基于MF-BVAR模型的中国宏观经济混频递归实时预测和传导机制研究

doi:10.16088/j.issn.1001-6597.2020.05.010

摘要/Abstract

摘要： 传统VAR模型需要变量保持同频的限制,这使得原始数据的完整性遭到破坏,从而影响模型的性能。本研究引入可以融合季度变量和月度变量的MF-BVAR模型,以GDP、PPI、CPI和PMI的递归样本为例并将其划分为三组不同的季度内信息组,比较了其与MIDAS模型和同频模型的性能,且通过混频格兰杰因果关系检验、脉冲响应分析和方差分解分析剖析了变量之间的传导机制。实证结果表明:(1)MF-BVAR模型关于PPI、CPI、PMI和GDP的实时预测误差分别较QF-BVAR模型降低约70%、80%、75%和20%;(2)基于定量和统计检验的角度均表明MIDAS和MF-BVAR模型的实时预测和短期预测性能均显著优于QF-BVAR模型,且MF-BVAR模型更适合于实时预测而MIDAS在短期预测表现更优,此外混频递归样本表明月度信息利用越充分,模型的预测误差越小;(3)GDP与CPI、PPI和PMI、CPI与PMI之间存在双向的因果关系,且传导时长呈现差异性,而PMI和GDP、PPI与GDP、PPI和CPI的正向传导时长均为3期,反向传导路径阻塞,存在不对称性。

关键词: MF-BVAR模型, 中国宏观经济混频数据, 实时预测, PMI, PPI, CPI

Abstract: The traditional VAR model limits the variables to the same frequency, which destroys integrity of the original data, thus affecting performance of the model. In this study, MF-BVRA model, which well integrates quarterly and monthly variables, is introduced. The recursive samples of GDP, PPI, CPI and PMI are divided into three of different quarterly information groups. The performance of MF-BVAR is compared with that of MIDAS and co-frequency model. The transmission mechanism among variables is analyzed by mixing Granger Causal Relation Test, impulse response analysis and variance decomposition analysis. The empirical results show that: 1) compared with QF-BVAR, MF-BVAR’s real-time prediction errors in PPI, CPI, PMI and GDP are about 70%, 80%, 75% and 20% lower respectively; 2) real-time and short-term prediction performance of MIDAS and MF-BVAR, quantitative and statistical tests tell, are significantly better than that of QF-BVAR, and the latter does better in real-time forecasting, while MIDAS is good for short-term forecasting; the more monthly information is fully used, the smaller the prediction error will be, the recursive samples tell; (3) there is a bidirectional causal relationship between GDP and CPI, PPI and PMI, CPI and PMI, and the conduction time is different; while the forward conduction duration of PMI and GDP, PPI and GDP, PPI and CPI are all three periods, and the reverse conduction path is blocked without an asymmetry.

Key words: MF-BVAR Model, mixing data of China’s Macro economy, real-time prediction, PMI, PPI, CPI

中图分类号:

F222.1

桂文林, 程慧. 基于MF-BVAR模型的中国宏观经济混频递归实时预测和传导机制研究[J]. 广西师范大学学报（哲学社会科学版）, 2020, 56(5): 105-127.

GUI Wen-lin, CHENG Hui. Research on Real-time Prediction and Transmission Mechanism of Mixing Recursion of China’s Macro Economy Based on MF-BVAR Model[J]. Journal of Guangxi Teachers Education University (Philosophy and Social Sciences Edition), 2020, 56(5): 105-127.

