Multifamily Investment Firm Develops Data-Driven Screening Approach
CONTI Capital says better data and machine learning leads to better decisions.
Real estate investment company CONTI Capital says that better transaction and psychographic data combined with machine learning tools has led it to greater efficiency with a data-driven screening approach that it calls the CONTI Index. In a white paper, the company calls it a “site selection model designed to identify outperforming metropolitan areas, submarkets and zip codes for multifamily investment opportunities.”
Prompting the company was a view that the Covid-19 pandemic had created “idiosyncrasies” and not “structural shifts” in housing markets. “We required a tool to see through the current market dislocations, to identify which market trends were permanent and which were transitory,” they wrote.
Four factors led to the ability to create the data-driven approach they now use. Listing platforms “improved the quality, granularity, and timeliness of market and submarket rental rate data.” Third parties were making real-time demographic data tied to “individuals that live within specific geographies” rather than waiting for more abstracted information from the Census Bureau. Psychographic data and methods, “typically used in marketing campaigns to categorize individuals into segments based on shared characteristics,” allowed better understanding of consumer behavior. Finally, a combination of better-quality data with a shift from econometric to machine learning models allowed analysis without pre-existing explicit assumptions about the data.
The CONTI Index uses two models: one that works from macroeconomic impact on property markets and a second that is operates at a market-level. The company then takes the top 50 U.S. apartment markets by total unit count and looks at each by supply and affordability; number of prime renters; local labor markets; industry metrics like rent growth, occupancy rates, NOI growth, and IRRs; quality of life; and market fiscal health.
After identifying MSA-level target markets, a ZIP code-level analysis finds the “pockets within each market that are idea for multifamily investments.”
The point of the tool is not to make all the decisions in considering specific properties, but to be a first-step screening that “allows us to screen an immense number of opportunities, ensuring that we never pass on a deal due to our biases.” It also speeds analysis from “an hour or more, down to just a few seconds” to consider whether a specific opportunity should move to the next step.
That said, such an approach has to be tailored for any given investment philosophy. So, CONTI’s choice of top 10 markets—Dallas-Fort Worth, Atlanta, Austin, Charlotte, Orlando, Tampa, Houston, Nashville, Raleigh-Durham, and Phoenix—might make sense for them and not for another investment approach, perhaps one that looks for under-appreciated markets that might not draw as much competition, resulting in lower acquisition costs.