This is a pretty niffty detailed chart for finding time adjustments, we are done.

[Mike Ault]

I am willingly going way out on a limb here to say that the model of using out of area appraisers working under unrealistic turn times does not result in better, more accurate or credible appraisals, regardless of the enhancements.

Maybe it's just me.

I appreciate your sentiment, however out here in the sticks, I consider anywhere within a 60 mile radius of my office, "my area". I have MLS access for the entire area, know it well, have been appraising in it for over 30 years, and frequently I am one of the few appraisers available for relocation assignments in those small towns. And in between my office & most of my subjects, there is nothing but cattle and antelope roaming amongst mesquite bushes. "Distance" is not a good metric to gauge competence in sparsely populated areas. There is another large population center 120 miles south of me, and I leave it to the appraisers living there to appraise anything within 60 miles of them.

 

[Terrel L. Shields]

Back in the days I lived 50 miles outside Dallas

In 1975, 50 miles outside of Dallas is metro now. I remember when the I35 Y was miles outside of town. I've been backed up north of there trying to get downtown and had to go at 20 mph for miles.

 

[alebrewer]

And that begs the question. What kind of trendline do we run? Linear? Logarithmic? running average? Or, maybe a 2nd order derivative is just the ticket?

Not sure, but you better have a pretty chart to let them know what you're doing - and why! I was thinking log trend with exponential smoothing.

 

[Mike Ault]

Not sure, but you better have a pretty chart to let them know what you're doing - and why! I was thinking log trend with exponential smoothing.

I'm waiting to see what you do. I'm pretty sure that alamode will come out with a "Titan Analytics solution" soon as well. Too much money to be made for them to just sit this one out.

 

[alebrewer]

I'm waiting to see what you do. I'm pretty sure that alamode will come out with a "Titan Analytics solution" soon as well. Too much money to be made for them to just sit this one out.

agreed. I use Spark/Synapse and I've no doubt they're building one of these as we speak. What's gonna be real fun is trying to make sense of month by month deltas when you have a total of 20 observations for the entire year...

 

[DWiley]

agreed. I use Spark/Synapse and I've no doubt they're building one of these as we speak. What's gonna be real fun is trying to make sense of month by month deltas when you have a total of 20 observations for the entire year...

That would be an exercise in futility with that little data. Perhaps that is why the is no requirement for monthly analysis - despite continued assertions to the contrary.

 

[alebrewer]

That would be an exercise in futility with that little data. Perhaps that is why the is no requirement for monthly analysis - despite continued assertions to the contrary.

I've never asserted that the GSE's require(d) it. What I have asserted is that, with publication of that cool graph - and subsequent illustration(s) - the GSE's created the perception that this kind of analysis is now required. Wait until the providers deploy their 'solutions'. Got a hundred spot they will look much like the graph published by the GSE's. Want some of that action?

 

[Sadie]

No official word from Fannie that this is a requirement. Imo a neighborhood trend line is what they want, but then they post an example based on this, which is insane -every sale tracking at a different amount. it becomes about price smoothing, not how a trend is truly influencing a market.

Appraisal volume is down, but the managers at the entities want to keep their jobs, so they come up with ever more demands about precision - for the smaller and smaller amount of appraisals that will be performed.

IMO, the Fannie examples are best viewed with healthy skepticism. They don't seem to be based on reality, more like wishful thinking.

 

[Sadie]

Rather than adjusting each sale at a different rate, which seems more about the sensitivity of an individual price, we can develop trend lines that show when the market changes ( if it changes over the course of a year), Such as prices were stable for the first six months then either declined or increased. An 8 month old sale should not be adjusted for its first six months on the market, but adjusted after the six-month mark (for example)

Each market is different and some prices are dependent on the season. What the guidance boils down to is support your adjustments or lack of adjustments.

 

[Sadie]

And what happens when their broad brush market analysis shows prices are decreasing in the greater Baltimore area but the subject's market segment is stable or increasing (or vice versa). Are we gonna get hit if we do it correctly? And if we do it correctly will we be hit with discrimination if FNMA broad brush statistics show different than the subject's market?

Doesn't your MLS system have stats that you can customize to fit your subject? Like the new home market, or age restircted areas? Sometimes those markets hum along in their own little world,