Is home-buying getting less personal? Or just more scientific?
Real estate companies, lenders, insurers, property owners and those who provide data for all of these stakeholders are increasingly using artificial intelligence in every aspect of buying, selling and home financing. Algorithms churn through millions of documents in seconds: property values, debt levels, home renovations and even portions of a homeowner’s personal information. AI helped in many markets during the COVID-19 pandemic’s worst days and since, when supply for sale dried up and prices surged, because it meant agents could cast their net wider.
Speed and accuracy around buying and selling increasingly matter as the US real estate market remains dominated by higher prices – though trends always vary by region – and a rush by many would-be buyers to get in before additional projected interest-rate increases materialize.
One data firm, CAPE Analytics, which produces artificial intelligence-powered geospatial property data, believes it can further remove human emotion and error in valuing property, in particular by providing a wider-lens snapshot of exterior and surrounding conditions that will impact a home or The value of commercial building.
CAPE’s AI-powered platform, AIRE, launched this week, will fill critical gaps found in traditional data sources, it argues. For instance, AIRE will include distance to highways and bodies of water. It can provide late buying-process data in the event the property has signficantly changed due to wildfire or other severe storms. AIRE even captures the view visible from each property.
“All of which are features that are missed by existing data sets and even human-led exterior inspections,” said Raj Dosaj, head of real estate for CAPE Analytics.
“As desktop and hybrid appraisals increasingly become the norm, the most accurate and complete data sources must be brought to bear in order to maintain accurate valuation assessments,” he said.
This expanded data collection has become increasingly important to real estate, home lending and property insurance markets as climate change-induced storm severity, drought conditions, flooding and wildfires change communities. Over 600 fires had broken out in New Mexico and Arizona by early May. And that follows extreme wildfire conditions in Colorado, California, Oregon and elsewhere in recent years.
Read: Climate change fueled 3rd costliest losses ever in 2021 – less than half of that property was insured
CAPE’s AIRE platform combines computer vision, aerial imagery, granular geospatial data and proprietary data sources.
CAPE says it found inconsistent reporting from the traditionally-used automated valuation models (AVMs), standard exterior broker price opinions (BPOs) and property condition reports (PCRs). All told, 70% of property issues identified by AIRE are missed by existing human-driven solutions, the company believes.
“‘Many valuation and appraisal tools… have trouble assessing atypical homes – or even typical homes in atypical locations – which can be exacerbated when real estate markets become more volatile.’ ”
This leads to risk miscalculations during underwriting, loan and real estate trades, volatile insurance premium-setting, and misprices real estate investment portfolios. And, ultimately, it means that home buyers and sellers aren’t getting the full picture.
Read: Is now a good time to buy a home? Most Americans don’t think so – and Gallup says it’s the first time more than 50% of people feel that way
Technology is of course changing the way house hunters look for homes, with the advent of more web-based listing and valuation estimate sites, including Zillow Z,
Realtor.com, Opendoor, HomeLight, UpNest and others. Some sites, such as Zillow, even acted as an investor by scooping up many of its own listings for flips, before having to sell at a loss. The offers iBuyer strategy, however, did not hurt early-year earnings, as MarketWatch reported at the time.
But CAPE’s Dosaj argues that the property valuation snags for Zillow and others are painful symptoms of data lags, especially when the pandemic and an accelerated work-from-home trend forced highly unusual market conditions, including very low interest rates, difficulty touring homes, and a spike in demand from many buyers to leave congested areas.
“Many valuation and appraisal tools… have trouble assessing atypical homes – or even typical homes in atypical locations – which can be exacerbated when real estate markets become more volatile,” Dosaj said.
“These same issues extend to appraisers and even consumers looking to buy or sell homes as it is challenging for humans to subjectively determine the impact of various positive and negative factors on a home’s value,” he said.