Stock Price Prediction Based on Conditional Flows Scenario Generation

Stock price forecasting is an important issue in the financial field. However, most of the existing studies were focused on the prediction of a single stock, ignoring the correlation among different assets. A possible way to solve the above problem is to provide a set of scenarios which include the future returns of several stocks, instead of a single one. The flow-based model is a kind of deep learning model proposed in recent years, which has powerful data generation abilities. In this paper, we use a flow-based conditional generative model to forecast the stock price scenario. We use real stock market data to verify the proposed method. The simulation results show that the model based on the proposed method can capture the complex dependence of the future stock relationship and provide more accurate and diversified forecasting results.