Killing the 1004MC

[PhiloFarnsworth]

Is there any organized movement within the appraisal community working to send the 1004MC back to the depths of Hell from whence it came? Are there any attempts to make the case that the entire concept of the from and the instructions are such as to be entirely unclear and misleading?
It seems to me that, if one is to refer to a table of numbers based on a virtually uncontrolled sample set, with laughably small sample sizes as a "statistical analysis" is, on its face, fraudulent.
Maybe its just the character of my area, but in order to obtain any sample size that could have a chance of being statistically analyzed, the criteria for inclusion of comparables must be so broad as to make the individual differences among the comparables swamp any significant change in market values. Perhaps there are some places where huge subdivisions with true homogenous character may provide some meaningful data, but I have yet to find one in my work in SE PA.
The local MLS I use provides the ability to pull out the 1004MC grid for any set of criteria one might set up. An interesting exercise is to run multiple analyses, each time making small changes to the criteria. Each criteria set might arguably be a correct one when trying to get a minimal data set. Each run returns wildly different and contradictory results. Even using more wide ranging and generalized internet sources such as Zillow, etc shows the silliness of this reliance on "scientific" statistics. Zillow breaks down each county by municipality and gives, among other things, adjusted year to year changes for each. Looking at this list, am I really to believe that two adjacent townships that are in reality, identical in general market conditions may show a 14% drop in one and a 5% increase in the other. Reporting this data as anything other than the results of a random number generator is undeniably misleading.
I am an appraiser. I am not, nor do I claim to be, a statistician (in fact, without spell check, I can't even spell it). The reliable and applicable information on local market conditions that I should impart in my reports are limited to: 1)General, region-wide data that characterizes the market conditions in the wider area, county or group of counties. Published MLS county data is a start, and may be expounded upon with the rare piece of data one might find to back up other contentions about particular sectors of my market. Again, any real reliable data I can find is inevitably region-wide. 2) Reporting of those cases where a group of data shows a local deviation from what is typical for the region. These cases are not typical; most neighborhoods do not have enough data to establish any significant difference from the region as a whole. 3) Report any relevant issues that may have an effect on market trends.
All this is typically covered in the narrative on page 1 of the 1004. I have no problem with the addition of the MC form as a separate page; the comments section below the grid on the MC form add some real information also. But I think the grid should be made optional, to be used only when it clarifies the report and does not mislead the reader into thinking they are looking at a meaningful analysis.

 

[Lee Lansford]

...But I think the grid should be made optional, to be used only when it clarifies the report and does not mislead the reader into thinking they are looking at a meaningful analysis.

Other than the reporting of the data, I have from the get-go not been X'ing many "overall trend" boxes as the data are too few to distinguish a "trend".

Fannie has recognized (at this moment, I don't have the pronouncement in front) that such situations would exist and alerted appraisers that in such instances the appraiser is obligated to supplement the 1004MC with additional market data and analysis.

 

[JTRock]

Anyone that has taken a statistics course knows that the form is flawed.

 

[Pilgrum]

"The purpose of this addendum is to provide the lender/client with a clear and accurate understanding of the market trends and conditions prevalent in the subject neighborhood. This is a required addendum for all appraisal reports with an effective date on or after April 1, 2009." ​

There is the main problem with the 1004 MC. The data is for the NEIGHBORHOOD. And COMPARABLE sales at that. It has nothing to do with the MARKET AREA and is, therefore, flawed from the beginning. Further the information is supposed to match that recorded at the top of page two of the URAR. Again leaving the reader with an unreliable view of the subject's market.

"Instructions[FONT=Arial,Arial][FONT=Arial,Arial]: The appraiser must use the information required on this form as the basis for his/her conclusions, and must provide support for those conclusions, regarding housing trends and overall market conditions as reported in the Neighborhood section of the appraisal report form. The appraiser must fill in all the information to the extent it is available and reliable and must provide analysis as indicated below. If any required data is unavailable or is considered unreliable, the appraiser must provide an explanation. It is recognized that not all data sources will be able to provide data for the shaded areas below; if it is available, however, the appraiser must include the data in the analysis. If data sources provide the required information as an average instead of the median, the appraiser should report the available figure and identify it as an average. Sales and listings must be properties that compete with the subject property, determined by applying the criteria that would be."[/FONT][/FONT]​

How often have you seen anyone explain that the data is unreliable? What do you do if the subject's neighborhood has only 4 or 5 comparable sales over the preceding year?
Most of the time in my area data is so limited as to render the 1004 MC meaningless.

