统计与数据科学学院知识库

我国在十九大报告己经明确提出要健全货币政策和宏观审慎政策双支柱调控框架，保持经济稳健发展和金融市场的稳定是我国目前的两个重要目标，其中资产价格特别是房价可能存在的泡沫风险，会对金融市场稳定造成重大影响。鉴于核心CPI可以更好地反映总体物价水平，且贝叶斯TVP-VAR模型可以从时变层面对货币政策的效应进行分析，故本文主要研究内容分为两个部分，其一是货币政策的价格效应，其二是货币政策对房地产价格的冲击分析。

关于我国货币政策的价格冲击效应方面，鉴于核心CPI是从货币政策的角度定义的，本文探究了我国货币政策的价格效应在不同时点上的差异。首先采用小波分解剔除短期波动得到八类价格指数的长期趋势。然后针对长期价格趋势运用动态因子模型提取它们共同的部分，并将其定义为核心CPI。最后将核心CPI作为价格变量，针对M2、核心CPI等变量通过构建贝叶斯框架下的时变参数向量自回归（TVP-VAR）模型，测度了我国货币政策在不同时点上的价格效应，并与货币政策对CPI的冲击进行比较。实证结果表明：我国货币政策对整体物价的调控具有一定的短期效应，且在不同的时点上存在显著的差异性；核心CPI可以更好的作为制定货币政策时的参考。

关于我国货币政策的房价冲击效应方面，通过贝叶斯框架下的TVP-VAR模型分析了数量型货币政策和价格型货币政策对房价的冲击效应，实证结果表明：不管是金融危机时期，还是经济发展新常态时期，两种货币政策对房地产价格的调控都表现出较好的效果，且房价也在短时间做出反馈；同时经济发展处于低迷期时，价格型货币政策对于房价的冲击效应大于数量型货币政策。

总之，通过本文的研究，核心CPI可以更好地代表我国总体的物价水平，应构建更加完善的核心CPI体系，为制定货币政策提供更好的依据；基于TVP-VAR模型对货币政策价格冲击效应和房价冲击效应的时变特征进行了刻画，为政府健全货币政策和宏观审慎政策双支柱调控提供参考。

In the report of the 19th National Congress of the Communist Party of China, our country has clearly stated that it is necessary to improve the dual-pillar regulatory framework of the monetary policy and macro-prudential policy, and to maintain the stable development of the economy and the stability of the financial market are two important goals of our country at present. Among them, asset prices, especially housing prices, may exist the risk of bubbles will have a significant impact on the stability of financial markets. In view of the fact that the core CPI can better reflect the overall price level, and the Bayesian TVP-VAR model can analyze the effect of monetary policy from the time-varying level, the main research content of this paper is divided into two parts. One is the price effect of monetary policy, and the other is the impact analysis of monetary policy on real estate price.

Regarding the price shock effect of our country's monetary policy, given that the core CPI is defined from the perspective of monetary policy, this paper explores the difference in the price effect of our country's monetary policy at different time points. Firstly, wavelet decomposition is used to eliminate short-term fluctuations to obtain the long-term trends of eight types of price indices. Then, their common components of long-term price trends are extracted by using the dynamic factor model and defined as core CPI. Finally, taking the core CPI as the price variable, the time-varying parameter vector autoregression (TVP-VAR) model under the Bayesian framework was constructed for variables such as M2 and core CPI to measure the price effect of our country's monetary policy at different time points. And compare with the impact of monetary policy on CPI. The empirical results show that: our country's monetary policy has a certain short-term effect on the overall price regulation, and there are significant differences at different time points; the core CPI can be better used as a reference when formulating monetary policy.

Regarding the impact effect of our country's monetary policy on housing prices, the impact effects of quantitative monetary policy and price-based monetary policy on housing prices are analyzed through the TVP-VAR model under the Bayesian framework. The empirical results show that: Whether it’s the financial crisis or the new normal of economic development, both monetary policies have shown good effects on the regulation of real estate prices, and housing prices also give feedback in a short time at the same time, when economic development is in the downturn, the impact effect of price-based monetary policies on housing prices is greater than that of quantitative monetary policy.

In a word, through the research of this paper, the core CPI can better represent our country's overall price level, and a more comprehensive core CPI system should be constructed to provide a better basis for formulating monetary policy; based on the TVP-VAR model, this paper describes the time-varying characteristics of monetary policy price shock effect and housing price shock effect, which provides a reference for the government to improve the monetary policy and the dual-pillar regulation of the macro-prudential policy.