By Frank Rohde • June 27, 2024
Leveraging Artificial Intelligence to Enhance Returns
We leverage artificial intelligence to create a differentiated real estate investment portfolio with great customers and assets.
Living in San Francisco - as I do - one can't order a sandwich without running into a conversation about AI. So I wanted to explain where and how we use AI as part of the Ownify platform.
Underwriting Our Customers
Ownify has built a powerful underwriting engine that leverages a set of models to a) prevent fraudulent customers from applying in the first place, b) assess their credit worthiness, and c) predict their available cash flow to support the monthly Ownify payment.
As part of that, our platform ingests several hundred variables for each applicant and ultimately renders a decision that results in a "home budget" for each applicant. In order to train our models and set up constraints, we bought a 1M+ record set of historical performance data on low-down payment mortgages, solving for the "cold start" problem in machine learning.
Our results thus far have been very encouraging, with zero late pays across our portfolio.
Forecasting Home Valuations
Once we've qualified an "Owni," we evaluate the appreciation potential for each home that the Owni submit to Ownify for potential purchase. To render a prediction here, we've built two sets of models:
A) A macro-level model that scores the attractiveness of each of the 300+ metro markets in the US based on economic fundamentals such as housing supply / demand, job creation, household formation, regulatory friendliness, and other housing economics. This model helped us identify the Triangle region in NC as our launch market.
B) A property level model that uses several hundred variables to assess the current value and appreciation potential for each individual home.
As much as we trust data, we also conduct an in-person inspection of each home before we close. Our results have been encouraging, with an average 6% appreciation across the portfolio in 2023.
Driving Internal Efficiency
How We Mitigate Risk
Of course, we recognize the potential risks of AI and the importance of mitigating bias in our models. We are proactively managing these risks by building inclusion directly into our technology; for example, our underwriting engine is able to underwrite small business owners, gig economy workers, and other non-W2 employees (where most traditional underwriting engines fail). We also keep a 'human-in-the-loop' when making key business decisions.
At the end of the day, homeownership is a deeply tangible, physical, and emotional experience. However, when applied correctly, AI / ML can help us create better returns, higher efficiencies, and cost savings for our Ownis and investors.