摘要
2017年新疆人民政府工作报告提出2017年将完成全社会固定资产投资1.5万亿元以上,同比增长50%。作为固定资产投资重要资金来源的银行信贷将增长多少成为社会各界关心的问题。本文针对新疆2017年固定资产投资可能增长50%这一突变情况,采用多元线性回归模型、VAR模型和面板模型等3种计量模型来预测新疆信贷增速。研究发现,3种模型的样本外预测精度均较为理想。结合模型预测结果和形势判断,本文更倾向于面板模型的预测结果:2017年新疆各项贷款新增1882亿元、增长12.4%。
In 2017 Xinjiang government work report, it indicates that the fixed asset investment will reach 1.5 trillion and increase by 50% in 2017. The public majority are concerned about the credit growth, which is an important capital source for the fixed asset investment. This paper adoptes three classical ecometric models to predict the potential credit growth of 2017 if the fixed asset investment in Xinjiang increase by 50%. The three models are MLR( mutiple linear regression)model, VAR model and panel data model. The results of the study show that the out-of-sample prediction performance of the proposed models are good. Based on these forecasting results of models and the situation analysis, actually this paper is more inclined to the result of panel data model: the loan of Xinjiang in 2017 will add 188.2 billion yuan and increase by 12.4%.
出处
《金融发展评论》
2017年第8期1-13,共13页
Financial Development Review
