1004mc Soon Not To Be Required By Fannie Mae

[Joe Flacco]

That may be true but also the parameters chosen for the data can have a huge effect on the results. Something as simple as moving or expanding a boundary, or not including REO sales can show differing results. The adjustments for timing can be huge, which is why perhaps many appraisers are reluctant to adjust for conditions.

What I have noticed is that usually almost all of the price gains for the year occur during the spring season. So if the overall trend (YoY) is increasing then I'll report increasing but after the spring sales start closing, adjustments are often unnecessary for the rest of the year. That price jump going from winter into spring gets almost the whole years increase adjustment.

 

[NC Appraising]

You have never seen actual trend conflict with the town and zip code trend? Like when declining number of distressed sales results in zip code median price increasing while prices are actually stable? Or when prices at the low end are jumping sharply while mid- higher priced homes are stable but all together shows median price for zip code increasing? It is not that simple. The best you got is that paired sale of the same house over 2 years.

That data in the graph both bar and line can be further delineated by GLA, site size, etc.. I'm not getting paid for this.....you wanted something I gave it to you.

Most of the time the sales grid will be your best supporting data IF you use active/pendings a and a mix of older and newer sales.

If you only use comparable sales that closed within the past 90 days and not even research older sales then yes, I can see appraisers being conditioned to a stable market.

 

[glenn walker]

Stable increase- or Stable Decline - You have been Thunder Struck : ) LOL

 

[Joe Flacco]

That data in the graph both bar and line can be further delineated by GLA, site size, etc.. I'm not getting paid for this.....you wanted something I gave it to you.

Most of the time the sales grid will be your best supporting data IF you use active/pendings a and a mix of older and newer sales.

I think like you said before looking up the sales history in the subdivision is probably one of the best ways to do a quick look at trend. But then when there is nothing very.recent in the subdivision going into spring season it still gets pretty complex.

 

[Joe Flacco]

How do you adjust for this sudden spring market pop before the data shows? May and June closings showing the breakout which means probably March, April, May contracts. Only way I can think of is to get contract prices out of listing agents that were pending at the beginning of Spring. Active listings were indicating that this might happen however.

 

[Evincere]

Really, CU, through the UCDP has all of the data from the appraisal that you prepared and was uploaded to it plus the data from all of the other GSE appraisals done in that market area going back since 2012.....there is zero chance that the CU moodel cannot model current trends at least as well and probably better than the appraiser since CU has more extensive complete market data than the appraiser has.

So if Appraisers per FAnnie have been filling out 1004MCs since 2012 with what Fannie considers to be worthless / useless, non informative data how worthless / useless will the current trends be that CU models spit out. Will it not be regurgitating the same garbage that was put into it? The machine cannot know what it doesn’t know.

Unless the Fannie has some secret AI it’s all GIGO

 

[Dublin ohio]

Unless the Fannie has some secret AI it’s all GIGO

No secret. Fannie has data from all the comparable sales and data from subject properties from appraisals that have been submitted. No reason for Fannie to use 1004mc data for CU. They have more than enough data to develop trends and more sophisticated systems than those available to the typical residential appraiser.

 

[George Hatch]

What I have noticed is that usually almost all of the price gains for the year occur during the spring season. So if the overall trend (YoY) is increasing then I'll report increasing but after the spring sales start closing, adjustments are often unnecessary for the rest of the year. That price jump going from winter into spring gets almost the whole years increase adjustment.

In line with the above, in many market segments the volumes are not that large so the appraiser winds up with small datasets. But the other situation that occurs with many of the markets in our region is that most of the action in pricing occurs during fairly brief periods interspersed with longer periods with less activity but more stable pricing/.. The pricing increases aren't occurring on the strictly linear line but in fits and spurts. Enough so that at what point of the year we're at the preceding increments may not line up with the majority of the activity.

Really, 2yrs at 6-month increments might demonstrate more obvious trend indicators, both by expanding the time periods involved (less overlap between the spasms) and by doubling the volumes per period - thus making the medians more usable. So there's your tweak to the 1004MC or its equivalent.

 

[Joe Flacco]

In line with the above, in many market segments the volumes are not that large so the appraiser winds up with small datasets. But the other situation that occurs with many of the markets in our region is that most of the action in pricing occurs during fairly brief periods interspersed with longer periods with less activity but more stable pricing/.. The pricing increases aren't occurring on the strictly linear line but in fits and spurts. Enough so that at what point of the year we're at the preceding increments may not line up with the majority of the activity.

Really, 2yrs at 6-month increments might demonstrate more obvious trend indicators, both by expanding the time periods involved (less overlap between the spasms) and by doubling the volumes per period - thus making the medians more usable. So there's your tweak to the 1004MC or its equivalent.

They should pay you to help them with a new form.

 

[George Hatch]

Meh, there are probably half a dozen different ways to set that analysis up, and lots of people who could do it. Bear in mind, when Fannie first brought the MC form online there wasn't a single software solution for it, and we had to include instruction on the difference between an average vs a median because a lot of appraisers didn't know the difference. Nevertheless, by the time the form became mandatory a few months later there were at least 3 freeware solutions for it that I know of and several other available solutions that were available for cheap, and almost everyone had access to at least one functional solution.

Fannie changed their program and the appraisers adapted forthwith because the alternative was to starve. As it turns out, sink/swim is all the motivation appraisers need to do the work. It's when they're not being threatened with extinction that some of those appraisers will look for corners to cut.

Anyways, with appraisers having already figured the existing MC out, tweaking the existing worksheets to use longer timespans is literally limited to editing the if/then equations in several cells on one line for the different time frames, then copy and paste than line through the remainder of the worksheet. That's maybe an hour's worth of work, including the debug.