Time-varying neural network for stock return prediction
This work addresses the challenge of adapting neural networks to time-varying financial market data, which is incremental as it builds on existing methods to handle dynamic environments.
The authors tackled the problem of neural network training in time-varying contexts, specifically for stock return prediction, and demonstrated that their online early stopping algorithm outperforms current approaches in tracking functions with unknown dynamics, showing concrete superiority in predicting monthly U.S. stock returns.
We consider the problem of neural network training in a time-varying context. Machine learning algorithms have excelled in problems that do not change over time. However, problems encountered in financial markets are often time-varying. We propose the online early stopping algorithm and show that a neural network trained using this algorithm can track a function changing with unknown dynamics. We compare the proposed algorithm to current approaches on predicting monthly U.S. stock returns and show its superiority. We also show that prominent factors (such as the size and momentum effects) and industry indicators, exhibit time varying stock return predictiveness. We find that during market distress, industry indicators experience an increase in importance at the expense of firm level features. This indicates that industries play a role in explaining stock returns during periods of heightened risk.