互联网金融收益率的统计预测研究——以余额宝为例

Statistical Prediction Research on Internet Financial Return Rate——TakingYu’ EBao as an example

以余额宝每万份收益数据为研究样本,构建ARIMA模型进行研究,判断模型的拟合效果,对余额宝每万份收益的发展趋势进行预测.研究结果表明,余额宝每万份收益具有一阶单整的性质,当期余额宝每万份收益的变化受到上一期余额宝每万份收益的影响.此外,模型预测的准确度受到国家政策、市场利率、规则调整等不确定因素影响.可以根据ARIMA模型的预测大致判断出余额宝每万份收益的未来走势,为互联网金融市场众多参与方提供决策参考.

This paper takes per ten thousand copies of return on Yu’ EBao as the research sample,constructs ARIMA model to study,judges the fitting effect of the model,and forecasts the development trend of per ten thousand copies of return on Yu’EBao.The research results show that per ten thousand copies of return on Yu’EBao have the first-order monolithic nature,and the change of per ten thousand copies of return on Yu,EBao in the current period is affected by the last earnings of Yu’ EBao.In addition,the accuracy of model prediction is affected by uncertainties such as national policy,market interest rate,regulation adjustment and so on.According to the forecast of ARIMA model,we can roughly judge the future trend of Yu’ EBao,and provide decision-making reference for many participants in the Internet financial market.