1004MC & Neighborhood Section page 1

[DWiley]

If you fine-tune too much you're left with too little data...

No comment in this thread better illustrates the problem - What do you mean by "too little data"? Too little for what? The data is the data. As my stats professor made us say every day in my experimental deisgn class, "There is no such thing as a valid sample."

The grid at the top of the 1004MC is supposed to report only comparable sales in the subject's neighbrohood. Period. If there are no sales, report that. If there are 150 sales, report that. It is a pretty simple concept.

If the grid does contain enough data points for trend analysis (and it probably won't) then supplement the grid with whatever it takes to support the reported trends. I have to do that most of the time. There is nothing really complicated about it. Supposedly, we have had to to this since we have been appraising, so I don't see why it should be such a big deal.

For all who are now so perplexed about what a "neighborhoood" is - If this term is so nebulous, how in the world have you been providing "neighbohood" boundaries all these years?

I have said many times that I am not a big fan of the form. We wouldn't even have this form if so many appraisers weren't so lazy. None of that means that I am free to ignore assignment conditions.

For anyone who doesn't want to fill out the form the way Fannie says, the solution is simple - don't do Fannie work. On the other hand, if you don't want to take that drastic step, then don't be surprised if a state board rings you up for certifying that you are complying with appicable assignment conditions while at the same time you are getting on Internet web sites and proclaiming to the world that you are intentionally ignoring them.

Appraisers tend to have high self confidence, which makes it very difficult for us to admit it when we are doing something wrong. After the 1004MC form was released I discovered that I had not been filling the form out the way Fannie wanted. Like many, I used to think that the trend boxes on page 1 applied to the whole neighborhood. Now I know they are specific to the subject's submarket. I was wrong. I changed. I suppose I could just whine about it, but changing seemed more productive.

Maybe some of you can drop into the seminar I am doing on "The Top 20 Mistakes Appraisers Make in Completing the 1004 and the 1004 MC" in April. There are sure to be some fireworks.

 

[residentialguy]

ResGuy, I think you should consider rolling with the punch and use manually selected comparables from the one year period to generate the 1004MC data on NorthstarMLS & use that output to fill in the top of page 2, URAR.

I do that. The only time that I don't is when I only have a few comps. Then I expand the search parameters (as FNMA suggests) and put those results in the MC analysis. But I still show the actual comps on top of page 2 to show the reader how little I had to choose from, which would show why I had to expand search and/or time frame.

 

[residentialguy]

Maybe some of you can drop into the seminar I am doing on "The Top 20 Mistakes Appraiser Make in Completing the 1004 and the 1004 MC" in April. There are sure to be some fireworks.

Sounds like fun! You are buying the round-trip plane ticket, right?

 

[CANative]

No comment in this thread better illustrates the problem - What do you mean by "too little data"? Too little for what? The data is the data. As my stats professor made us say every day in my experimental deisgn class, "There is no such thing as a valid sample."

The 1004MC is not designed for sorting out the fewest ideal comparable sales. It's for supporting the concusion of your pricing trends and inventory analysis.

Ideal matches are good but less than ideal properties are included in the overall maket. If there are a lot more fairly similar properties than there are ideallying similar properties the trend analysis will be skewed.

The form is too simplistic. I reduces appraisal to mathematical calculations and that's NOT appraisal.

 

[Mike Garrett RAA]

I think Danny should fly to Colorado Springs and put on his seminar!

Back to the comparable issue...shouldn't we apply the same criteria we use for selection of comparables used in the grid, ie., similiar models, bracketing the square footage by, say, 10%. net/gross adjustments of 15%/25%, etc?

 

[DWiley]

The form is too simplistic. I reduces appraisal to mathematical calculations and that's NOT appraisal.

The grid at the top of the form is simplistic. The entire form is not - at least not when completed correctly.

I just finished a report for a property with few comparables sales. The grid at the top of the 1004 MC was very simple - the first line was filled with 0's and 1's

The supplemental data I added in the middle of the 1004MC was not simple at all.

 

[CANative]

Danny... I'm not saying I don't prepare this form to Fannie requirements (although I've probably screwed a few of them up). I'm just saying it's a very poor tool for the most part and will likely cause problems in the future (IMO). I'm pointing out what I think the weaknesses are.


Here's what I think should happen.

The form should be discontinued as an addendum and incorporporated directly into the Fannie Form as a Page 7 or 8 instead of the usually 6 or 7 page form. The two lines at the top of Page 2 should be eliminated. The forms producers should integrate an excel program with the ability to import the appraisers MLS refined search results. This would then auto-populate the form. Page 2 of this form could be used for a spreadsheet of all properties used to populate the form and a large comment section for explaining the search parameters and what the results mean. And since it's excel driven, page 2 could also contain some graphs or charts to look at.

 

[DWiley]

I think Danny should fly to Colorado Springs and put on his seminar!

Now THAT would be fun (for me). I think you would get a kick out the examples I have lifted from actual cases.

Back to the comparable issue...shouldn't we apply the same criteria we use for selection of comparables used in the grid, ie., similiar models, bracketing the square footage by, say, 10%. net/gross adjustments of 15%/25%, etc?

As a starting point, yes. But, in the adjustment grid we often have to use sales that are not comparable and/or are not in the subject's neighborhood. Such sales should not be reported in the 1004MC grid.

I appraised an historic home on an acreage tract. The sales I used in the comparison approach were NOT included as part of the data in the 1004MC grid because they were not in the subject's neighborhood.

 

[residentialguy]

The supplemental data I added in the middle of the 1004MC was not simple at all.

Hey, as long as it fits on those 8 lines... :new_all_coholic::new_all_coholic:

 

[DWiley]

I'm not saying I don't prepare this form to Fannie requirements

That comment was not aimed at you.

Some have posted very clearly that they do not follow Fannie's instructions and do not intend to. To those I am saying that there might be consequences.

Here's what I think should happen...

If you are good with the search tool I think you will find that I said a similar thing long ago. As analytical tools improve I think we will see continual expansion of the form to include support for many things, not just market conditions. For example, I see a day when the standard residential report will have to include support for the adjustments.