Applying the Fintech revolution to appraisal.

Author:Wall, Kevin
Position::THE MORTGAGE IMPERATIVE
 
FREE EXCERPT

WHEN FIRST AMERICAN APPROACHES A LENDER about our collateral valuation services, one of the first questions we're asked is, "How do you select your appraisers?" Appraiser selection may be the single most significant factor in determining appraisal quality--yet this critical element of collateral valuation continues to operate much as it has for decades. Current methodologies for selecting appraisers are only marginally effective at assessing appraiser quality and vary widely from one company to the next. The time has come for a revolutionary approach to collateral valuation led by big data and analytics.

The financial services sector is often accused of being slow to innovate. The last decade, however, has seen a gradual but steady march toward modernization. Bleeding-edge developments in financial technology (fintech) that first took root in consumer finance and Wall Street trading are now beginning to transform the mortgage industry. There's no better place to start than with the underlying security that supports the entire mortgage lending process: collateral valuation.

The last seven years have brought several noteworthy developments in residential appraisal regulation. Both 2009's Home Valuation Code of Conduct (HVCC) and the 2010 Dodd-Frank Wall Street Reform and Consumer Protection Act that supplanted it delivered much-needed clarity around how lenders should and should not interact with appraisers to ensure appraisal integrity and independence. Big data first arrived on the appraisal scene when standardization of real estate appraisal data was significantly improved following the 2011 implementation of the Uniform Appraisal Dataset (UAD), part of Fannie Mae and Freddie Mac's larger Uniform Mortgage Data Program[R] (UMDP). Fannie Mae's Collateral Underwriter[R] is already leveraging analytics supported by this big data to deliver automated risk management of appraisal reports.

Appraiser selection, on the other hand, remains largely unaddressed in the appraisal process. While it's clear that lenders bear responsibility for understanding and managing appraiser selection--even when appraisal fulfillment is contracted through an appraisal management company (AMC)--there is no industry consensus as to how lenders should actually navigate this task.

The requirements around regulatory compliance apply regardless of the appraiser engagement model that best fits the lender valuation strategy. Many lenders elect to engage appraisers directly. Others find value in outsourcing the entire process to AMCs. A growing segment combines a hybrid of both direct appraiser engagement backed by outsourcing to AMCs.

Appraiser selection is the first step in the appraisal process, and it plays a critical role in determining appraisal quality. Choosing the right appraiser from the get-go can greatly reduce lender exposure to substandard appraisals and the downstream risks that go with them.

For instance, a questionable appraisal may trigger multiple reviews as it is passed from the appraiser, appraisal firm or AMC to the lender and on to the investors. At best, this drawn-out appraisal review process is a drag on productivity that eats into lender margins, and in certain cases on noncompliant appraisals, lenders may eat the cost of reappraisal. In the worst case, appraisal hang-ups can impede closing--costing lenders and borrowers real money--or even lead investors to request a buyback of the loan.

The question is, how do lenders select top-quality appraisers--or at least ensure their AMCs do? There is currently no standardized method for identifying top-quality appraisers or validating appraiser selection during appraisal review. Interagency guidelines dictate only that institutions should "establish selection criteria and procedures to evaluate and monitor the ongoing performance of appraisers."

As a result, lenders and AMCs across the industry each use their own internal metrics and data points to rank order or score appraisers. Typically an appraiser ranking is determined using a combination of characteristics, such as turnaround time, daily volume, certification/licensing status, customer service, revision rate, fee and proximity to the subject property.

The current approach is problematic for several reasons. First, the architecture of today's appraiser-selection process places too much emphasis on turn time and cost and not enough on appraisal quality. Just because an appraiser is fast and charges a lower fee than other appraisers, it doesn't follow that his or her appraisals are of a superior quality. Second, current methods offer no visibility into the quality of appraisers not currently on a panel.

The time has come for our industry to apply so-called big data and sophisticated analytics to solve these problems and potentially revolutionize collateral valuation.

Fannie Mae's May 2016 Mortgage Lender Sentiment Survey observed that more than half of lenders surveyed think the industry would benefit from a "digital disrupter"--the equivalent of an Uber[R] or Google[TM] that will redefine the way our entire industry does business.

What would digital disruption look like in the appraiser selection space? Today's top companies make it easy for customers to see what they're buying before they buy it. Much as Airbnb[R] users read guest reviews to inform their choice of lodgings, those selecting the appraiser should be able to identify the best appraiser for the job--and big data is the key to making it happen.

By pooling data from multiple sources--not just what's available to the engaging party, but also data from a source used by all appraisers, such as appraisal software providers--we can develop a more precise and objective process for identifying the best appraiser at the time of order assignment based on the historical quality of each appraiser's work. If the "best" appraisal is one that uses well-selected comparable properties to generate highly accurate collateral values, regardless of turnaround time or fees, then the best appraisers are those who consistently deliver accurate appraisals. Thus, if big data and analytics allow us to meaningfully and objectively score appraisals for quality, then by extension, appraisers can be assessed, too.

Revolutionizing appraiser selection will require both vision and resources. Few companies own the data assets required to build a more robust model for appraiser selection. In fact, most AMCs and lenders have access only to their own appraisal data. The data analysis expertise necessary to build and interpret meaningful appraiser selection models is similarly hard to come by.

Yet, if we can overcome these challenges, so that appraisal quality metrics are not only aggregated, but standardized and accessible, our industry could improve not just appraiser selection, but the entire appraiser recruitment process. The right data analysis could provide compliance with engagement of the best appraiser who also selects the best comparables and produces supported and defensible values. This is much the same way automated appraisal review tools help lenders, investors and AMCs validate appraiser work today.

The more widely and freely our industry's combined data is shared with everyone involved in the appraisal process, the more applications we'll find for using it to our mutual benefit. Digital disrupters like Uber and Lyft have fundamentally changed not just how we catch a ride, but also how we communicate with the driver, how we pay for the service and how we evaluate our experience. Big data and analytics have the power to improve the appraiser selection process resulting in reduced costly reviews, buy backs, delayed closings and regulatory risk.

Kevin Wall is president of Santa Ana, California-based First American Mortgage Solutions, a primary provider of collaborative and customizable data, analytic, valuation, title, settlement, risk-mitigation and quality-control solutions for loan origination and servicing. He can be reached at kwall@firstam.com.

To continue reading

FREE SIGN UP