Actual Age/effective Age Vs. Condition

[Fred]

Makes no sense to me

Verne, I have another one. This is from a report done by a regional firm brought in because of their "expertise" in a specific situation.

The report says: "Subject's effective age is 40 since it competes with properties that are predmonantly 40 years old."

After considering that, thik about this. The report estimates 100 year life for the replacement improvements and 40/100 just happens to be the ratio that gets the cost approach withon a few percent of income capitalization. The real zinger for me is that this is a historical district where subject is a hodge-podge of add-ons that might average 100 and competes with other properies that are 50-150. I don't see any 40-year old competitors. That's why I say, "effective age" has a limitless number of meanings.

 

[Austin]

Just to answer the argument about linear and curved age trend lines, I took the same age vs. price graph and changed from linear to exponential. Note that the R square increased by only 6% from 19% to 25%. In other words the argument about linear and curved lines in the final result is not significant even in this entire range of properties.

 

[Austin]

There are a lot of other factors regarding the graphs I posted that you may not have noticed.
1. By changing age from linear to exponential we explained an additional 6% of price influence from age. If I did the same for GLA it would be even more because quality of construction is a huge factor. Already we have explained 98% of price variance in this entire market with size and ages alone. Just these two adjustments put us right in the middle of the noise zone and all we need to do is reconcile.
2. There is one other adjustment I can make to refine the results even more. How much better can you get from explaining 98% across the spectrum? You have to filter the data to get down to the molecular level of dealing with a few comparable sales.
3. Notice how much the high value sales skew the data. If we removed these sales and treated them separately the text statistics would blow the roof off.

What does this prove? Arguments about effective age, how much to adjust for a carport, how much a larger lot contributes, etc., are just academic exercises in futility. The sequence of adjustments come from the market and number one on the list is generally GLA. Not always thought which is why we must test every data set before doing anything to determine what we are dealing with.

 

[Mike Boyd]

AGE and CONDTION deserve separate analysis and are, therefore, subject to separate adjustments.

Condition is cosmetic, age is structural as well as operational effectiveness.

Attempting to designate EFFECTIVE age of comparable sales would require an in depth analysis of each property and worthy of a narrative report and the resulting fee of 10 times the standard fee for a SFR.

I go by the comments made in the MLS printout regarding what is new. You can bet if it has a brand new heating/cooling system, new kitchen cabinetes and tops, new roof, etc. the listing agent will make reference to those items in the printout.
If it has new paint and floor coverings and that is the extent of the comments, you can bet nothing significant has otherwise been done.

I never use effective age in the market comparison grid. I might type in actual age/updated for the subject as well as comps. If I have not concluded that a comp has been updated I will make a + adjustment based on market reaction in that market area to age differences and the quality range.

If we let them, those who make the deciscions of what should be required in an appraisal report will become more and more like congress, thinking that new laws have to be made regularly to demonstrate and support their employment status.

There may be situations when extensive narrative reporting is necessary. In those cases you have a more complex property to evaluate and your fee should be increased to reflect that complexity.

 

[Fred]

Austin,
I didn’t say anything about “non linear.” I just didn’t see size and age on the same graph as advertised, which would be a three-dimensional graph. Right? The ANOVA table works for me.

I don’t have any problem with your analysis. However, I have just never seen real market data where some single number, like $85/SF-$90/SF explains so much that no one would ever need an appraiser. Are those houses or fungible goods like bushels of wheat?

You have been riding the square-footage explains everything' horse, since I came on the forum in 2002. It is interesting to read you writing that “view” and this other stuff don't make any difference in your study. I don’t know your area and maybe there is nothing to look at there that excites the desire that leads to perceived utility, and scarcity, etc. For example, where I work, in the last 12 months, the lowest half-acre home site price is about $45,000 (maybe $2.00/SF) and the highest half-acre home-site price is $1 200,000 (about $54/SF).

Does anyone besides Austin work in a market where price per SF is all anyone has to know?

 

[Austin]

Steven:
Five will get you ten I can go into any market in the country and find the same thing. There is nothing out of the ordinary about this market area with one exception-me. This data is right out of the MLS book and stretches over a two year period. The method as just shown works across the entire range of prices and even better on filtered comparable data. I am surprised that the result is this good over that range because quality of construction is the next most significant value influencing variable. If you look at those data clusters you can pick out the upscale subdivisions in clusters. Our local tax assessors have a 36 point quality of construction rating that is reported on the tax data card. I have a graphical study around here somewhere showing how it correlates with M & S.

Don't try this one at home folks because this data is about three years old. The attached graph is an analysis of quality of construction categories from the residential cost manual. I correlated it with the local tax assessors rating system. So if they have a B+4 rating I know exactly where to look in M & S. I can't find it right now but I will post it if I can find it.

 

[Tim Hicks (Texas)]

Interesting. I made up a very unscientific graph regarding the interest of linear regression graphs on the forum. The original topic had 36 readers and over a few posts it quickly dwindled down to two readers (I wonder which two?). Then after general apathy to no readers.

Here is my graph.












This post is purely in jest and I only wanted to see if I had the ability to make a fancy graph like Austin. I think that regression analysis has many merits, but unfortunately can be manipulated like any statistic. Hopefully, Austin will realize this is just internet teasing by an inferior counterpart and will not take it personally. I appreciate his efforts to educate us less dedicated individuals.

 

[Peter LeQuire]

The 3-D effect is sweet.

 

[JSmith43]

Tim,

Is that an autocad rendition of a San Fransisco sidewalk?

 

[Fred]

Austin,

I can go into any market in the country and find the same thing.

You can DO the same thing and so can I, but would we really FIND the same thing. Wasn't there a movie called "Flatliners?" Are you looking at a market or the vital signs of a corpse?

If you are right and that lack of variability is the norm, then AVM is a lot more viable than I ever thought it was. I don’t get to work with that kind of line-up-like-ducks, self-reconciling, kiddie-data Those are cookie-cutter appraisals you have there: 600 sales and looks like more than half are within the (I almost hesitate to use the word) "range" of 50k to 150k.