New report from the Oxford Future of Real Estate Initiative discusses how AI could make buying a house more efficient.
Imagine a world in which house sellers, buyers and mortgage lenders have access to a public list of all properties on the housing market, alongside independent and public valuations for each listing. The process of buying a house would be faster – as in-person appraisals would be unnecessary- and liquidity would significantly improve.
With this ideal in mind, the Oxford Future of Real Estate Initiative (FORE) has investigated the benefits and limitations of using Automated Valuation Models (AVMs): little-used software-based pricing models for property valuation, which are cheaper, more efficient, and more consistent than human appraisals.
The team interviewed industry and government practitioners, discovering that reliability and transparency seem to be major barriers to mass rollout and adoption of AVMs. Currently, most AVMs are developed by privately-owned companies offering valuation services, who, understandably, want to protect their trade secrets. This secrecy around how individual AVMs are run – referred to by the authors as the ‘black box’ - is vital to developers’ competitiveness.