参考文献

[1] 王双正. 基于VAR模型的通货膨胀与经济增长关系研究[J].经济理论与经济管理,2009(1):21-27.
[2] 周文,赵果庆. 中国GDP增长与CPI:关系、均衡与“十二五”预期目标调控[J].经济研究,2015(5):4-17.
[3] 隋占林. 中国经济增长与通货膨胀预测及影响因素研究[D].大连:东北财经大学经济学院,2016.
[4] 刘金全,张鹤. 经济增长风险的冲击传导和经济周期波动的“溢出效应”[J].经济研究,2003(10):32-39+91.
[5] 刘金全,郑挺国. 我国货币政策冲击对实际产出周期波动的非对称影响分析[J]. 数量经济技术经济研究,2006(10):3-14.
[6] 齐丽云,何跃. 基于PMI和PPI的GDP预测模型[J].统计与决策,2013(16):12-14.
[7] 杨兴武,沈颂东. 中国采购经理指数(PMI)对GDP影响的传导路径分析[J].统计与决策,2016(5):124-128.
[8] Foroni, C, Ghysels, E, Marcellino, M. Mixed-frequency vector autoregressive models[J].Advances in Econometrics,2013(6):247-272.
[9] Camacho M. Mixed-frequency models with Markov-switching dynamics[J].Economics Letters,2013 (12):369-373.
[10] Foroni C, Guerin P, Marcellino M. Markov-switching mixed-frequency VAR models[J]. International Journal of Forecasting,2015(3):692-711.
[11] 刘金全,解瑶姝,龙威. “新常态”下中国财政政策与货币政策选择的新视角[J].经济问题,2016(3):14-21.
[12] 张劲帆,刚健华,钱宗鑫,等. 基于混频向量自回归模型的宏观经济预测[J].金融研究,2018(7):34-48.
[13] Schorfheide F, Song Dongho. Real-time forecasting with a fixed-frequency VAR[J].Journal of Business& Economic Statistics,2015 (10):366-380.
[14] Zadrozny P. Gaussian likelihood of continuous-time ARMAX models when data are stocks and flows at different frequencies[J].Econometric Theory,1988 (4):108-124.
[15] Mariano R S, Murasawa Y. A coincident index, common factors and monthly real GDP[J].Oxford Bulletin of Economics and Statistics, 2010 (2):27-46.
[16] Kuzin V, Marcellion M, Schumacher C. MIDAS vs. mixed-frequency VAR: Now casting GDP in Euro area[J].International Journal of Forecasting,2011 (4):529-542.
[17] Foroni C, Marcellino M. A comparison of mixed frequency approaches for now casting Euro area macro- economic aggregates[J].International Journal of Forecasting, 2014 (7):554-568.
[18] Ghysels E. Macroeconomics and the reality of mixed frequency data[J]. Journal of Econometrics, 2016 (2):294-314.
[19] Ghysels E, Hill J B, Motegi K. Testing for Granger causality with mixed frequency data[J]. Journal of Econometrics, 2016 (1):207-230.
[20] 陈浩东. 基于混频数据模型对通货膨胀与经济增长影响的实证研究[D].长春:吉林大学商学院,2012.
[21] 刘汉,刘营,王永莲. 经济景气指标与实际GDP增长率的混频预测[J].统计与决策,2017(21):29-33.
[22] 周德才,邓姝姝,左玥. 中国金融状况指数混频编制与应用研究——基于MSMF-VAR模型的一个经验分析[J].南开经济研究,2018(2):148-163.
[23] 顾光同,许冰. 中国工业增加值的半月预报:基于宏观月度数据[J].系统工程理论与实践,2018(8):183-1993.
[24] Nergo M D, Schorfheide F and Wouters R. On the fit of new Keynesian models[J].Journal of Business and Economic Statistics,2007 (1):123-162.
[25] Sims C A. Macro economics and reality[J].Econometrica,1980(1):1 -48.
[26] Geweke J. Using simulation methods for Bayesian econometric models: inference, development and communication[J].Econometric Reviews,1999 (1):1-126.
[27] Carter C K and Kohn R. On Gibbs sampling for state space models[J]. Biometrika, 1994(3):541-553.
[28] 刘金全,刘汉,印重. 中国宏观经济混频数据模型应用——基于MIDAS模型的实证研究[J].经济科学,2010(5):23-34.
[29] 宋科进. 我国制造业PMI、非制造业PMI与经济增长关系的实证研究[J]. 发展研究,2014(10):8-12.
[30] 解瑶姝. “新常态”时期中国经济增长与通货膨胀的关联机制研究[J].经济问题探索,2018(12):11-20.
[31] Negro M D and Schorfheide F. Bayesian Macro-econometrics[M].Oxford University Press, The Oxford Handbook of Bayesian Econometrics,2011.
[32] Doan T, Litterman R B and Sims C A. Forecasting and Conditional Projection Using Realistic Prior Distributions[J].Econometric Reviews,1984(1):1-100.
[33] Foroni C, Marcellino M and Schumacher C.U-MIDAS: MIDAS regression with unrestricted lag polynomials[J].Journal of the Royal Statistical Society,2013(12):57-82.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed