We wanted to adjust the City Challenge and instead of Fixed Income look at Commercial Real Estate assets. We built a tool to help investment professionals in REITs and traditional commercial real estate firms make better investment decisions using machine learning.
Right now, prospecting for new properties to acquire into a REIT portfolio or a traditional open- or closed-ended fund is a very manual and inefficient process, and a team of analysts spends about 80% of their time on it. Unlike stock and bonds, the commercial real estate asset class has not adopted a data-driven investment approach.
The acquisition process takes about 18 weeks per acquisition and leads to underperforming the firms acquisition targets and leaving millions of dollars on the table due to the uninvested capital. There is currently $300B of uninvested CRE investment capital which equates to median 25% of pledged under-deployed funds in the US.
We prototyped a platform which uses machine learning to analyze and model growth in neighborhoods and build prediction models to help these professionals better understand asset growth to decrease the amount of manual and tedious work they need to do before they go in and conduct a deeper dive analysis on selected assets.