1004MC Addendum Error

[VegasWayne]

Resguy, SFREP is calculating the median sale price as % of list price by using the numbers in the table instead of the correct way, finding the ratio for each sale and then finding the median. When Wintotal released the first version of the 1004MC form it would automatically make the calculation the same way. The 2nd version of the form removed this feature because they found out it was being done incorrectly.

You may want to check for duplicates in the listings, the same address with two or more listings. This is less of a problem since they changed the directions to only include the active listings on the last day of the period, but could be the cause of the different counts.

 

[residentialguy]

Yeah, there are duplicate addresses, but they are for different listings (cancelled, expired, etc)

Here is a portion of the addresses

 

[George Hatch]

I'm curious why there's any variation at all between the Matrix worksheet and mine. Maybe I've got an error in there somewhere. I wonder if we're handling pendings differently? Or maybe we're handling the tallying protocol relative to the exact date differently.

This is fairly straightforward math - that means there shouldn't any deviation unless we're using different "rules". Maybe I've been using the wrong rules. Kind of a bummer. I know that when me and my beta tester checked my worksheet by hand we didn't see any errors.

Something isn't right. I'll bet it's the pendings.

 

[David Mescon]

Appraisers should not lose sleep over the accuracy of MLS data when completing the 1004mc form. We are not responsible for this data base and have no control what is posted or not. I add comments on the form itself that the MLS data is often incorrect or that certain data is often not included and that I am not liable for the effect this has on my results and reporting.

Maybe I'm just old fashioned, but if my signature is on the report, the accuracy of each element matters to me. I'm slow compared with many other appraisers, but I do things like hand vet each and every listing and I toss the obviously incorrect stuff.

 

[Delta85]

Maybe I'm just old fashioned, but if my signature is on the report, the accuracy of each element matters to me. I'm slow compared with many other appraisers, but I do things like hand vet each and every listing and I toss the obviously incorrect stuff.

So in completing a 1004mc you research and verify all the data that comes up with the realtors and through assessors records? All the sales and listings?

 

[VegasWayne]

I'm curious why there's any variation at all between the Matrix worksheet and mine. Maybe I've got an error in there somewhere. I wonder if we're handling pendings differently? Or maybe we're handling the tallying protocol relative to the exact date differently.

This is fairly straightforward math - that means there shouldn't any deviation unless we're using different "rules". Maybe I've been using the wrong rules. Kind of a bummer. I know that when me and my beta tester checked my worksheet by hand we didn't see any errors.

Something isn't right. I'll bet it's the pendings.

My guess would be that there is a difference in how many days back it is to the the prior periods. The form uses months and most applications I've seen convert this to a standard number of days. Since February has 28 days there would be a difference between three months and 91 or 92 days that most people use. The trends would be just as valid with either method.

 

[David Mescon]

So in completing a 1004mc you research and verify all the data that comes up with the realtors and through assessors records? All the sales and listings?

No, I don't go so far as to verify each listing. I do, however, read each one, pretty thoroughly. I often find reasons to disqualify a particular listing after doing so. It gets rid of a lot of the way out there stuff, like, for example, mixed use properties that are found under the single family category of listings.

 

[Elliott]

Here is your worksheet



Here is Matrix



Here is SFREP's

Now, would someone explain how their narrative analysis of the 3 different MCs would be different because of the data?

 

[George Hatch]

I use T-91 days for the 1st period, T-183 days for the second period and T-365 days for the 3rd period. That right there could account for the differences.

Another possibility for the variance between mine and the Matrix version may involve the protocols for the cutoff dates. Depending on the composition of a dataset, it may make a difference which day the worksheet uses as the cutoff date, meaning one day there may be 28 actives and the next day there might be 33 actives. If we're handling those dates differently, that one day difference could explain the 51/13 count for closed sales between those two periods in my workshheet vs the Matrix tally of 54/12 - maybe the one sale closed right on the borderline of that period and Matrix picked up another couple sales from the prior period that mine excluded. Or vice versa.

The addition or subtraction of that one sale between those two periods could have a corresponding effect on their respective medians, too.

I know in my worksheet I incorporated the "<=" references to the specific dates so as to count a sale up to and including that date instead of cutting it off on the day prior to that cutoff. Same with the listings - the number as of that one date.

ResGuy, if you ran a couple more iterations on the Matrix version by moving the date around by one day forward or one day earlier, I wonder if those results might match up better?

 

[DTB]

All this consternation over a bogus 1004MC that nobody give's a rat's azz about. Except fellow appraisers. The whole real estate world is laughing at us.