Freddie Mac Research - Appraisals

[DWiley]

FYI

20210920_home_appraisals - Freddie Mac


Freddie Mac Research Explores Causes for the Appraisal Valuation Gap for Homeowners in Minority Neighborhoods | Freddie Mac (gcs-web.com)

 

[TerryRohrer]

What proportion of the various gaps discussed are explained by these variables? For example, if each contributed 1.25% to the appraisal gap, that could explain the entire concluded gap (and in my view. 1.25% is a modest amount). Just wondering how the motivations of realtors, lenders, buyers, sellers, home inspectors, and all other housing market participants are being measured and explained leading up to the entire "gap" being explained solely by appraisals.

"Differences in comparable sale distances, comparable reconciliation, variances in comparable sale prices and possible systematic overpayment for properties by minorities can explain only a modest amount of the observed appraisal gaps for the minority tracts."

 

[Joe Flacco]

It's a very compelling research note.

Some other possible causes of the gap to be explored include

1) The rate of change in prices during the sample period (2015-2020) in the tracts

It is highly likely that there is a correlation with between higher rate of change in values with higher percentage of appraisal values being lower than contract price.

2) The percentage of appraisers located in the tract in which the property being appraised is located.

With the small percentage of appraisers being minorities, it is likely that the appraisers appraising in the minority tracts are not located in the minority tracts. The appraisers may not adequately understand the motivations, preferences, or expectations of the market participants in the tract.

 

[Joe Flacco]

3) The percentage of appraisals that are lower than contract price due to the practice of subtracting seller concessions from comparable properties dollar for dollar in the different tracts.

 

[TerryRohrer]

The appraisers may not adequately understand the motivations, preferences, or expectations of the market participants in the tract.

What would these include that would not be discernable by analysis of the sales in the tract? It may well be that if each appraiser had the staff and budget of Freddie and/or Fannie and/or HUD, each could develop models in every neighborhood, for every assignment, and be able to measure the impact of every detail on price to the nearest $1 every time. But, I doubt it. That leaves me concerned that there is any validity to research that suggests the entire variance among 20 million transactions involving several times that number of individuals (buyers, sellers, realtors, lenders, families, relatives, or everyone who may typically or occasionally or always is somehow involved in decisions regarding residential real estate) can conclusively be boiled down the motivations of fewer than 40,000 appraisers with any certainty.

 

[Joe Flacco]

4) The percentage involvement of Appraisal Management Companies in the appraisal assignments in the different tracts.

 

[alebrewer]

I'm sure this was considered, but in reading through the section on 'overpayment', I didn't see where overpayment due to inclusion of settlement costs was addressed. IOW, it can be assumed that, in lower income neighborhoods, purchasers of homes may not be as prepared financially to pay for closing costs associated with procuring a loan. One means of easing this burden is to roll said costs into the sales price of the home via seller concessions. Is it possible that this is a statistically significant reason appraised values did not meet contract sales price for minority buyers moreso than for white buyers?

 

[Joe Flacco]

What would these include that would not be discernable by analysis of the sales in the tract? It may well be that if each appraiser had the staff and budget of Freddie and/or Fannie and/or HUD, each could develop models in every neighborhood, for every assignment, and be able to measure the impact of every detail on price to the nearest $1 every time. But, I doubt it. That leaves me concerned that there is any validity to research that suggests the entire variance among 20 million transactions involving several times that number of individuals (buyers, sellers, realtors, lenders, families, relatives, or everyone who may typically or occasionally or always is somehow involved in decisions regarding residential real estate) can conclusively be boiled down the motivations of fewer than 40,000 appraisers with any certainty.

I'm just throwing out some ideas that are not explored in the research note that could cause the gap and likely to have strong correlation.

 

[Joe Flacco]

I'm sure this was considered, but in reading through the section on 'overpayment', I didn't see where overpayment due to inclusion of settlement costs was addressed. IOW, it can be assumed that, in lower income neighborhoods, purchasers of homes may not be as prepared financially to pay for closing costs associated with procuring a loan. One means of easing this burden is to roll said costs into the sales price of the home via seller concessions. Is it possible that this is a statistically significant reason appraised values did not meet contract sales price for minority buyers moreso than for white buyers?

It is not minority buyers vs white buyers. It is properties located in black or latino tracts vs properties located in white tracts.

 

[alebrewer]

It is not minority buyers vs white buyers. It is properties located in black or latino tracts vs properties located in white tracts.

Thank you for that clarification.