Fannie Mae recently released its Collateral Underwriter automated risk assessment tool, which evaluates the risk associated with appraisals. It does not accept or reject appraisals, nor does it provide lenders with an estimate of a property's value.
CU is at the forefront of the field we can describe as "the science of valuation." To assess risk, the tool analyzes large amounts of data from a variety of sources, including the UAD dataset. Essentially, CU is a very sophisticated data analysis system with a built-in automated valuation model. One of its checks is for overvaluation. If CU's analytics indicate an appraiser has overvalued a property, that appraisal will definitely be flagged for further review. Interestingly, undervaluing a property is not flagged.
In addition to property eligibility and policy compliance, CU analyzes four key appraisal components:
Data Integrity
- Are physical attributes and transaction terms accurately reported? Fannie Mae has robust data on most properties, and they'll be checking to see if appraisers have been consistent with themselves, as well as their peers, when coming up with property valuations. Appraisers need to effectively keep track of whether they've been consistent with themselves, and one way to do that is by maintaining a database of the properties they've appraised.
- Ensuring consistency with their peers presents more of a challenge. The thinking behind it is that like-minded appraisers should produce similar results, and if an appraiser is not consistent with their peers, Fannie Mae will flag the report. It may be that an appraisal report flagged for not being consistent with other reports turns out to be the most accurate. The key is to be able to support and defend the accuracy of the data in their report.
Comparable Selection
- Are the selected comps reflective of the subject property? CU ranks appraiser-rated comps against a pool of available comps found most suitable by Fannie Mae's data and algorithms. Previously held notions on geographic range won't apply anymore — appraisers should know that the comps Fannie Mae deems the most appropriate are not necessarily those within one mile of the subject property. It will depend on the market. Also, the standard time frames appraisers are accustomed to may no longer apply. Fannie Mae will consider comps that are up to a year old.
Adjustments
- Are adjustments based on typical market reaction? I can't emphasize enough that the hard and fast rules appraisers are accustomed to using will not necessarily apply anymore for adjustments. Regression analysis will. For example, appraisers should be cautious about using traditional gross living area adjustment factors of $25 to $35 per square foot. There's nothing wrong with using a $25 per square foot adjustment factor if it is representative of the market. However, in many cases across the country, $25 per square foot is not representative of the market. If appraisers continue to use an adjustment factor of $25 per square foot, their appraisals will likely be flagged. Of these four components, I believe adjustments will be the cause for most concern.
Reconciliation
- Are the most recent comps given the most weight in reconciliation? This is an area where appraisers may get flagged and will need to start defending themselves. It's becoming increasingly important that appraisers have analytical support for the adjustments that they make.
Fannie Mae's CU assesses the risk associated with an appraisal by analyzing large amounts of data and sophisticated models. Appraisers need to start doing the same. In order to properly defend and support your appraisals, arm yourself with the same tools. Technology firms are rolling out sophisticated analytic tools with regression analysis that are easily integrated into appraisers' existing form-filler software.
The bottom line? Appraisers who rely only on their experience and judgment without verifying and supporting their appraisals with market analytics will likely find their reports will be flagged by Fannie Mae. Those who don't make use of analytics similar to CU will find themselves struggling to support and defend their appraisal reports. The only way to offset the advantage these review systems have is to employ similar market analytics of your own. It's what I call computer-aided appraising: an appraiser with expert knowledge of the market and armed with good market analysis software will have no problems producing an accurate and defensible property valuation. Don't be left behind.
Jeff Bradford is the founder & CEO of
