Un supported adjustments

[Sadie]

Before you use any method to arrive at an adjustment, you need to understand the methodology. Standard 1 says in part that you must understand and correctly employ recognized methods and techniques. If you don't know how adjustments are derived, perhaps you shouldn't use that method.

 

[Mike Ault]

This discussion reminds me of Jesuit priests discussing "how many angels can dance on the head of a pin?". Pointless esoterica. Appraisals for areas where there is enough conforming sales activity to actually isolate & quantify month-to-month market appreciation/declines in a credible manner will certainly be performed by GSE AVMs. Appraisers operating in the rest of the world will now have to deal with newly minted "underwriters" demanding to see that Freddie Mac graph somewhere in the report and nonlinear market change adjustments with no smoothing on the grid. It would help if the GSE's could disseminate a lesson in statistics alongside their expectations in this regard, including what constitutes a "statistically valid sample size" prior to demanding a granular analysis thereof.

 

[J Grant]

That is NOT what the chart illustrates. Not at all.

It looked that way to me. How else could one month show 4% and then another month 2% and a next month flat, and each month moves in sync with the price of sale A, B and C?

Idk how else to read it - what trend program is so sensitive that it can show a reliable 4% up or down one month then 2%, the next month, then flat the next month?
the time frames are so short

Normally, a trend last for several months or six months, or a year or longer.
The exception is when a market makes a sudden turn, or there is a good benchmark of when a trend stops, such as prices were rising for 5 months and then the market went flat for 4 months, and then began rising..

Normally, there are not month-to-month measurable changes.

 

[alebrewer]

Normally, there are not month-to-month measurable changes.

of course there are. It just takes enough data points to confirm that it really is a delta in the market, as opposed to outliers skewing the trend.

 

[J Grant]

This discussion reminds me of Jesuit priests discussing "how many angels can dance on the head of a pin?". Pointless esoterica. Appraisals for areas where there is enough conforming sales activity to actually isolate & quantify month-to-month market appreciation/declines in a credible manner will certainly be performed by GSE AVMs. Appraisers operating in the rest of the world will now have to deal with newly minted "underwriters" demanding to see that Freddie Mac graph somewhere in the report and nonlinear market change adjustments with no smoothing on the grid. It would help if the GSE's could disseminate a lesson in statistics alongside their expectations in this regard, including what constitutes a "statistically valid sample size" prior to demanding a granular analysis thereof.

If they make the exact same or similar software available, we can provide it.

If not, then we can only provide what we can have access to.
Though IDK if an AVM could creidhly crate such short term time snetive adjutmsens and if so, are they even beneifial in a messy and imperfect market.

 

[J Grant]

of course there are. It just takes enough data points to confirm that it really is a delta in the market, as opposed to outliers skewing the trend.

Then we need to ask what valuation is - a math problem that uses "data points," many of them not similar to the subject and in a program, many of them not verified, or is it a market value, with the most similar comps used and local market conditions considered to develop a MV opinion?

Imo, a lot of this is jut profiteering for the sake of income to tech companies, software providers, or GSE's with much time on their hands now that appraisal volume is reduced. They have to spend their time doing something, and tweaking a chart to show a fraction more accuracy of a statistic is one way to do it,

Math can always be read back to prove itself as "accurate", but if it misreads the market, what did it accomplish?
nd what person can tell, than a highly trained programmer / experienced appraise who is too expensive to use, whether it misread the market?.

 

[J Grant]

Even if the chart is based on a reaction to data and not to the individual comp sale prices, the overly sensitive or month-to-month different rates of adjustments are price-oriented, not market-value-oriented.