Would You Guys Buy This Arguement?

[Metamorphic]

I'm working on a very high end rural property. There's not a lot of recent comps....yadda yadda yadda, you know the story. I've found a great comp, across the street from the subject, as similar as two high end customs can be. I want to use it, but its a year old sale. I'm planning on using it as a 4th comp, but I think its necessary to present a basis for depreciating, appreciating, or even doing nothing with the comp on the basis of its date of sale.

Most of the methods I've been using for tracking market trends haven't worked. The area is very non-conforming WRT to age (new to turn of the century), condition, lot size (0.15 acre to 10), size (I pulled a batch of sales and got an 800 sf and an 8000 sf) etc etc. If you filter the data to control these variables pretty soon you dont have a big enough data set to do anything with.

So I got to looking at the data and found I had 19 properties that had had been sold twice in with in a 4 year period. On this graph I marked their sales prices and dates. I fitted a red line to properties that decreased in value, a green line to properties that increased in value. As you can see nothing below $800k has depreciated.

On this basis of this data do you think I'm justified in saying that there's no basis for making an upward or downward adjustment to the year old comp. FWIW, the comp in question is actually the highest value sale on the graph.

 

[Louis Pompeo]

You're kidding..........right?

 

[jay trotta]

ahhhhhh the "scientific" part of the business........ROFL

and yer in Calif. right - anywhere near Timothy Leary's house ?

 

[c w d]

On this basis of this data do you think I'm justified in saying that there's no basis for making an upward or downward adjustment to the year old comp. FWIW, the comp in question is actually the highest value sale on the graph.

Excuse my peers!

Whats the average change in value over the 4 year period? If it's a net zero change then I'd say your data is not telling you anything. If it's a positive change then I'd say your data is telling you what the trend was for the past four years. Remember in the past four years the market experienced its largest price gains which would skew your results. Even if its giving you a negative average change its still going to be skewed by the gains in market price early in your data set.

In my experience rural areas are somewhat insulated from big changes in the overall market. But they are not immune. Compounded with the lack of sales determining a trend is difficult if not impossible so don't feel like you're missing something when it may not be there.

You can derive "market condition" or "time" adjustments from viewing the price sales and compare those to current prices of comparable active listings. Use a decent sized data set and you can determine a ratio (percentage) for a "market condition adjustment" or "time" adjustment. I personally do not utilize this method but, it seems acceptable to a large enough number of our peers here on the forum that you'd get the benefit of the doubt from a reviewer.

 

[Lloyd Bonafide]

Were the higher-priced comps improved or remodeled between their sale dates? Also, it looks like the two higher sales sold around 1-07 and 6-07 - and probably have nothing to do with the current market.

I don't know where in the region the property is located, but if you don't make a negative time adjustment, the typical reviewer will probably toss the dated sale out of consideration.

 

[Lawrence R.]

This one is tough.

I tend to agree with CWD, your analysis is like trying to pinpoint how much the temperature changed in the last 20 minutes of the ice age...

In the grand scheme of things, the recent turn in the market is not statistically significant against the past 4 years of rising prices(well, slightly, if you have a ton of data--which in this case you don't)

I still think you would be better served to look at listing information maybe quarter over quarter for the past year and see what pricing trends have done as well. You may see more clearly that sellers expectations are dropping at a rapid pace, even if the homes have yet to sell.

If your data is telling you something 180 degrees out of phase with what your gut, and your peers, and the national data is telling you, there is probably something wrong with the data. In this case, I think that the current snapshot in time is too short to reveal statistically significant trends by observing sales alone.

Run some tests on listings. It may clear up the picture.

 

[incognito]

This is a $250 job AMC job, right? If you are into the analysis, I would also look for past sale prices of current comparable listings. You may find some trend there.
If I were analyzing your chart, I would interpret a general downtrend in the past 2 years. The homes that first sold more than 2 years ago may have net positive increases in sale price, but are likely skewed up by large increases in 05. If you disregard all but the resales that have taken place since 2006, I see a general down trend. That is the salient data, the way I see it. I see
-2% to as much as -20% per year indicated by those sales. Of course, the data is raw and unadjusted for additions, condition, etc. None the less, I see support for a negative time adjustment. YOU get to choose how much!


As for FNMA, this statement might help when some underwriter does not think sales over 1 year can be used:

FNMA guidelines DO NOT preclude the use of comparables over 1 year old. Per the FNMA HANDBOOK FOR APPRAISERS:
"Generally, comparable sales should have been settled or closed within the last 12 months. However, you may use older comparable sales if you believe that they are appropriate for the situation, and the selected comparable sales are the best indicators of value for the subject property."

