News Sentiment Analysis for Liquidity Risk Reduction

Abstract

Abstract Recently, the low quality of banks' assets has caused many problems for banks and the economy of some countries. The lack of efficient risk management is the main reason for the decline in the quality of banks' assets. One of the most important risks in banks is liquidity risk. The Basel Committee on Banking Supervision (BCBS) introduced the liquidity coverage ratio (LCR) as part of the Basel III reforms for the short-term recovery of internationally active banks against liquidity shocks, after the crisis. As the LCR is designed to ensure that banks hold a sufficient reserve of high-quality liquid assets, predicting LCR position at the right time can prevent banking serious problems in the future which has not been addressed in previous research. This study applies the Sentiment Analysis approach as qualitative measures and investigated its impact on LCR. A news integration method combined with text representation techniques, the result of which is fed into well-tuned Deep Learning Algorithms such as Convolutional Neural Network, is introduced as a unique predictive machine suitable for this problem. Finally, we find that current news contain information related to the change in LCR of next month, an insight that helps to improve prediction of its position with an accuracy of 88.6%. The outcome indicates that risk managers can benefit from complementing established qualitative measures that are related to future liquidity risk changes.