Statistically-Supported Appraisals Can Weather A Perfect Storm

Statistically-Supported Appraisals Can Weather A Perfect Storm REQUIRED READING: The appraisal industry is confronting a level of scrutiny, competition and regulation that some have characterized as the ‘perfect storm.’ How appraisers navigate this storm will determine whether the industry sees clear skies ahead and flourishes again or whether it follows the path of so many other seemingly indispensable professions that are no longer in existence or relevant today.

The economic crisis of 2007-2008 exposed glaring flaws in the valuation processes of single-family appraisals. These flaws led to inflated valuations, appraiser fraud and identity theft, rampant pressure on appraisers and, most importantly, questions in the overall credibility of real estate appraisals. The aftermath of that crisis, however, may continue to cause uncertainty and less credibility as homeowners, lenders and appraisers all seek to understand the fundamental nature of housing values and prices.

For example, one major challenge facing appraisers today is whether identical houses in the same neighborhood can have different values. A decade ago, an appraiser with the task of valuing a home in a neighborhood would have a fairly straightforward job. While there might be nuances to every sale, they were generally uniform in terms of being representative of the general market.

But in the last five years, the opposite is true. Faced with a typical neighborhood that might consist of short sales, foreclosures, investor-owned rental properties, lender-owned properties and other ‘normal’ sales, how does an appraiser look at all of these various indicators to arrive at a value?

With all the varied types of properties, which sales should an appraiser consider and which should be given less weight? Frankly, what do they all mean? The appraisal industry is struggling to deal with these very issues as it attempts to reconcile data that often cannot be reconciled.

Traditionally, the appraiser relied on the format of the appraisal forms to drive the process and eventually reach an ‘opinion of value.’ However, considering the widespread use of sophisticated automated valuation models (AVMs) and abundant access to data, the appraiser is still noticeably lagging behind other valuation services by following a ‘forms approach’ to valuation.

In the 'hood

Over the last four years, more analytics have emerged in the traditional appraisal, including more specific data related to market conditions at the neighborhood level.

The fundamental building block of an appraiser's analysis begins with compiling information on a particular neighborhood. An appraiser may ask, ‘What is happening with the housing stock? What is happening with market trends? And what percentage of sales is traditional, and which ones represent short sales, foreclosures and other distressed housing?’

The answers to these questions help to form and drive the appraiser's opinion of value. Trying to make sense of conflicting information is part of the appraiser's skill set. For example, do foreclosures have an impact on a neighborhood's value? Are short sales having an impact on the level of value in a neighborhood? Many times, there is an absence of clear-cut answers.

Now, more than ever, there is an abundance of data that has increasingly become available for real estate: public record data, multiple listing service data and even consumer-facing sites such as GoogleEarth, Trulia and Zillow, all offer rich and comprehensive data that can be used by appraisers to more completely understand neighborhood trends and valuation.

One solution that is increasingly being offered in the marketplace is coupling an AVM with an appraiser inspection and overview. While this does provide a client with more data, pairing AVMs or ‘black box’ analytics with the appraiser does not fundamentally enhance the credibility of the value conclusion, nor does it enhance the credibility of the appraiser's market analysis skills. In addition, many appraisers are concerned that much of the analysis for these types of products are done by statisticians who may have little or no understanding of the nuances of local neighborhoods.

A second solution to the problem of understanding value at the neighborhood level, however, comes from the growing inclusion of more sophisticated data analytics into the appraisal process. Appraisers are increasingly turning to solutions that incorporate techniques such as regression analysis to gain greater insight into markets. By incorporating regression analysis into the process, appraisers are shifting from the art of appraising to the science of appraising.

A different approach

Instead of comparing a few properties, appraisers can compare hundreds of properties for analysis with a statistical degree of accuracy, strengthening the final value conclusion.

The analytics should be created by the appraiser as opposed to having the appraiser interact with a ‘black box’ AVM. When the appraiser creates the regression model or valuation model specifically for a subject market, the process is known as interactive valuation modeling, and the result is an interactive valuation model (IVM).

IVMs are superior to AVMs because the appraiser with local market expertise builds and ‘fine-tunes’ the model to accurately account for the localized market characteristics. Contrast this to an AVM, where the AVM may not even have adequate market boundaries, let alone account for the local characteristics.

Lenders have sought solutions that incorporate some of this neighborhood-level analysis, and some banks, such as U.S. Bank, the number three-lender in the country, has begun using a collateral valuation report (CVR) option in lieu of traditional appraisal.Â

The CVR incorporates the best of both worlds. It uses all of the key elements that make use of the current market environment: data, analysis, regression, deep drilling into neighborhood-level analysis and other sophisticated information. Within this report format, the neighborhood or market can be more precisely defined by overlying a polygon shape depicting the market boundaries on a street or satellite map. This simply eliminates any confusion that might arise in trying to identify the boundaries.

Appraisers have begun to discover the need to understand and harness this analytical power and incorporate it into their appraisal process. This requires new levels of education and training programs on an industry-wide basis.

As appraisers adopt increasingly more sophisticated valuation techniques that are augmented by robust data and statistical analysis, consumers, lenders and others who rely on an understanding of housing values will benefit from reports that demonstrate greater accuracy, a more transparent and credible process, and ultimately, a more reliable and supportable valuation.

For the residential appraisal industry, confronting the ‘perfect storm' of increasing scrutiny, regulations and constraints, it is a beginning toward a more accurate valuation – and not a moment too soon!

Mark R. Linne is chief strategic and valuation officer for Farmington Hills, Mich.-based AppraisalWorld Inc. and co-author of three books for the Appraisal Institute: ‘A Guide to Appraisal Valuation Modeling,’ ‘Practical Applications in Appraisal Valuation Modeling’ and his newest book ‘Visual Valuation:Â Implementing Valuation Modeling and Geographic Information Solutions.’ He can be reached at (734) 458-4977.


Please enter your comment!
Please enter your name here