All projects
Forecasting U.S. Equity Returns
Compare how accurately regression and machine learning models forecast the next month’s U.S. equity market return.
- Backtesting
- Machine Learning
- Time Series
In collaboration with AmStat, we will use approximately 15 publicly available market and macroeconomic predictors to forecast the next month’s U.S. equity market return. We will compare regression models with two machine-learning models on held-out periods, assess forecast accuracy across market conditions, and evaluate the performance of a forecast-based trading rule net of transaction costs.
Related project with AmStat: Out-of-Sample Signal Performance