宏观经济信息与金融市场关联性——来自混频动态条件相关系数模型的证据

宏观经济信息是金融市场之间相互传递的重要信息内容,有效利用宏观经济信息是否有助于更好地理解金融市场关联性？为此,本文运用混频动态条件相关系数(DCC-MIDAS)模型分析了我国四个重要金融市场之间的动态相关性如何受到纳入的宏观经济信息的影响。结果发现:(1)工业增加值和货币供应量M2负向影响金融市场关联性,经济政策不确定性和通货膨胀水平反之。前三者是影响金融市场关联性较为重要的因素,而通货膨胀的重要性相对较低;(2)宏观经济信息作为市场关联性的长期成分相较基于市场信息的模型可以获得至少1.45%的效率提升。(3)工业增加值和通货膨胀对金融市场关联性的影响较为稳健,货币供应量M2和经济政策不确定性的影响表现出周期性特征。经济上行阶段工业增加值、货币政策信息带来的效率提升更为明显,经济下行阶段政策不确定性相对重要。本文研究结论对于加强金融监管协调和防范金融市场风险共振具有参考价值和指导意义。

Macroeconomic Information and Financial Market Connectedness: Evidence from A DCC-MIDAS Model

Understanding the correlation between financial markets is key for the effective implementation of coordinated supervision. The financial sector is the bloodline of the real economy, and enhancing the capacity of financial services for the real economy requires the coordination and alignment of financial markets with the macroeconomy. A well-informed and well-run financial market should be able to effectively reflect the characteristics of the macroeconomy. Making good use of macroeconomic information and accurately estimating dynamic correlations between financial markets are important for the accurate implementation of cross-market coordinated supervision and the efficient monitoring and early warning of risk resonance between markets.This paper uses the DCC-MIDAS model to incorporate macro-level low-frequency variables into the analytical framework of high-frequency correlations among financial markets and uses the covariance matrix estimation accuracy comparison method proposed by Engle and Colacito (2006) and Laurent et al. (2013) to compare the macroeconomic information model and the long-term market volatility information model in terms of the estimation efficiency of dynamic conditional correlation coefficients. Additionally, it systematically investigates and discusses the cyclical characteristics of the impact of each macroeconomic variable on the correlation between financial markets. This paper presents direct evidence of the impact of macroeconomic information on financial market correlations and bridges a research gap regarding multi-market correlations. The study more accurately captures the causal factors causing financial market resonance, and the mechanism of macroeconomy influencing financial risk, provides feasible ideas for the implementation of coordinated cross-market regulation in the context of different shocks, and presents a basic framework for the construction of real-time monitoring indicators for financial risk mixing by combining macroeconomic information and financial market data.This paper uses monthly data of China's industry value added, M2, consumer price index, and economic policy uncertainty index from January 2006 to June 2018, totaling 150 sample points, and daily yield data of the stock market, money market, foreign exchange market, and bond market during the corresponding sample period, totaling 3,258 sample points. The following findings are obtained:(1) Industry value added and M2 negatively affect financial market correlations, and economic policy uncertainty and inflation levels conversely. The robustness of the results is not affected by the way macroeconomic information is introduced into the financial market correlation analysis framework; (2) AS macroeconomic information is a long-run component of market correlations, the macroeconomic information model achieves an efficiency increase of at least 1.45% over other models based on market information. Real economic performance, economic policy uncertainty, and M2 are the most important factors affecting financial market correlations, whereas inflation is less important; and (3) the impact of industry value added and inflation on financial market correlations is relatively robust, whereas economic policy uncertainty and M2 show cyclical characteristics. During economic upturns, loose monetary policy is more likely to trigger financial market correlations, economic policy uncertainty inhibits financial market correlations, and the efficiency improvements brought by industry value added and monetary policy information are larger. Meanwhile, the efficiency gains from economic policy uncertainty and monetary policy information are higher during economic downturns.Based on the main findings, the following policy recommendations are proposed: (1) cross-market coordination and supervision should be informed by the role of macroeconomic factors in the analysis of inter-market correlations and prevent the upside of financial market correlations in advance. Gradually form a systematic set of financial market risk resonance monitoring indicators; (2) cross-market coordinated supervision should be informed by the dynamics of the economic cycle, flexibly apply policy tools according to the stage of the economic cycle, curb financial market risk resonance with macro policies, and dynamically adjust the proportional weights of each macro variable in the monitoring indicators; and (3) pay attention to the channel role of macroeconomic information in market connections, gradually release macroeconomic data, focus on its economic signal role, be wary of the market transmission of market panic, and guide investors to pay attention to the operation of China's macroeconomic fundamentals.