Un supported adjustments

[J Grant]

Imo, these tools seem more applicable to time adjustments, and I would only use a software program that allowed me manual control over the data that is entered,

 

[Surf Cat]

Imo, these tools seem more applicable to time adjustments, and I would only use a software program that allowed me manual control over the data that is entered,

Like some here, I am a fan of and primarily use Spark. It works very well for data 12 months back for time adjustments. There's a more recent inclusion I discovered here on the Forum, Valuation Labs, which is free to use (refreshing in comparison to many appraisal tools that want to gouge the appraiser). I created a custom export from my MLS and gave it a whirl. My initial export did not work so, I sent it to the creator of the program and they got back to me very quickly. You have to be dead accurate and label the export "exactly" like example CVS.....no extra spaces (like I had) or it won't work.

It requires 2 years of data for it to start to calculate. The time adjustments in comparison to Spark I found to be different. Seeing how I've only used it a couple of times.....it might be user error. It's a pretty slick tool although I have to play with it some more so, the jury's still out on that one. Another tool I use to verify time adjustments is "InfoSparks" which is included with my MLS. It too produces different results however, it's more broad and you can't dial in the exact range of GLA, year built, and market area like you can with Spark. But you can go 5 years back in time which is beneficial.

Different tools, different results. It's hard to meet the demand of the new trend of "exactness" that some clients want. Case in point, the sample of the way the GSE's want time adjustments in that graph where there was a different time adjustment each month.

Appraising Real Estate isn't an exact science. As I garnered more experience in doing this... I thought that the appraisal should be concluded with a tight range, not an "exact" number. That way, the lender has some wiggle room instead of being handcuffed to a single, solitary, number. I mean....that's what they're doing now with waivers.

 

[J Grant]

Like some here, I am a fan of and primarily use Spark. It works very well for data 12 months back for time adjustments. There's a more recent inclusion I discovered here on the Forum, Valuation Labs, which is free to use (refreshing in comparison to many appraisal tools that want to gouge the appraiser). I created a custom export from my MLS and gave it a whirl. My initial export did not work so, I sent it to the creator of the program and they got back to me very quickly. You have to be dead accurate and label the export "exactly" like example CVS.....no extra spaces (like I had) or it won't work.

It requires 2 years of data for it to start to calculate. The time adjustments in comparison to Spark I found to be different. Seeing how I've only used it a couple of times.....it might be user error. It's a pretty slick tool although I have to play with it some more so, the jury's still out on that one. Another tool I use to verify time adjustments is "InfoSparks" which is included with my MLS. It too produces different results however, it's more broad and you can't dial in the exact range of GLA, year built, and market area like you can with Spark. But you can go 5 years back in time which is beneficial.

Different tools, different results. It's hard to meet the demand of the new trend of "exactness" that some clients want. Case in point, the sample of the way the GSE's want time adjustments in that graph where there was a different time adjustment each month.

Appraising Real Estate isn't an exact science. As I garnered more experience in doing this... I thought that the appraisal should be concluded with a tight range, not an "exact" number. That way, the lender has some wiggle room instead of being handcuffed to a single, solitary, number. I mean....that's what they're doing now with waivers.

Note, Danny from the GSE's commented here that they do not fully stop want or require ) the time adjustments on that graph - yet the graph was posted on their website as an example, so go figure. They do seem to want a more detailed statistical time adjustment shown than the market conditions grid.

I found the GSE claim that not enough appraises made time adjustments weird. The market has been stable for the last 18 months in many areas - so where does a time adjustment apply? The letter then stated that a lack of applying time adjustments in minority areas led to under-valuations!

Which makes no sense. Perhaps they did not get the notice that a new sheriff is in town ( Trump) who will be going after DEI-related govt things - I am not a fan of his, but in some cases, it is warranted.

 

[Terrel L. Shields]

The letter then stated that a lack of applying time adjustments in minority areas led to under-valuations!

Yep, I have to take that with a rather jaundiced eye. I mean unless prices are changing at breakneck speed lack of a 1 or 2% adjustment isn't going to move the needle much. Our market when moving up, was typically 4-6% per year. That's what? half a percent or less per month? Six-month sale 2-3% at max?

I have never seen a market move so fast that its momentum was more than a year or two. When a big announcement of a theme park in NE Oklahoma land prices zoomed up. But the promoters had already bought the land they wanted using straw buyers. Everyone raised their prices but no one wanted to buy it. Prices then begin to float downward.

 

[litchke]

I don't put the raw data into my report for 2 reasons.
1. There is too much temptation to think, "here is the data, you figure it out", rather than taking the time to explain how you arrived at the adjustments.

2. As I have said here many times before, "people have just enough information to be dangerous". Putting raw data into reports...and doing a poor or incomplete explanation of what it means...... results in the reader making their own conclusions of what it means. If you just say, " Spark/Synapse was used to estimate the adjustments", the reader can say, "well I see the data and I say it means something different".

could you provide an example of something you would say for an adjustment that you applied?

 

[bsilver]

I think that before even discussing the pros and cons of using 'adjustment machines', some attention needs to be paid to exactly what 'support' consists of, whether it be support for adjustments, a cap rate choice, or a highest and best use determination. Support consists of reporting the relevant evidence (market data/analysis) and logic (rationale) utilized in developing the appraisal to the degree and extent required to provide credible assignment results in the context of the intended use. The appraiser must decide when developing the scope of work just how detailed the reporting of evidence and logical reasoning needs to be for the intended use. Intended users should be able to understand how conclusions were developed sufficiently enough so they believe the report to be credible.

