Machine Learning for Transparent and Sustainable Investing

Offentlig prosjektsammendrag

Investment management is performed in one of two forms: either discretionary, that is, selecting investments based on news that are processed by humans, or systematic, by defining rules or algorithms that drive the investment process. A growing example of the latter is passive index-based investing, where the investor follows a popular market index. Another growing example is quantitative investing, where investment strategies based on rules and data are executed by computers. In addition, new concerns in the asset management industry are the effects of climate change on the financial markets and the growing demand by investors for the consideration of Environamental (E), Social (S) and Governance (G) factors when choosing investments. Machine Learning for ESG-investing aims at improving the performance of quantitative investment strategies applied to the stock market, created with the help of machine learning techniques, instead of traditional, outdated statistical tools, and taking into consideration the role of ESG factors. An AI/ML-based model for quantitative investing will lower management fees for the investor thanks to automation. It will make markets more efficiently priced through the more intensive and rational use of information. And it will increase the transparency of the investment process, since the results of the model can be interpreted in terms of traditional factors and the investors can gain insight into how ESG factors affect markets. The results from this research project can be applied to construct cheaper, more efficient investment portfolios that align with well defined ESG-based investment goals.