Predictors of Sustainable Investment Motivation: An Interpretable Machine Learning Approach

Sosnovskikh, Sergey ORCID: https://orcid.org/0000-0002-3744-740X, Valko, Danila and Meyer-Alten, Raphael (2025) Predictors of Sustainable Investment Motivation: An Interpretable Machine Learning Approach. Sustainable Development. ISSN 0968-0802

Published Version Available under License Creative Commons Attribution. Download (1MB)

Abstract

This study investigates the determinants influencing retail investors’ capital allocation to sustainable financial products, focusing specifically on Germany—a pertinent case due to its strong commitment to sustainability, supportive regulatory environment, and substantial market growth. We utilised two surveys conducted in 2020 by the 2° Investing Initiative and Choyze GmbH in collaboration with the German Environment Agency to obtain measurement invariant combined dataset. Utilising robust and generalised linear modelling and interpretable machine learning technics, our analysis identified that three primary motivation components – personal values, social and environmental impact, and investment return – exhibit significant overlap. The findings demonstrate that investment motivations are consistently predicted by sustainability interests. In contrast, socio-demographic factors (age, gender, education, household income) exhibit inconsistent pattern across investment motivations and exert weaker influence. As a result, we propose robust models to predict the sustainable investment motivation and highlight their consistency and applicability in the green finance sector.