摘要
文章用隐因子模型和无套利简约宏观金融模型分别对中国国债收益率曲线进行拟合和预测,发现将宏观因子加入利率期限模型后,国债收益率曲线的长端拟合和预测效果均有提升。经脉冲响应和方差分解分析后发现,CPI增长率主要通过影响短期收益率来影响收益率曲线,而工业增加值增长率主要通过影响长期收益率来影响收益率曲线。此外,简约宏观金融模型可以形成对宏观经济的预测,尤其是对工业增加值增长率的预测较为准确。
This paper applies the latent factor model and the no-arbitrage parsimonious macro-finance model to respectively fit and predict China＇ s bond yield curve, and discovers that both the prediction effect and fitting of the long end of yield curve are improved after macro factors are added to the term model. Impulse response and variance decomposition analysis find that CPI growth rate influences the yield curve mainly through influencing the short end, while industry value added growth rate via the long end. In addition, the parsimonious macro-finance model can he used to predict macro economy, especially accurately for industry value added growth rate prediction.
出处
《统计与决策》
CSSCI
北大核心
2017年第17期159-164,共6页
Statistics & Decision
关键词
简约宏观金融模型
国债收益率
宏观因子
parsimonious macro-finance model
bond yield rates
macro factors
