Research On The Influencing Factors And Forecasting Methods Of The Rate Of Return Of Internet Insurance Current Financial Products

With the steady increase in the per capita disposable income of Chinese residents and the gradual opening up of financial management concepts,the public's reliance on the Internet has been continuously improved,and Internet finance has gradually become a new investment method.Internet insurance financing is a new mode of Internet financial management and insurance financing.This article takes Internet insurance current financial products as the research object,and analyzes its characteristics of “stable capital,wide investment range,high liquidity and low threshold”,and combs its products.Characteristics and development status.Then this paper mainly studies from the following aspects:First,the set of empirical modal decomposition(EEMD)model is used to decompose the selected Internet insurance current financial products yield series into a set of eigenmode functions and trend items with different frequencies,which have different correlations with the original sequence.Among them,the trend item has the highest degree of correlation with the original sequence,and the variance impact is also the largest.Second,the EEMD-QR model is constructed to study the influencing factors of Internet insurance current financial products.According to each component feature and t-test,the highfrequency/low-frequency components and trend components are recombined,and the degree of correlation with the original sequence and its characteristics are used to investigate the influencing factors of the large impact on its components.It can be found that the product yield is mainly affected by the trend item and the low frequency component,and is qualitatively analyzed and further quantified in combination with market changes during the study period.Through the quantile regression(QR)model,an empirical study on the factors affecting the profitability of Internet insurance current financial products.Studies have shown that interbank lending rates,bond market conditions,insurance institution strength,exchange rate and securities secondary markets have varying degrees of impact on yields at different quantile levels.Thirdly,according to the non-linear and non-stationary characteristics of the profitability of Internet insurance current financial products,the EEMD-LSTM nonlinear combination model is constructed to predict the profitability of Internet insurance current financial products.This paper predicts and integrates the EEMD's decomposition of the original product's rate of return.At the same time,the high-frequency component of the profit rate of the Internet insurance current financial products is mainly caused by the market random fluctuation,so the EEMD-LSTM model that eliminates the high-frequency component is constructed.The EEMD-LSTM model with all component integrations and the single LSTM model are separately predicted.The empirical results show that the EEMD-LSTM prediction effect of eliminating high frequency components is significantly higher than the other two models,which also explains that the profit rate of Internet insurance current financial products is mainly driven by low frequency components and trend items.Through the research of this paper,we will enrich and improve the theoretical methods of Internet wealth management products,especially in the insurance sector,and provide ideas and reference for studying other Internet wealth management products.At the same time,correctly understanding the influencing factors of the revenue of such products and making predictions can have certain guiding significance for the Internet financial platform,the corresponding financial institutions and the regulatory agencies,and can better help investors rationally choose appropriate products for investment.