This paper came out in 2017 but is making the rounds again today. The authors believe they found evidence of appraisal overvaluation in the sales comparison approach in appraisals in 2015 and 2016. I read the article and replied below. Directly below is their summary.
"Abstract
Home appraisal came under scrutiny for contributing
to the home-price bubble and enabling the origination
of risky mortgages that led to the post-2006 foreclosure crisis. Subsequent regulations tried to minimize or
eliminate conflicts of interest and improve valuations.
Nonetheless, our study of appraisals completed in 2015
and 2016 find that appraisal bias still occurred. Our analysis delves into the underlying appraisal development
to identify causes of appraisal bias. Contributing factors
are that comps are generally higher valued than the subject property, and appraisers are more likely to comparatively adjust upward lower priced comps but less likely
to adjust downward higher priced comps.
KEYWORDS
housing price determination, residential appraisal"
"Using a large sample of appraisals ordered to underwrite purchase-money loan applications
in 2015 and 2016, we find that about 69% of comps were represented by transactions valued at
more than the subject property. After controlling for the availability of market transactions and
a variety of loan and property characteristics, we continue to find a greater likelihood of selecting more expensive comps."
https://millersamuel.com/files/2022/06/AppraiserOvervaluation.pdf
My reply: The first issue to consider is 2015 and 2016 markets were increasing in value. It makes sense there would be more higher than lower priced comps available. In a decreasing market you would probably see the opposite.
The second issue is lenders, AMC, underwriters demand that the subject size, bed/bath count and main amenities are bracketed with comps currently listed, pending or sold within three months of the appraisal. Three sold, one pending and one listed comps are mandatory. If you intentionally select properties which are larger and have more bed, baths than subject, they will most likely have sold for a higher price than subject. Sometimes you can bracket size with one comp but need another to bracket bed/bath count. This means you have potentially two out of the three or 66% sold comps which will be, must be superior to subject by design. You still need a comp or comps that bracket the lower end of size, bed/bath count...
The third issue is supply and demand. How many months of housing supply were available at the time? Was it similar to 2020/2021 where there weren't enough homes available for sale so most properties sold instantly over market? There's no box in the grid to adjust for that.
The fourth issue is there were many developer flips during this time. The properties were bought low, improved with new kitchen, baths, paint, landscaping...then sold higher. Of course they could be in better condition all around than subject if the subject wasn't also a speculator flip.
The fifth issue is home attributes which affect value that are not covered by the adjusted factors in the grid or cannot be exactly measured. Those attributes could be curb appeal, staging, marketing, landscaping, superior architect, superior design/remodel, superior layout, custom lot shape, specific superior view, upgraded non line item features... We can only include adjustments which we make using matched pair statistical analysis using math. It's not always possible to extract a value for staging, marketing, landscaping, upgraded non line item features, three tone paint, superior layout, distinct view...
The writers' conclusion below,
"Our evidence of calibration bias, although helpful to understand appraisal overvaluation, cannot answer the question of how it occurs without further research into the development of calibrated values. To that end, a hedonic pricing model can be a helpful tool to assess the accuracy of a price adjustment associated with a given property attribute or amenity while also ensuring consistency and accuracy in the aggregate. Incorporating hedonic pricing techniques into appraisal development is limited. Given the amount of detailed property information and modeling technology that are available today, the appraisal practice could benefit from employing data-driven analytics to help assess appraisal accuracy in an efficient and quantifiable manner. Future research in this direction can develop a better understanding of the underlying appraisal valuation process."
I'm glad they admitted they don't know how the alleged "over valuation" occurred. Maybe there was no overvaluation. Maybe it was the result of what I stated above or maybe it was something else. Would have been nice to do the same in a decreasing market for comparison. While data-driven analysis is always good you will still need a human to appraise real estate, homes for now because humans are the ones buying, paying for and owning them. The contract price is based on the human's level of desire for the property.
There are some things a human can see, find valuable that an AVM robot or statistics cannot. For example a Mercedes is worth more than another similar car because it has a superior design and is more appealing to the eye and market. So is a Louis Vuitton, Gucci, Chanel purse for the same reasons. Kelly Blue Book values and company sales records are the proof of this. The items are generally made out of the same amount of similar materials yet one is worth more than the other. Design, appeal, every factor in market reaction can't be easily calculated in a form appraisal.
Citation: Mayer YG, Nothaft FE. Appraisal overvaluation: Evidence of price adjustment bias in sales comparisons. Real Estate Economics. 2022;50:;862–881. https://doi.org/10.1111/1540-6229.12351
Mary Cummins of Cummins Real Estate is a certified residential licensed appraiser in Los Angeles, California. Mary Cummins is licensed by the California Bureau of Real Estate appraisers and has over 35 years of experience.
Mary Cummins, Mary K. Cummins, Mary Katherine Cummins, Mary, Cummins, #marycummins #animaladvocates #losangeles #california #wildlife #wildliferehabilitation #wildliferehabilitator #realestate #realestateappraiser #realestateappraisal #lawsuit real estate, appraiser, appraisal, instructor, teacher, Los Angeles, Santa Monica, Beverly Hills, Pasadena, Brentwood, Bel Air, California, licensed, permitted, certified, single family, condo, condominium, pud, hud, fannie mae, freddie mac, fha, uspap, certified, residential, certified resident, apartment building, multi-family, commercial, industrial, expert witness, civil, criminal, orea, dre, brea insurance, bonded, experienced, bilingual, spanish, english, form, 1004, 2055, 1073, land, raw, acreage, vacant, insurance, cost, income approach, market analysis, comparative, theory, appraisal theory, cost approach, sales, matched pairs, plot, plat, map, diagram, photo, photographs, photography, rear, front, street, subject, comparable, sold, listed, active, pending, expired, cancelled, listing, mls, multiple listing service, claw, themls, historical appraisal, facebook, linkedin DISCLAIMER: https://mary--cummins.blogspot.com/p/disclaimer-privacy-policy-for-blogs-by.html
