Online margin trading strategy based on LSTM prediction information
With the growing maturity of the domestic margin market and the continuous ad-vancement of financial technology,the development of intelligent margin trading strategies has emerged as a critical topic and challenge in the field of quantitative finance.This article employs long short-term memory(LSTM)networks to construct expert strategies based on predictive information and proposes an online margin portfolio trading strategy that integrates expert opin-ions.Firstly,multiple technical indicators are utilized as input variables to anticipate the trend of stock prices using the LSTM neural network model.Next,an expert specializing in investing in a single stock is considered,and buying or selling strategies for each expert are formulated based on the LSTM model's prediction results.Then,a weight optimization model based on expert performance is proposed to determine the weight of each expert by solving the model.Finally,to demonstrate the efficacy of the proposed strategy,historical trading data from stock markets is utilized for empirical analysis.The results demonstrate that the developed strategy is capable of achieving better performance than some benchmark strategies,even when taking transaction costs into consideration.
NETL
NSTL
万方数据
