When lender ask for additional comparables?

[OminousRun]

Yes! :icon_twisted:

I examine solds and active listings reducing each from over 70 down to the 15-20 most comparable per search (and often doing multiple searches) using what I know about the subject to reduce the over 70 down to 15-20. I then use the data on the 15-20 from each search to reduce it down to the most comparable 6-12. I type in those 6-12 and weed out even further until I have what I feel I need. So, I have "used" 15-70 per search and thus I believe I hit your statement requirement.

You mention 30 due to statistical significance. That is fine, but remember that 30 would be expected to include the extremes, extremes we appraisers weed out as "not comparable". When doing a statistical analysis (such as a 1004mc) we may need those 30 to lend significance, but for the report we have continually excluded the less comparable and thus 30 is no longer the significant number. There is a reason "3+" is the "standard" for a typical (non-complex) property. :icon_mrgreen:

I'm happy to see you analyze large data set (which I do too). I'm sure many of the best appraisers on AF (Webbed included!!!) analyze large data sets, for sound statistical purposes, when completing an appraisal report. The more data you analyze, the lower the variance should be, resulting in a more accurate final estimate....

However....

We all know there are certain flaws with the way most appraisers complete their appraisal reports. Usually they start with the opposite methodology. They look for 1 or 2 "good comps". What's a good comp? I'm not sure. They find what they believe is the perfect comp (it may very well be) and build their report around that one comp. Unfortunately that method may lead to problems in the report, and may lead to problems with the UW, etc.

I find the more comps I put on a report and the more comps I analyze, the more accurate my appraisal becomes........it's an amazing thing!!!

 

[CANative]

Here's the market data for Covelo. Two years worth. It takes me 1.5 hours to get to the turn off to Covelo and then another hour to get to Covelo. I've done about 30 appraisals there over the years.

Please analyze this for good statistical support. Keep in mind that no matter what the subject property is these are the only sales that can be used because they are the only sales there are.

 

[OminousRun]

Yeah, the sales comparison method starts to break down when there are so few sales.

There are other methods that can be used (not typical appraisal methods), like time series.....but that's a little above my pay grade right now!!

 

[HF Mudd]

KJR2008, in answer to your question, I believe that the recent sale of the subject can provide good evidence of market value if the conditions surrounding the sale are truly representative of typical market activity.

I don't believe that I have ever used a prior sale of the subject as a gridded comparable sale - mainly because the conditions of that prior sale are usually NOT representative of typical market activity. The prior sale in most cases is an estate, short, interfamily, REO etc. sale which usually does not represent typical market activity. Even if the prior sale was representative, I would probably examine the transaction in the reconciliation process without placing it on the grid. I'm curious why the OP felt it necessary to grid this sale.

I'm more concerned with the OP's statement in his second post:

"only one other sale was supportable of the appraised value"

KJR - could you please expand on this statement? Are you saying that only one other comparable utilized supports your appraised value, are you saying that only one other sale in the market supports your appraised value, or are you saying something else entirely that I'm not understanding?