Depending on the client you may have to show all the sales data, all the analysis, and explain it all in excruciating detail. Other clients may accept a description of the analyses with the conclusion stated, and a brief statement on the rationale for its reasonableness. And some analyses have more market support than others. A square foot adjustment has more numerical data to produce a conclusion than the somewhat awkward layout you felt required an adjustment. Bottom line, you need to include at the very least, a description your evidence and the rationale for your conclusion.

As to utilizing analysis software, really it's an AVM, is fine as long as it results in a credible report given the intended use. It sounds like the OP simply attached the output and referenced it for the adjustment explanations. Having not seen the report, I'm guessing as to what was and wasn't included. What is your evidence and what is your rationale for its credibility?

In using an AVM, simply pointing to the output says very little about either the data or how it was analyzed. Did you describe the methods each analysis used to derive the figures? (not the math, the concepts). Did you sort through the data that the program used? Did it have any outliers that should have been removed? Did it reject data it should have used? Was the market area it derived the observations from too broad or too restrictive? What database did it use? Is the database reliable? How many sales did it select from? How many were rejected? Were any statistical results accompanied by associated variance or confidence stats? Did the paired sales analysis use good pairs, or should adjustments have been made? These are examples. This seems like more than it is. But I suspect that these sorts of items were absent your report.

One last comment on market condition adjustments. I have reviewed many dozens of reports over the past few years, and I definitely saw a lack of time adjustments. Now, granted, the reports I was reviewing were unusual properties and the time frame researched generally spanned two to five years. Report after report contained statement like, the market did not markedly appreciate or depreciate, hence no market condition adjustments were required. What's the matter with that? It's not that an adjustment wasn't made. There was no evidence presented. Over and over I saw this. It's probably a true statement! There are so many indicators that one could simply cite, MLS stats, state or county stats, FRED, rent levels, land sales, resales, etc etc. Anything but nothing! Someone stated a 6% per year appreciation so a 1 or 2% adjustment is silly. It is. But where'd ya get the numbers? Just drop where you got the knowledge in the report, analysis complete. I'm not pointing fingers, and this has just been a nitpicky annoyance for me recently.

 

[Terrel L. Shields]

In using an AVM, simply pointing to the output says very little about either the data or how it was analyzed.

The algorithms are secret. Don't tell anyone.

Someone stated a 6% per year appreciation so a 1 or 2% adjustment is silly. It is. But where'd ya get the numbers?

No one says you have to calculate that - but like an extraordinary assumption and a simple assumption. Some things are simply logical, intuitive, or common public knowledge. I don't state the size of the water line in front of the house, but I could look that up and "support" my assumption that the water pressure is OK when I see water flowing out of a faucet.

NWA real estate market remains healthy; Skyline Report turns 20 - Talk Business & Politics

The Northwest Arkansas real estate market experienced strong home sales and low vacancy rates in multifamily and commercial properties in the first half of 2024, according to Arvest Bank’s new...

talkbusiness.net

This is a sample. You have to log in and a subscription is required unless you can find a friendly banker to provide it to you. After all, only Fannie requires you to include support in the report, doesn't it? I mean the data are readily available 'by reference' (I believe that is noted in USPAP.)

 

[J Grant]

Yep, I have to take that with a rather jaundiced eye. I mean unless prices are changing at breakneck speed lack of a 1 or 2% adjustment isn't going to move the needle much. Our market when moving up, was typically 4-6% per year. That's what? half a percent or less per month? Six-month sale 2-3% at max?

I have never seen a market move so fast that its momentum was more than a year or two. When a big announcement of a theme park in NE Oklahoma land prices zoomed up. But the promoters had already bought the land they wanted using straw buyers. Everyone raised their prices but no one wanted to buy it. Prices then begin to float downward.

The problem with the chart is that it is backward.

It is not applying a market trend line condition trend to the sales.

It is taking each sale and its individual sale price and then "adjusting" a trend line tailored to that particular price to make the sales all come within a close range to search other .

So if Comp B sold 5% higher the market trend adjustment for it is 5%, and if Sale B a month earlier sold for 2%, higher, the adjustment for it is 2%. The reality, though, is within a trend; individual properties can sell for prices both higher and lower than the trend line. This is what happens in an imperfect market, so leave it alone. Or, the higher or lower price can be do to a certain feature that needs an adjustment.

In lieu of an adjustment, for small differences like 1-3%, for example, we can reconcile on the higher or lower range.

Also, if at some point in the time frame, such as the first 4 months, the market appreciated and then flattened to stable, that is a reason to adjust comps selling within the different time frames sidfiernely (with no adjustment for the sales that were negotiated when the market stabilized.

 

[Terrel L. Shields]

The reality, though, is within a trend; individual properties can sell for prices both higher and lower than the trend line.

Exactly and without a huge amount of data (like rural sales) then the "noise" - deviations from the trendline- can be large enough to look like the change in prices was larger than it really was.

 

[DWiley]

The problem with the chart is that it is backward.

It is not applying a market trend line condition trend to the sales.

It is taking each sale and its individual sale price and then "adjusting" a trend line tailored to that particular price to make the sales all come within a close range to search other .

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