Optimizing a Portfolio of Impact Investments: A Focus on Practicality and Financial Performance

Abstract

In this thesis, two key challenges that are currently hindering the growth of the impact investing field were addressed: (i) the lack of a standardized way to classify investments as ‘impact investments’, and (ii) the “lack of appropriate capital across the risk/return spectrum” and “shortage of high quality investment opportunities”81. Data Filtering and ‘Impact Investments’ Classification To address the first challenge, the following was accomplished:  A method was devised for investors to quantitatively evaluate whether an investment should be considered an ‘impact investment’, based on the two key characteristics that are commonly used to define ‘impact investments’: ‘financial performance’ and ‘impact’  The method is applicable to U.S. equity and works by assigning standardized, easily obtainable, and reliable proxies for these two characteristics: o Financial performance – annualized geometric mean monthly returns as a measure for return, and annualized standard deviation (or least absolute deviation) of monthly returns as a measure for risk. Both these measures form the proxy for financial performance o Impact – an overall ‘impact’ score for each company/investment, calculated using a weighted sum of environmental, social, and governance indicators from MSCI’s KLD database. The weights are determined by the investor, based on their ‘impact’ preferences (i.e. what kind of ‘impact’ they value the most: social, environmental, or corporate governance). This ‘impact’ score is the proxy for ‘impact’  Investments with an insufficient level of ‘impact’ can then be filtered out, leaving the investor with a set of ‘impact investments’. The ‘impact filter level’ (i.e. stringency of the impact filter) can either be set directly by the user, or is determined by a built-in statistical method based on the ‘impact’ data  To ensure maximum ease-of-use and practicality for the user, which is a key goal of this thesis, the method was automated into a tool made up of code files on R, whereby the user only have to input four lines of inputs to generate a set of ‘impact investments’ and the investments’ corresponding ‘financial performance’ and ‘impact’ data This addresses the first challenge hindering the growth of the impact investing field, not only in terms of theory, by offering a standardized process for investors to classify investments as ‘impact investments’ in the U.S. equity universe, but also in actual practice, by offering a flexible tool which investors can use in practice to actually filter and classify investments as ‘impact investments’ by simply inputting their preferences.