A machine learning algorithm applied to macroeconomic factor investing
This paper examines the extent to which macroeconomic indicators can be used to determine the optimal allocation of an extended Fama French 5-Factor model which includes the risk-free rate. The study is based on Modern Portfolio Theory (MPT) as developed by Markowitz(1952) and Smart Beta Investing....
| Autor principal: | |
|---|---|
| Formato: | masterThesis |
| Idioma: | eng |
| Publicado em: |
2022
|
| Assuntos: | |
| Texto completo: | http://hdl.handle.net/10362/144695 |
| País: | Portugal |
| Oai: | oai:run.unl.pt:10362/144695 |
| Resumo: | This paper examines the extent to which macroeconomic indicators can be used to determine the optimal allocation of an extended Fama French 5-Factor model which includes the risk-free rate. The study is based on Modern Portfolio Theory (MPT) as developed by Markowitz(1952) and Smart Beta Investing. The algorithm combines MPT with two Machine Learning (ML) Algorithms (K-means Clustering and Random Forest) to predict the macroeconomic state and arrive at the according optimal ‘tactical’ portfolio allocation of each security over the investment period. The research contributes to the existing literature of ML Algorithm performance applied to Smart Beta macroeconomic strategies. |
|---|