Confident Risk Premiums and Investments using Machine Learning Uncertainties

Abstract

This paper derives ex-ante (co)variances of stock-level and portfolio-level risk premium predictions from neural networks (NNs). Based on the precision of risk premium forecasts, I provide improved investment strategies. The confident high-low strategies that take long-short positions exclusively on stocks with precise risk premium forecasts deliver superior out-of-sample returns and Sharpe ratios than traditional high-low strategies. Mean-variance strategies that incorporate covariances of return predictions also outperform existing strategies. Risk premium variances reflect time-varying market uncertainty and spike after financial shocks. Cross-sectionally, the level and precision of risk premiums are correlated, thus NN-based investments deliver more gains in the long positions.

Keywords: Stock Return Predictions, Neural Networks, Standard Errors, Risk Premiums, Confidence Intervals, Investment Strategies, Machine Learning Uncertainties

JEL Classification: G11, C13, C57, C58