The Measurements Of Systemic Financial Risk And Its Impacts On Macroeconomy Forecasts And Asset Pricing

In the past decade,many events in Chinese financial markets have been enough to arouse people's attention,such as the abnormal fluctuation of stock market in 2008 and 2015,and the rapid rising of interest rates in 2013 and 2016.The fluctuation of the financial market had aroused the high attention of the Communist Party of China and the government of the People's Republic of China.The nineteenth Congress of the Communist Party of China clearly demanded that we should improve the financial supervision system and ensure that no systemic occur.In order to achieve this goal,it is necessary to research on the measurement of systemic financial risk and its impacts on macroeconomy forecasts and asset pricing.This paper attempted to supplement the existing literature in these areas.The contribution of this article is mainly embodied in three aspects: first,it established comprehensive indexes to measure systemic financial risk which can reflect the level of systemic financial risk more comprehensively.Second,it built a prediction model based on the systemic financial risk indexed to forecast the future economic situation and market trend.Last,this paper analysised how systemic financial risk affects the yield of assets like bonds and stocks,which would supplement the shortcomings of the existing asset pricing theories about systemic financial risks.Eighteen indicators of systemic risk were constructed,covering four major categories of institution-specific risk,comovement and contagion,volatility and instability,liquidity and credit.By using two method,namely the principal component quantile regression and partial quantile regression,this article synthesized single indicators into two types of comprehensive indicators,which contain more information about systemic risk.And analyzing the relationship between these indicators and the trend of macroeconomic and financial variables one by one,many intuitive discoveries were found.For example,the WDD may have its own specific advantages for the prediction of the "big crash" in the stock market.Although the CoVaR can forecast the big reversal of the stock market too,but lagging behind the WDD.After constructing the forecast model and forecast accuracy index,the predictive power of single indicators and comprehensive indexes on macroeconomy had been tested.The results showed that the forecasting model constructed based on systemic risk measures has poor predictive power for GDP growth and industrial added value growth.However,if using the model to forecast the change of GDP growth or industrial added value growth,the predictive power would be significantly improved.Especially in the forecast of extreme cases of "hard landing" of the economy,they showed far better characteristics than the autoregressive model.Through the horizontal comparison between the indicators or idexes,it's found that the predictive power of the comprehensive index is better than almost all of single indicators.Meanwhile,the prediction effect of the partial quantile regression method is slightly lower than that of the principal component quantile regression method.And what should be noted is that although these measures and models have good predictive power for the economic downturn,it is difficult to predict the economic recovery and prosperity.This paper also used systemic risk measures to forecast the return of the stock market and the results showed that the measures also have predictive power for the extreme cases of the stock market.This means there may be some interesing relationships between systemic risk and asset pricing,which call for more analysis.These researches will help policymakers to anticipate changes in the economic situation from a systemic risk perspective,so as to enhance the foresight and scientificity of economic policy.The estimation results of the Two-tier Stochastic Models showed that in the bond market,3-month repurchase interest rates were slightly elevated;3-month treasury bond rates were slightly lower;while the 10-year treasury bond rate and 10-year AAA corporate bond rates were obviously overvalued.Based on four rates above,the liquidity spread,the term spread and the default spread can be obtained.Further analysis of the spreads showed that there is a two-way Granger causality between the pricing deviation of the rates and systemic risk.What's more,the pricing deviation of different bond rates would significantly affect the corresponding spread,and systemic risk could exacerbate these effects;The impact of pricing deviation on the three spreads during the monetary tightening will be slightly greater than during the monetary easing.These conclusions may not only complement the existing literature about the bond rate deciding,but also help investors optimize the bond portfolio from a systemic risk perspective.For the stock market,this paper had measured the systemic financial risk exposure of the stock,and then use Portfolio-based analysis to research the relationship between the systemic financial risk exposure and stock returns.The results showed that while using the market value weighted average method to calculate portfolio yield,the stock portfolio with the higher degree of systemic financial risk exposure in the cross section will have a higher return rate too,which is consistent with the expectation of classical theory.Further analysis also found that larger companies are more likely to have a higher degree of systemic financial risk exposure,and as the degree of systemic risk exposure increases,the portfolio's sensitivity to market factors and value factors also increases.These findings,while improving existing asset pricing theory,also help prompt investors to focus on systemic risks.In the stock pricing system,we should add systemic risk exposure factor to multiple factor models,so as to improve the effectiveness of pricing.