Federal Reserve Bank Of Philadelphia - Appraising Home Purchase Appraisals

[Randolph Kinney]

We argue that institutional features of home mortgage lending cause much of the information in appraisals to be lost: some 30 percent of recent appraisals are exactly at the home price (with less than 10 percent below it).

This information loss can be attributed to the appraiser attempting to respond on behalf of the lender to two important institutional issues related to mortgage underwriting and securitization. First, the government-sponsored enterprises (GSEs) and federally regulated banking institutions set up ranges for the loan-to-value (LTV) ratio across which underwriting and mortgage insurance requirements differ.

Second, the GSEs and federally regulated banking institutions require that the home value, which is the denominator of the LTV ratio, be calculated as the lesser of the sale price and the appraisal. As a consequence of this minimum value rule, an appraisal that is below the contract price (the agreed-upon offer price) may increase the down payment needed to stay within the borrower’s desired LTV range.

We model the decision of whether to report the actual appraised value versus whether to bias the appraisal upward (as high as to the contract price) as a tradeoff between the cost of potentially losing the mortgage transaction and the cost of losing informational value from the appraisal.

https://www.philadelphiafed.org/-/m.../publications/working-papers/2017/wp17-23.pdf

The assertion is clear that the lender has a bias to make the loan and the appraiser has a bias knowing the contract price to hit so the loan can be made. The assertion is 50% of the appraisals should be below contract price that would be expected if the appraisals were unbiased. Approximately 30% of appraisals in both datasets precisely equal the contract price.

 

[J Grant]

Anyone who is not a hypocrite has to realize there is no win for appraisers now. The entire system still favors appraisers who hit value, then a paper like this makes the genius observation that many appraisals come in at SC price. The AMC still is biased toward picking deal friendly appraisers. Add in reconsideration of value, the "value rebuttals", Tidewater, appraisers mocking other appraisers who come in lower than SC price etc.

Wouldn't it be ironic, despite an AVM being "better" at predicting defaults ..if lenders and RE agents see the AVM's deriving lower value estimates than appraisals, the lenders and RE agents clamoring to get appraisals back?

The purpose of an appraisal is not to predict default risk, the purpose of appraisal is to develop an opinion of market value. If lenders use AVM's as a tool for default risk analysis, that is a use lending purpose appraisals were not designed for.

Perhaps a hybrid report part appraisal, part AVM would serve a future purpose meaningful to the users and clients..

 

[Eli]

I can't tell you how many tidewater notices I send. I would estimate that 5% meet contract price (1% would be more like it) after I send the tidewater notice. I just did one where the contract price was $675. The AV was around 620's.

I am doing one now where the AV will be around 20k to 30 K more than contract price.

These were both new construction.

 

[hastalavista]

One interesting thing I didn't see. The premise is that without bias, 50% of the appraisals are expected to fall below the contract price, and 50% above the contract price. But I didn't see a test of that theory using AVMs? I'd like to know if the AVMs show the so-called expected 50-50 bell-curve? (is that in there)?


BTW, I didn't see in the paper that AVMs were suggested as a substitute for appraisals. I saw that they may be useful as a better default indicator (but that improvement seemed to be marginal).

 

[Eli]

One interesting thing I didn't see. The premise is that without bias, 50% of the appraisals are expected to fall below the contract price, and 50% above the contract price. But I didn't see a test of that theory using AVMs? I'd like to know if the AVMs show the so-called expected 50-50 bell-curve? (is that in there)?


BTW, I didn't see in the paper that AVMs were suggested as a substitute for appraisals. I saw that they may be useful as a better default indicator (but that improvement seemed to be marginal).

It's like the wind. Lol

I am being serious when I say I think LA and the FtC have some people nervy. No joke.

That includes FRB because they are not exempt.

 

[hastalavista]

From the paper...



It seems to me that the authors have identified one reason why the contract price is exactly 30% of the time. The contract is a data point (how many times have we argued that). It can be the defining data point if it falls within the range.

Here's a new test:
Simulate a purchase on 200 properties; have the purchase negotiation "match" the market dynamics (not impossible to do). Then, have those 200 properties appraised without the simulated purchase contract as part of the disclosure.
The question is: What is the percentage of the simulated contract prices that fall within the adjusted range of the valuation?

 

[George Hatch]

AVMs don't need to perform better or even equally as well as appraisals in order to compete. They're already faster and cheaper so they already have that going for them.

The disincentives the lenders impose on appraisals that don't rubber stamp the contract price are well known to appraisers, so some of the failures are attributable in part to that factor.

Regardless, the way forward for us is to play to our own strengths and to do better at what we are able to add to the valuation.

 

[djd09]

The Propaganda Kings. And that ain't no joke. Soon as i read HVCC, it couldn't be more disingenuous.

 

[Deleted member 134708]

I came across a rather interesting working paper recently published by the Federal Reserve Bank of Philadelphia - WORKING PAPER NO. 17-23 APPRAISING HOME PURCHASE APPRAISALS, which can be found at the following link:

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwjLtoTihJbWAhXMMyYKHQ9ZALQQFggmMAA&url=https://www.philadelphiafed.org/-/media/research-and-data/publications/working-papers/2017/wp17-23.pdf&usg=AFQjCNGSxv91e7E3Aem9aWmAK7SlBaVyTw
The paper's authors conclude that appraisals have an upward bias, there is confirmation bias on purchase related appraisals, and AVM's are more predictive of default than appraisals.

The paper is highly technical, but worth the read. While I don't agree with some of the conclusions and some of the methodologies utilized, I have no doubt that papers like this one have been and will continue to be utilized to justify PIW's/Appraisal Waivers and the current trend away from requiring appraisals in every mortgage loan transaction.

In order to counter the trend away from appraisals, Appraisers as a group are going to have to be able to refute the data and/or methodologies and conclusions from this other papers. Simply arguing that appraisals are better and do a better job assessing collateral risk simply because appraisers say so without any data to support their argument is not going to be a winning argument and is not going stop the trend away from requiring appraisals on every loan. This is why it important for residential appraisers to join a professional organization (such as the AI or NAIFA) and push the leadership of these organizations to fund/conduct there own studies regarding the issues raised in the Philly fed's paper.

Spot on Tim.

My take, put a fork in residential appraisers, they will never come together as a group. If hasn't happens yet, not gonna.

Even if they did come together, their argument would be an opinion based on feelings with zero facts going against a Fed paper with all the facts et al

 

[TRESinc]

In order to counter the trend away from appraisals, Appraisers as a group are going to have to be able to refute the data and/or methodologies and conclusions from this other papers. Simply arguing that appraisals are better and do a better job assessing collateral risk simply because appraisers say so without any data to support their argument is not going to be a winning argument and is not going stop the trend away from requiring appraisals on every loan. This is why it important for residential appraisers to join a professional organization (such as the AI or NAIFA) and push the leadership of these organizations to fund/conduct there own studies regarding the issues raised in the Philly fed's paper.

but i am entitle to have people send me work because after all, i have a license!