Determining the true value of a property is essential to the mortgage loan process. Generally, appraisers perform physical property inspections and match house characteristics against comparable properties sold nearby to calculate an accurate value.
In addition to these physical appraisals, originators often leverage automated valuation models (AVMs) as part of their quality control measures.
AVMs have been used for refinances, loan modifications, portfolio and valuation risk monitoring, home equity lines of credit, marketing purposes, loan pre-qualifications and largely as a tool for appraisal or evaluation reviews. In each of these applications, AVMs have the unique advantage of being completely isolated from all parties involved in the transaction.
An AVM will produce the exact same result regardless of who enters the required input data. Therefore, the use of AVMs in validation provide a time- and cost-efficient value-add to the mortgage workflow, or in some cases as a reasonable source of the value estimate itself.
However, in the run-up to the financial crisis, the traditional appraisal process was often subverted, with home values manipulated to fit parameters set by lenders and little attention was paid to correcting inflated appraisal values based on other objective opinions on value, such as an AVM.
When those subverted loans were later bundled into securities, the inflated values were eventually discovered, but not until the damage was already done. Inaccurate property values put the whole loan ecosystem at risk.
How did home valuations get so off track?
In the years preceding the 2008 market crash, some loan originators veered away from the careful evaluation of information that was critical to the loan process, including borrower assets and income, credit history and validating the true value of a property. In many cases, the industry saw appraisals completed by individuals without appropriate geographic competency, or overinflated appraisals that were influenced by lender or broker bias to meet a certain loan amount.
Ultimately, in either scenario, the supportable value estimate was going unchecked and contributing to the instability of the mortgage market.
The Center for Public Integrity documented appraisal inflation in a 2009 article, where they interviewed a number of appraisers, including Florida appraiser Mike Tipton. “Tipton is among dozens of appraisers who have told the Center for Public Integrity that for years lenders across the United States have pushed them into inflating the value of homes to justify higher mortgages.”
The fallout was only realized after homeowners began to default on their loans in large numbers. By that time, the asset had been sold to investors, who based their purchase on the appraised value, only to find that it was unsubstantiated.
In the aftermath of the crisis, determining true property value became a high priority for mortgage stakeholders at all levels. Post-crisis, multiple regulatory bodies jointly updated the Interagency Appraisal and Evaluation Guidelines in 2010 to address supervisory matters relating to the values and valuation methods underlying financial transactions.
At the same time, Fannie Mae and Freddie Mac (GSEs), initiated the Uniform Mortgage Data Program (UMDP) under which appraisal and loan data would be standardized and the data collected for stronger appraisal quality and risk management capabilities.
In a white paper from 2011, Veros concluded that, “The updated Guidelines continue to recognize AVMs and evaluations as valid tools. In fact, there are several circumstances, such as portfolio analysis, abundance of caution liens, real estate secured business loans, and loan workouts, where an AVM is the preferred way of documenting collateral value.” Further, the efforts being made by the GSEs began to lay the groundwork for a more data-driven mortgage market.
However, lenders were under pressure from federal regulators on every aspect of their loan operations. Federal regulators warned financial institutions to reform their appraisal processes in light of the new 2010 standards and started reviewing appraisals as part of their examinations to determine, as the FDIC put it, “whether the methods, assumptions, and value conclusions are reasonable.”
Unfortunately, lenders were uncertain how to use the interagency guidelines regarding appraisal valuation tools and stay in compliance, so many lenders stopped using AVMs in certain lending channels out of an over-abundance of caution.
Fortunately, the market and technology have evolved significantly since then. Even during the market retraction, AVM developers continued to enhance their models, seeking out new data sources and expanding benchmark testing for the performance and accuracy of their value estimates.
Today, AVM innovation has coalesced into a valuation product category with impressive and measurable accuracy rates. The accuracy of AVMs has now been tested and sustained over multiple years and diverse geography and property conditions.
In addition, AVM methodology is more transparent than ever, with well-documented due diligence literature available on data sources, modeling techniques, testing procedures and safety measures. These resources ensure the soundness and reliability of AVMs.
Utilizing collateral valuation tools is part of the larger effort by the GSEs to expand their use of data and analytics tools to serve the mortgage market. In October 2016, Fannie Mae announced its Day 1 Certainty program, which shields lenders from buyback risk under certain conditions. As part of the program, Fannie Mae provides property inspection waivers on about 20% of limited cash-out refinances, giving lenders freedom from reps and warranties on property value, condition and marketability.
Using big data to make informed property valuation decisions is key to this initiative, and proof that the efforts to standardize and collect appraisal data through the UMDP are paying off.
"Advancements in technology have enabled these capabilities — we couldn’t have done this several years ago,” Andrew Bon Salle, executive vice president of single-family business at Fannie Mae, said at the time of the announcement.
Freddie Mac is set to roll out an appraisal-free mortgage loan to some consumers this spring. “Instead of using professional appraisers, Freddie plans to tap into what it says is a vast trove of data it has assembled on millions of existing houses nationwide, supplement that with additional, unspecified information related to valuation, and use the results in its assessments of applications,” The Chicago Tribune reported in October 2016.
Freddie Mac has utilized AVMs for more than 20 years internally, and offers its AVM tool, Home Value Explorer (HVE), to the industry.
“AVMs have become an integral part of today's mortgage market, and AVM technology has advanced the world of automated valuation services from novelty to necessity. Today, AVMs are efficient, effective and essential tools in loan manufacturing. AVMs help expedite processes, lower costs and minimize risk,” Freddie Mac notes on its website.
As the GSEs utilize data and analytics tools, including AVMs, with more confidence, lenders are sure to incorporate them more frequently as well. This evolution is well-timed in the midst of rising interest rates and an appraisal community already challenged with extensive turn-times and excessive upcharges.
With all these advantages, isn't it time to start bringing these tools back?