 

[Metamorphic]

Were the higher-priced comps improved or remodeled between their sale dates? Also, it looks like the two higher sales sold around 1-07 and 6-07 - and probably have nothing to do with the current market.

I don't know where in the region the property is located, but if you don't make a negative time adjustment, the typical reviewer will probably toss the dated sale out of consideration.

Checking on the remodel status is my task for this morning, I was too burnt to do it last night. However, most of the high end properties on that chart are newer which doesn't suggest the likely hood of a remodel with a big change in value, and its the high end numbers that are the one's I'm looking too most for value.

Yeah, I realize I need to address the age of that sale to make it work, but I'm having trouble finding a basis for doing anything. Amongst the newer high end properties there's just not enough sales detect any trend. The magnitude of the market related depreciation you might expect to see is completely confounded by things like 180 degree panoramic views of the sierras, half a dozen excess acres of land that may or may not be usable, and all sorts of subtle locational preferences (access, proximity to town, big pines vs mixed oaks etc). When I start looking at the properties individually I'm left wondering if that one recent sale that looks a little low is showing me depreciation or or perhaps its a slog to get to from the highway, or maybe the "super" view the agent promised in the MLS is just "ok", or maybe its a big fancy home built amongst a bunch of skanky cabins, or maybe its 7 acres is all located on the side of a hill to steep to climb.

I wouldent mess with all of this is this comp wasn't so good otherwise. Its the only one in the same development which is a little bit special in the area. I've got another comp that's in a similar, but nice development that's at least a couple hundred thousand over where I'm looking. I've got one that's not in any development that's a couple hundred below where I'm going, and my third is right at the price I'm heading to, but was sold partially completed (I've got to get a handle on that one today). And that's it. Anything else that's potentially comparable is old than a year, or in a totally different price range.

 

[Metamorphic]

Excuse my peers!

I did that long ago. I've been on various net forums for a long time and what you find is that on every one there's folks who participate in order to sometimes teach and sometimes learn, and there's some people who just want prove how much they know, but don't have time to teach or participate meaningfully so they just kibitz.

Whats the average change in value over the 4 year period? If it's a net zero change then I'd say your data is not telling you anything. If it's a positive change then I'd say your data is telling you what the trend was for the past four years. Remember in the past four years the market experienced its largest price gains which would skew your results. Even if its giving you a negative average change its still going to be skewed by the gains in market price early in your data set.

Yeah, I realize that those nice straight lines can be covering up a "hump" in the value of the properties. Basically what I was experimenting with here is a paired sale analysis, but instead of doing 2 houses where everything is the same except for the pool, I'm trying to hold all the improvements the same and let the time be the variable. My hope was when I started graphing these is that I would see a trend where stuff that had its 2nd close in the last year would be red lines, and stuff that closed for the second time 2 years ago would show a green. That would definitively decreasing trend. On the other hand if it was all green, then you could, at the very least say that "while the market in the state is in decline, in this market the trend is not sufficiently negative to be documented on a basis as long as 4 years."

In my experience rural areas are somewhat insulated from big changes in the overall market. But they are not immune.

That's the feeling I'm getting here. The market is insulated. The high end houses may be insulated to the point of effective immunity.



You can derive "market condition" or "time" adjustments from viewing the price sales and compare those to current prices of comparable active listings. Use a decent sized data set and you can determine a ratio (percentage) for a "market condition adjustment" or "time" adjustment. I personally do not utilize this method but, it seems acceptable to a large enough number of our peers here on the forum that you'd get the benefit of the doubt from a reviewer. Roscoe also suggested looking at listings. I guess I'll be doing that today too.

1234567890

 

[Metamorphic]

I tend to agree with CWD, your analysis is like trying to pinpoint how much the temperature changed in the last 20 minutes of the ice age...

In the grand scheme of things, the recent turn in the market is not statistically significant against the past 4 years of rising prices(well, slightly, if you have a ton of data--which in this case you don't)

That's not entirely true around here. In nearby markets where there's more sales and a little more conformity, I can pretty easily document a 1/3 reduction in value starting the last quarter of 05. (I'm starting to get an occasional data set that suggests a flattening is in progress in the more desirable middle-class'ish areas BTW.) So if the subject market has gone through the same 1/3rd'ish roller coaster as the rest of the market, I'm pretty sure I'd be able to see it, even with the small data set. In the 4 year term I graphed, would have 1 year of up and 3 years of down.

I still think you would be better served to look at listing information maybe quarter over quarter for the past year and see what pricing trends have done as well. You may see more clearly that sellers expectations are dropping at a rapid pace, even if the homes have yet to sell.

I'll try that today. When you do that do you present it graphically, or in a table or do you just discuss it and keep back up in the file?

1234567890