Koedds: A National Real Estate Investment Analysis
Sean Kouma, William Edds
TL;DR
This work addresses identifying counties with high real estate growth potential while managing risk by merging historical growth data with the $SVI$ at the county level. It constructs a normalized investment score through a weighted combination that includes $1 - SVI_{county}$, using Redfin metrics dating back to 2012 and presenting results via a Monocle GeoJSON visualization. Although the experimental bench marking reveals only a weak link between $SVI$ and growth ($r = 0.126$), the approach highlights Midwest regions (notably around Iowa and Ohio) as promising, and demonstrates how a data-driven, geospatial scoring framework can inform both investor decisions and consumer guidance. The paper also contemplates a future shift toward compound ML-driven models and industry adoption to enhance predictive power and decision support in real estate markets.
Abstract
With costs and risks increasing for investors and home buyers alike, additional analysis of the housing market is required to help individuals make the right choice. In addition to traditional market analysis, other aspects such as the economic vulnerabilities of the local community must be taken into account to further ensure real estate buyers receive a positive return on investment from their purchases as well as ensuring that traditional home owners get the best price for their future home.