1004MC, Lack of Data an Appraiser Guess

[prasercat]

I submit the report, 1004MC has 1 sales in each period. The listing data is similar. I don't check the trend boxes and explain why.

The client is telling me that HUD requires the boxes to be checked because the appraiser should be familiar enough with the market to basically guess the trends.

This sounds fishy, but in this environment, anything seems possible.

Does anyone have a link to such a statement from HUD?

Basically, what Denis said.

Guessing is the Art of appraisal and how skippies turn appraisals so fast.

Skip the art at all cost.

This situation can be really tough. If you go outside of what you would normally consider a competitive area to spot areas well outside the normal boundaries to those areas that have similar prestige, you have to be more specific in your search. See the differences between including REO's and not in your analysis. Just REO's, check their trends, etc. It seems to me that under 40 sales makes this analysis rather unreliable, better 60 to 100. If things get really tuff and you are getting alot of very different properties (million dollar homes with 100k homes, etc), try to find what is unique about those odd balls "non-comps" and exclude them in the search, it may be excluding a location, a subdivision, a town, those on a river, those that are short-sales, those without photos, those with fixup, etc.,even those outside a large price bracket (this is something one could use for really far outlying values). Try different searches until a meaningful number is achieved, remove clear outliers (such as, price, price/square foot, age, etc). Note, when market times are short and REO's are selling at a similar price to non-REO's and they are few in number, I just include them in the mix. Most often, lately, this is not the case, so a separate study is done, one for the REO trends and one without. In any case, excluding fix-ups, those without photos, is something I almost always do.

While the neighborhood stats on page 1 and above the first sales grid should reflect the stats from the 1004MC study, this may not be the case when you have to expand the market so much to get enough data. I find it better to zoom in on my area where I would normally put the boundaries and use those stats on these two pages since they are more indicative of the subject immediate neighborhood (which is more relevant to competition and regression/progression) - If you don't do this, they'll wonder why you have found so many "comp sales" in 12 months, yet you say they are scarce, etc.

I then explain in the 1004MC this diference in the data pools (or use a reference to see the addendum that can't be missed), why this deviation in procedure was applied and that in order to find sufficient data points (due to a lack of comparable sales) regarding the market surrounding and most affecting the subject, it was necessary to expand the market beyond the typically considered boundaries and parameters. I then explain the search critieria used in each "neighborhood".