Time adjustment in the declining market

[Lee Lansford]

You would prefer they used matched pairs?
Sales & ReSales of the same property?
Some other statistical method?

If housing in the neighborhood are, first, not EXTREMELY homogeneous and, second, if analysis does not include careful consideration of REO/Short Sale activity vs. those that are not in those categories, percentage changes in average and median prices (from period to period) are not percentages that should be applied when appraising a particular property in the neighborhood.

The segment of the market that the subject occupies may have price movement that differs from that of the neighborhood as a whole. The appraiser should search for pending/contingent sales activity of competing housing as one indicator of movement in prices. Current offerings may have less reliability because such properties may never sell for a price any where near the asking price, but there is that little thing called The Principle of Substitution that has to be considered.

 

[Doug Meyer]

Your adjustments should also be as of the pending date and not closing date to accurately reflect the change in market (if any).

 

[timd354]

Just finished a report in a neighborhood I last visited in late October.
As of late October 2008, decrease in value 5/07-10/07 -vs- 5/08-10/08
was about 5-6%, and list-to-sell ratio about 95% IIRC.
NOW... 8/07-1/08 -vs- 8/08-1/09 periods show 29% decrease, and list-to-sell ratio now at 87%.

So... which is more correct, 29% annual decrease, using 6 month periods,
or should I look at the last 3 months only, where the % figure would be substantially higher?

I don't think there is a "right" answer here; use 6 months and smooth the data, or use 3 months
and have most radical change shown, which may or may not continue at same rate.
???? ????

There certainly are lot of different variables that need to be considered such seasonal variations in some markets, lack of sufficient sales data to be statistically meaningful over short periods in some markets, inclusion of REO and short sales in the data analyzed, etc. There is no "right" answer that applies to all situations. The best approach to take will change depending on the market and the available data for that market. Truthfully, sometimes the best and most credible approach will be obviously flawed due to the lack of a sufficiently large data pool for a particular submarket which neccesitates looking at a pool of data for the larger market and trying to apply the findings from that data to the submarket in which you are appraising. For instance I recently had an appraisal of a property where there were only 4 sales of similar properties in the past 12 months (and only 6 within the past 24 months) and 1 current listing. obviously, that is far too small of a data pool to get any meaningful trends so I had no choice but to calculate the rate of depreciation from a larger pool of data which I expanded to include houses that were not similar to the subject property but were still within the same market area.

This is where a competent appraiser can really shine. He needs to use his ability to craft the right solution to the particular market depending on the data available and explain what he did and why he did it. One thing I know for sure is that the typical method that I see among many appraisers in my area of just pulling the year over year change in median sales price for an entire zip code and applying that rate of change to make a time adjustment is not appropriate in most cases.

 

[Riick]

Lee.. I understand that the general market might not reflect what is going on with the specific property.
However, our favorite customers seem to insist that if market values are going down,
that should be reflected with a supported (supportable?) negative time adjustment to the comparables.
(( Funny how they did not insist on this on the way UP ))

So, here is "M" the appraiser, in a market where for the last 6 months there have been a grand total of 30 moderately comparable sales
-- E.G.: 2 sty colonials within 2 miles, and with reasonably similar age, lot size and GLA.
IF those sales sold at an even rate, the 5 sales per month is too low a number (statistically) to be a reliable basis for drawing
a % of appreciation / depreciation for the month.
In fact, the 30 sales in six months are hardly enough to draw a statistically valid conclusion.
Still, the customer insists on some number.
So, the way I see it, you either I use what you can prove, which the customer (with their doubtless high powered computers and AVMs)
can the accept or reject, or you modify the data you find, based on your personal interpretation of what is going on, perhaps
using pending sales & competetive listings, addiing a SWAG element to the already rickety logical structure.

Stats are wonderful, WHEN the sample size is big enough.
When sales are slow, and they are surely slow around here, stats are unreliable.
In addition, we should all remember that it was high powered statistical models that said the sub-prime loans were NOT risky.
Low tech models that work, even if less precise, are good enough for this increasingly crochety old guy.

 

[timd354]

Lee.. I understand that the general market might not reflect what is going on with the specific property.
However, our favorite customers seem to insist that if market values are going down,
that should be reflected with a supported (supportable?) negative time adjustment to the comparables.
(( Funny how they did not insist on this on the way UP ))
quote]

Not only did tmany lenders not insist on market condition a/k/a time adjustments when the market was on the way up, many lenders simply would not accept any market condition a/k/a time adjustments. They have all now changed their tune and I personally cannot wait to sometime in the future when I run appraise a property where postive market condition a/k/a time adjustments are supported. I am sure that, suddenly, they won't want to accept market condition a/k/a time adjustments again.

