LSTM-Based Research on Stock Price Prediction in the Real Estate Market

After years of development in Chinese financial market, the stock market plays an increasingly vital role in resource allocation, among which listed real estate companies have made great strides in this market. Therefore, this paper aims to adopt a scientific method to analyze the stock price changes and give appropriate investment strategies to promote the stable growth of the capital market and increase the returns of investors. This paper is based on the efficient market hypothesis and uses daily time series data from January 1, 2020 to October 1, 2023. First, the daily rise and fall data of the real estate sector index(Wind All A Real Estate Enhanced Index） and five representative companies are selected from Wind and Tushare. Then, python software is used to test the stationarity of selected time series data in the past four years. While VAR model is constructed to validate the influence of selected index and the sliding window rolling technology is used to establish the LSTM model data to roughly forecast the future trend of the real estate stock. Therefore, the results offer a rough prediction of the future market in the real estate industry through the combination of macro and micro perspectives, hoping to provide a reference for investors to make decisions on the real estate stock market in the future.