摘要
本文利用百度搜索大数据、金融风险指标滞后项以及结构数据变量滞后项建立金融风险预测模型,研究发现:(1)包含百度搜索大数据的风险预测模型有助于提升金融风险预测的准确度,并且模型在金融风险上升时期的预测效果要好于金融风险下降时期的预测效果。(2)公众搜索宏观经济类风险、银行机构类风险和互联网金融类风险产生"追涨杀跌"的非理性行为,搜索非银行机构类风险的当期产生"追涨杀跌"的非理性行为,随后转变为"追跌杀涨"的理性行为。
The authors of this paper use the big data of Baidu search, lagged financial risk indicators and lagged structure data variables to set up a financial risk prediction model. The study of the paper shows that, （1） the risk prediction model with the big data of Baidu search helps improve the accuracy of financial risk prediction, and the prediction result of the model in the period of financial risk rising is better than that in the period of financial risk failing; （2） the public search for c risk, banking risk and Internet financial risk leads to a irrational behavior ＂chasing sell＂, and the search for banking risk leads to a irrational behavior ＂chasing sell＂ in the current period, then a rational behavior.
出处
《金融论坛》
CSSCI
北大核心
2018年第1期39-51,共13页
Finance Forum
基金
湖北省教育厅科研计划项目(B2017120)
湖北经济学院校级青年科研基金项目(XJ201507).
关键词
百度搜索
风险感知
金融风险预测
行为金融
Baidu search
risk perception
financial risk prediction
behavioral finance