 

[Alan H]

I've got to tell you that the whole idea of adjustments for a declining market can get carried away very easily.

I just went through a weeklong series of emails, phone calls and explanations over an assignment I completed in a gated community, where another appraiser did a 2nd appraisal and applied a blanket 13% adjustment to ALL comp values, based upon 'MLS Statistics'.

It turns out that the 'statistics' came from the difference in average sales data for the entire township.

As it so happens the properties within this PUD have historically sold for 20 - 30% higher than any of the odd properties surrounding it, partly because the ones inside the PUD were generally of a higher quality and newer, and partly because the PUD has some nice facilities. Also as it happens, a review of the actual sales in the township indicated that for the past 6 months the vast majority (80%) of sales were from outside the PUD and were of significantly inferior properties than our Subject.
This ### absolutely ignored all of this and used his MLS statistics to include a 1004MC in his report.

Although I included a summary of the market conditions within the PUD and the Township and the Region and their relationships, I did NOT include a 1004MC (because you can't put all of that on a 1004MC) and so the client came to me with questions.

After I had the the client reread the 'Market Conditions' in my report, they came back and asked me to explain how the other appraiser got the figures he did and I ended up spending far too much time 'reviewing' another appraisers work.

I'm now afraid that once the 1004MC becomes 'law' that I'll be spending a lot of time arguing my case and 'reviewing' other appraisers' work.

Hi-tech, lo-tech, whatever, one STILL has to compare apples to apples, unless of course your appraising oranges, in which case one must determine if oranges have increased in value while apples have declined.m2:

 

[timd354]

After I had the the client reread the 'Market Conditions' in my report, they came back and asked me to explain how the other appraiser got the figures he did and I ended up spending far too much time 'reviewing' another appraisers work.
m2:

I hope that your response was "I have no idea, you would need to ask the other appraiser how he came up with his figures"

 

[Riick]

After I had the the client reread the 'Market Conditions' in my report, they came back and asked me to explain how the other appraiser got the figures he did and I ended up spending far too much time 'reviewing' another appraisers work.

Jazus they've got big ones!
Where did they find the nerve to ask YOU how he got his figures?
m2: m2:​

 

[Alan H]

Jazus they've got big ones!
Where did they find the nerve to ask YOU how he got his figures?

m2: m2:​

I actually felt sorry for the lender, they were looking at the 1004MC like it was devine revelation or something and couldn't understand how our data could differ so drastically (he had a 13% blanket adjustment, I said that the values in the PUD were STABLE compared to 1 yr ago and only down 5% since 3 yrs ago, so I had NO adjustment for time). The lender wasn't so much looking for me to review the other appraisal as to help her understand what she was looking at and for.

By the end she said that she had lost all respect for the 1004MC (she didn't actually use those words).

I took the opportunity to explain to her that as of May 1, 2009 we will not be able to have these conversations any longer for any Fannie loans and now she is against HVCC. I told her to spread the word and tell everyone she knows what is comming. Who knows maybe it's not too late - and I'll try almost anything to stop that fiasco from happening.
:new_2gunsfiring_v1:

 

[Tucson Appraisals]

It is hard for me to see a licensed appraiser admitting to submitting reports without time adjustments when it was known before hand that it was a declining market. In my opinion that diminishes your credibility and you should not be accepting assignments until you have learned how to make time adjustments. No wonder your client is asking why you didn't make time adjustments....

As to how to make an adjustment, the decline is different from neighborhood to neighborhood, so why would anyone use statistics from an MLS? My guess is because it is quicker than figuring it out everytime...still better than not making any adjustment at all.

Here is some reasonable approaches to the problem that has worked for me:
If it is not in a rural area, I would take a sample of possible comparables (no less than 30 sales) within the last year and graph it in excell. See what the average decline is per month over the past year, then apply that decline percentage. Watch out to make sure..overall that the amount of decline per month is steady, if it increased in the last couple of months, then try to take that into account. If you are looking for the most accurate...from my experience this has been the best. As a secondary approach if you do not have a big enough sample is to simply compare newer sales to older sales and see what the reasonable adjustment would be. If you have conflicting information and can't make sense of it by doing it that way, then of course you could use the MLS statistics.

Again, I would not be accepting assignments until you figure out how to make these adjustments.