 

[Ca Ar Independent]

How often have you seen anyone explain that the data is unreliable? What do you do if the subject's neighborhood has only 4 or 5 comparable sales over the preceding year?
Most of the time in my area data is so limited as to render the 1004 MC meaningless.

Same in my area. I continue to run the 1004MC report from the comps that I select and note at the top of page 2. The form is a joke and yet another waste of time. While in the comments I say that the 1004MC is designed to indicate market trends only" (I occasionally to get UW questions asking why my value differed from the median values of the 1004MC!) and that due to the limited amount of truly comparable data, the results are not always a reliable guide of market trends. If I based my opinion of trend on the 1004mc only, I frequently get results that indicate an increasing market...which we definitely do NOT have. I do use it to calculate sale/list ratios for either 3, 6 or 12 months to apply to the active/pendings and for that it has been convenient.

 

[Lee Lansford]

"...

How often have you seen anyone explain that the data is unreliable? What do you do if the subject's neighborhood has only 4 or 5 comparable sales over the preceding year?
Most of the time in my area data is so limited as to render the 1004 MC meaningless.​

Not "meaningless" in that if comparable data in the neighoborhood are sparse, such provides some support to why an appraiser either has to go outside of the neighborhood for comparable data and/or include less than reasonable data for analysis.

If an appraiser--in situations where data populating the 1004MC are sparse--is not explaining that the lack of data does not support a credible opinion of an "overall trend", the appraiser is not doing his/her job (which does not surprise me all that much).​

 

[Pilgrum]

Not "meaningless" in that if comparable data in the neighoborhood are sparse, such provides some support to why an appraiser either has to go outside of the neighborhood for comparable data and/or include less than reasonable data for analysis.

If an appraiser--in situations where data populating the 1004MC are sparse--is not explaining that the lack of data does not support a credible opinion of an "overall trend", the appraiser is not doing his/her job (which does not surprise me all that much).[/quote]

I can certainly agree with that.
I alway look at neighborhood, market area and community data. I also compare sales prices in 6 month over 6 month, year over year and, at times quarterly over quarterly time periods. Also I compare COMPARABLE sales prices AND the market as a whole.
Sometimes there are some surprises.

 

[George Hatch]

One of the biggest shortcomings of the form is it doesn't bother to ask for the search parameters used so a reader can have a shot at recreating the search. It doesn't mention the limitations of using almost-unqualified data in the analysis or comment on the general limitation of using really small datasets. It doesn't even ask what tools the appraiser used, if any, to perform their analysis.

I'll put it this way, when commercial appraisers use Discounted Cash Flow models or put together an income/expense analysis we have to lay out our assumptions in those models so that the reader can understand how we got there. If we use a tool we have to be prepared to explain how it works.

Whether they keep the analysis or not there's a lot of room for improvement in the form and their instructions.

 

[Don Clark]

One of the biggest shortcomings of the form is it doesn't bother to ask for the search parameters used so a reader can have a shot at recreating the search. It doesn't mention the limitations of using almost-unqualified data in the analysis or comment on the general limitation of using really small datasets. It doesn't even ask what tools the appraiser used, if any, to perform their analysis.

I'll put it this way, when commercial appraisers use Discounted Cash Flow models or put together an income/expense analysis we have to lay out our assumptions in those models so that the reader can understand how we got there. If we use a tool we have to be prepared to explain how it works.

Whether they keep the analysis or not there's a lot of room for improvement in the form and their instructions.

I agree. First thing....do away with the "Neighborhood" analysis and go to a market analysis. Even in an area with 1,000's of neighborhoodsw I often cannot find enough neighborhood data to construct a meaningful 1004MC.

 

[PhiloFarnsworth]

I have to wonder how many of those who are compiling the data described, even when the data is plentiful, have a good grasp of what they are looking at. Large amounts of data will invariably generate statistics to admire and speculate on. But how many appraisers have a grasp of statistical methods good enough to judge their results. With the amount of variability from numerous uncontrolled parameters in typical samples can make it necessary to use hundreds of data points to arrive at statistically significant results. Time to junk this thing.