'The Optimal Mix' : deploying portfolio theory on real estate asset returns in mixed-use development

Abstract

Mixed-use has emerged as one of the most popular and demanded forms of real estate development in many metropolitan regions around the world. While mixed-use development broadly incorporates a variety of functions including, residential, commercial, and retail programs within one project, there is little science in determining the 'optimal mix' in mixed-use development resulting in a programmatic melange. Current practices largely determine the program mix through "gut intuition" or "rule of thumb", and value mixed-use projects by the returns of the individual components. This study seeks to develop an alternative model in defining an ideal program mix in mixed-use development that is based on an optimized and quantifiable portfolio value. The goal is to develop a framework for determining a recipe for mixed-use development in the hope of guiding future development practices in building more efficient, profitable and sustainable mixed-use developments across the United States. This study sees an opportunity to apply Modern Portfolio Theory, a widely adopted method in the finance industry that determines the most efficient allocation in a portfolio of assets, to identify an optimal program mix in mixed-use development projects. Mixed-use developments are inherently a portfolio of distinct real estate assets. Each component product type, such as residential, office, and retail can be thought of as individual assets within a mixed-use portfolio. These component assets offer varying returns and volatilities due to their individual characteristics and correlations with the market. If a mixed-use project is viewed as a portfolio, then an opportunity exists to optimize the project by adjusting allocations in the individual assets, resulting in an efficiently programmed project that maximizes total project returns for a given level of risk. Using market data, this thesis intends to identify the 'optimal mix' for fourteen markets across the United States. The study seeks to discuss the real-world limitations of implementing these program mixes in order to propose a new method to quantify and evaluate programming in mixed-use development; a method based on determining an 'optimal mix' that will generate the highest risk-adjusted returns for an investor, bringing to the forefront a new method in intelligent programming.