Systemic risk spillovers incorporating investor sentiment：Evidence from an improved TENET analysis
Description
The folder "code and data" contains the code for data processing and empirical results. It includes two folders, data is used to store data, and model is used to store running python and R code. 1.Data Description： 1.1.The folder "TENET network data at each time point" stores the adjacency matrix and other data of each time node in the TENET network. It is called in "Network topology analysis.R". 1.2.Ping An Bank Investor Sentiment (Bayesian Machine Learning).csv is Ping An Bank's investor sentiment data based on machine learning methods 1.3.Ping An Bank Investor Sentiment (Financial Dictionary).csv is Ping An Bank's investor sentiment data based on Financial Dictionary methods 1.4.Ping An Bank Investor Sentiment (Pre-trained Deep Learning (ERNIE)).csv is Ping An Bank's investor sentiment data based on ERNIE model. 1.5aligned_sentiment_indices.csv stores variables related to market sentiment, among which ISI, CICSI and Confidence index are derived from the CSMAR database, and BI is the investor sentiment index calculated by ERNIE based on Baidu AI platform. 1.6 The IIC.csv file contains data on tail risk spillovers within the financial sector. 1.7 The DS.csv file contains data on tail risk spillovers between any financial sector of a financial institution and any other financial sector. 1.8 The BIC.csv file contains data on how much risk each sector spillsover to others. 1.9 The BIC_receive.csv contains data on how much risk each sector receives from others. 1.10 The three files HHI.csv, NAS.csv, and AS.csv store network topology indicator data. 1.11 The code number.xlsx store the stock codes and abbreviations of all financial institutions. 1.12 The Stock Market Value.csv is the market value data of financial institutions, which is used to identify Systemically Important Financial Institutions (Härdle et al. (2016)). 2.Figure： 2.1Figure 1 can be obtained through the ''Sentiment Comparison of Three Approaches for Individual Financial Institutions.py''. 2.2Figure 2 can be obtained via ''Comparison of Market sentiment.py''. 2.3Figures 3 can be obtained through ''Change in average λ for systematic risk (compare to inclusion of sentiment variables).py''. 2.4Figure 4 requires you to choose to run ''Comparison of elemental standardisation treatments for TENET.py''. 2.5Figure 5 requires you to choose to run ''Comparison of average λ and spillover intensity.py''. 2.6Figure 6-11 are obtained by running ''Network topology analysis.R''.The same procedure is also run for Tables 5 and 6 concerning the rankings of risk emitters and receivers. 2.7Figure 12 is obtained by running ''Evolution of Cross-Sector Tail Risk Spillovers and Spill-Ins.py''. 2.8Figure 13 is obtained by running ''Tail risk spillovers between any financial sector of a financial institution and any other financial sector.py''. 2.9Figure 14 is obtained by running ''Tail Risk Spillovers within the Financial Sector.py''.