is not fully automatic from financial data. Industry classification, parameter estimation of factor returns covariance matrix and eliminate random noise in factor correlation coefficient matrix are usually decided by modeler's experience. The industry classification in this paper was made based on a machine learning methodology called quantitative clustering method from stock analyst forecast big data. Half-life parameter for factor returns covariance matrix is dynamically estimated from trading data intelligently combined with random matrix theory (RMT). Finally, this paper gives several numerical test results to verify the effectiveness of smart risk management methods in Chinese financial markets.