Portfolio Optimization with Financial and Social-Environmental Objectives: A Methodological Proposal for Managing Impact Investment Portfolios
investiment decisions, impact investing, sustainable finance, socio-environmental analysis, allocative efficiency, multi-objective optimization.
The main result of this study was the proposal and test of a new methodology for the management of investments made with the purpose of generating financial
returns and socio-environmental impact, bringing as a contribution the improvement of allocative efficiency for the management of this asset class by incorporating the benefits of diversification in its analytical tool, a practice already known by traditional portfolio managers, as well as the reduction of risks due to the learning effect. Thus, the objective of the present work was the proposal for a new methodology for impact investment portfolios selection. This methodology investigates ways to use available scientific evidence and it explores the effects resulting from the interdependencies of the combined results generated by each of the investment alternatives. Due to the scarcity of similar studies in the literature and the rare use of rigorous methods for implementing impact investments, this study, therefore, fills a gap to develop the theme of sustainable finance, whose relevance has been gaining importance over the last decades, as well as it illustrates the application of the methodology in order to positively impact management decisions in this area. The methodology developed was based on a multiobjective and multiperiod optimization model to measure the efficient frontier and select the portfolio according to the decision maker’s preferences. In addition, this study investigated a methodology, relying on the use of Bayesian meta-analysis algorithms, for the treatment and consolidation of the social and environmental impact estimates of investments, that were previously obtained by experimental and quasi-experimental methods or assessed by an expert’s opinion. The proposed model was applied in the selection of impact portfolios at the BNDES, using multi-objective genetic algorithms, concluding that, according to the objectives pursued, the institution's effective allocations in the sample period are Pareto efficient. However, the study identified allocative changes that have the potential to maximize the impact on job creation and increase company revenues, dimensions in which robust empirical evidence resulting from BNDES financial support was found.