Big Data’s Role in Appraisal Modernization

This piece originally appeared in the August 2023 edition of MortgagePoint magazine.

Matt Stepanovich serves as VP of Appraisal Modernization and Quality Control for SingleSource Property Solutions, where he oversees the company’s appraisal modernization initiatives. As a certified Residential Real Estate Appraiser with more than two decades of experience in the valuation and financial services industries, he has owned a successful real estate appraisal firm for over 15 years and has extensive knowledge and expertise in hybrid and alternative valuation products. He has also served as SVP of Sales and Operations for a regional real estate company, and VP of Sales for a national real estate inspection and hybrid valuation firm. He can be reached at [email protected].

Q: When did Fannie Mae and Freddie Mac begin collecting data for property valuations?
The GSEs implemented the Uniform Appraisal Dataset and began collecting various appraisal data points back in 2010. At that time, appraisal data was somewhat fragmented and inconsistent. In fact, that was part of the goal, to standardize appraisal data points across the board. The initiative was instrumental as it marked the inception of a more data-driven, systematic approach to property valuations.

Q: What was the benefit of implementing UAD?
The most significant advantage was that it gave both GSEs the ability to start systematically tracking data on individual properties. All of the data was housed in one centralized hub.

For the past five years, Fannie and Freddie have been able to monitor valuations and other data on specific properties. With that historical data in hand, requirements for the type of valuation that would be needed on specific transactions can be more easily determined.

The UAD also made the access, analysis, and utilization of property data significantly more efficient and reliable. This ultimately led to more precise, data-driven approaches to property valuations that reduced margins of error and provided a more nuanced understanding of the property’s value.

Q: How is historical data utilized?
The use of historical data in the property valuation process has always been a critical part of determining current market value. If you think of each property as having its own unique story to tell, the historical data serves as the chapters of that story. When an appraiser is tasked with determining a property’s value, the historical record of that property is crucial information for the appraiser. As they sift to analyze the data, they could be looking for clues in the form of data points that could be relevant to their current assessment. That might include details about past renovations, changes to the property’s layout, value fluctuations, or other data. Collectively, each of these elements contributes to a broader understanding of the property’s current value.

The same historical data can also help shed light on what the valuation requirements for the current report should be. For example, if the data points have a comprehensive and consistent history and the type of transaction qualifies, perhaps an inspection isn’t needed. Conversely, if there are data gaps or inconsistencies, perhaps an inspection or full appraisal is needed. Basically, historical data can be both a guide and a foundation that ultimately enables a more accurate and efficient determination of value.

Q: Are there any other benefits to using data?
Absolutely. For instance, some of the appraisal bias issues that have been raised are being investigated through UAD data. The Federal Housing Finance Agency recently examined over 47 million appraisals from 2013 to 2021 to investigate potential appraisal bias.

The agency developed a metric that identifies when properties appraise for less than a buyer’s offer and found undervaluation to be more prevalent in certain metro areas in which expected appreciation, property condition, housing availability, and buyer demand were factors. In such cases, the appraiser will have no idea about the ethnicity of the homeowner.

UAD data is also a valuable resource for other academic and industry researchers studying housing markets, lending practices, and appraisal practices and methodologies.

It’s used to understand real estate market trends and dynamics on a micro and macro level, which can be helpful for making strategy planning, policymaking, and investment decisions. And financial institutions and third-party vendors can leverage UAD data to develop proprietary models for property values and loan pricing.

Q: What type of data is collected by property inspectors for valuations?
Inspectors are able to collect a wide range of information that is important to the appraiser’s valuation. For example, they can provide a floor plan to help identify functional concerns related to the interior layout of the dwelling. The floor plan breaks down each level of the structure while indicating the interior walls, staircases, and flow of rooms. The digital floor plan is also used to help understand the structure’s gross living area. The inspectors also report the quality and condition of the dwelling’s interior and exterior for updates/damages, identify any marketability or location issues, and provide photos of all rooms from multiple angles.

Inspectors can capture virtually everything about the current condition of the property, including details about finishes, the type of flooring, the condition of the walls and windows, and any unique features. All of this data can then be relayed back to the appraiser to form an accurate and comprehensive valuation of the property, often in a much faster timeframe than a full-scale appraisal.

Q: How do you think the appraisal process might evolve over time?
As technology improves, one likely possibility is that artificial intelligence (AI) will be utilized to analyze data stored from the subject property as well as local neighborhood data and specific street metrics in order to come up with valuations. As we accumulate more and more data about properties, neighborhoods, and markets, AI also has the potential to play a transformative role in how we analyze and forecast housing market trends.

However, it’s important to note that even the most powerful technologies are still only tools. Human expertise will always be critical in the appraisal process to ensure every valuation makes sense and meets the needs of all parties involved.