One Unit Housing Trends - How long a period?

[Mike Garrett RAA]

Yes, my MLS does not include duplicates in the stat function.

 

[moh malekpour]

I usually do quarterly or seasonal comparisons like spring with spring and summer with summer or first 3 months of one year with first 3 months of another year but I don't limit the search to one season or one quarter of a year. I do at least 4 seasons or 4 quaters of one year to another year. I do it this way becuse I want to have enough data and also eliminate the seasonal effects.

 

[Richard Carlsen]

2005 avg 214K

2006 avg 194K

2007 (so far) avg 178K

All same zip code. All sales reported in MLS. I would say “declining” is most probable.

Good Morning Serena

One of the things that I remember from some of my appraisal classes is that "average" is something an appraiser should eschew except as a last resort.

You "data" could be the results of changes in buying in market sectors. For instance, 350 houses sold in 2004 for a $214K each and 350 houses could have sold in 2007 for an average of $178k each. But what if the market preference was for less expensive houses due to hard economic times and not just a reflection of the market rubber stamping the selections from 3 years before. What if say in the $90K to $140K price range, the number of sales increased by 25% while there were enough high end ($500K and up) sales to raise the average?

Use of gross MLS data is one of the reasons that I raised the question.

To my way of thinking, one has to examine the trends of the market (up or down in specific price range) as well as the gross numbers of sales with the best examples being re-sales of the same property based on the definition of Market Value. In this respect, I agree with and tend to do the same thing as Mike. (We fishermen all seem to think alike)

Generally in our market, the high end houses, both waterfront and off-water have suffered while the average sale price has declined due to the drop off of the higher end sales and the fairly steady demand in the lower end housing sector. This has resulted in a lower average selling price per unit that the media loves to latch on to as they publish their doom and gloom rags.

I guess what I'm saying is that just because the "average sales price" goes down, does not mean that the market is declining. IMNSHO, market trend comments are based on gross number of sales, sales trends within a market price sector and analysis of sales of houses of specifically similar size/style/location and utility with resales providing the most solid data for making comments on market trends.
The problem that those of us who service low density markets is, how many sales do you need in order to reliably make comments on trends. As an example, using the local MLS and using the following criteria; SFR sales for the past 12 months, frame construction, under 5 acre site, 1000 to 1200SF with 1.5 to 4 car garage, the search turned up exactly 5 sales ranging in price from $41,500 to $115,000. Do I hang my hat on only 5 sales for a whole year. If the year before was the same number, is that enough when the low end of the range is only 36% of the selling price of the upper end value?

 

[Michigander]

belaboring a point

I can't insert grids or any pics in here for some reason, but let me add to this.

The gross information on average sale price is for a zip code that includes copious new construction and plenty of data and shows a significant downward trend since 2005 (as expected in Michigan for the most part). Specifics include 2 sets of sales from 2005 until present of the exact same model house with little variation in amenities.

Model 1, 2005 had 4 sales with average sale price of $211,450
Model 1, 2006 had 6 sales with average sale price of $198,201
Model 1, 2007 so far has 4 sales with average sale price of $172,375

Model 2, 2005 had 6 sales with average sale price of $215,433
Model 2, 2006 had 5 sales with average sale price of $195,326
Model 2, 2007 so far has 5 sales with average sale price of $183,648

Tells you pretty much the same thing as the gross zip code search with average sale prices but this one gets down to specific nitty gritty. Either of these work in trending what is happening in a market. I would contend it is up to the appraiser who is trying to support their assertion (one direction or another) what is happening with the market. My contention is that the gross data I provided earlier is pretty similar to what I have just provided with specifics, model for model. What I am looking at is support, not proof to the minute detail of exact amount of decline. This is evidence enough to say this is a declining market.

 

[Mike Garrett RAA]

Based on your data...I would concur.

 

[Artemis Fowl]

Further belaboring a point

The original question was regarding the 1004 form and not the market in general. It has been my interpretation of FNMA guidelines that there is a distinction. If my subject is an $80k 800sqft bungalow in a neighborhood with prices ranging from $60k to $130k, do I care what happens with properties in the $200k-$250k price segment within the same zip code? If 300 $400k new contruction properties hit the market today is that going to cause an oversupply of properties similar to my subject? Isn't there a distinct market defined by the typical buyer? Am I way off base here?

XI, 402: Market Data Research (08/24/03)
The appraiser is responsible for adequately researching market data from all reasonably available and appropriate sources of information for the location and property type being appraised (including public records transfer information and, if appropriate, data from local real estate brokers who are not active in the local multiple listing service)—even if this results in the appraiser spending more time and incurring additional expenses in performing appraisal assignments in certain geographic locations or for particular property types. If the appraiser does not consider all relevant data, overlooks relevant data sources, or relies on incomplete data in the research and analysis stage of the appraisal process, the result may be a poor quality appraisal that could have a discriminatory effect. For example, when the only data that is researched and relied on is data obtained from a sales data reporting service or a local multiple listing service—and that data source was not used for most of the sales transactions in a particular neighborhood or market area—the appraiser may arrive at an inaccurate opinion of value. For specific information concerning sources of manufactured housing data, please refer to Section 304.01.
XI, 403.03: Trend of Property Values, Demand/Supply, and Marketing Time (01/31/06)
The appraiser must report on the primary indicators of market condition for properties in the subject neighborhood by noting the trend of property values (“increasing,” “stable,” or “declining”), the supply of properties in the subject neighborhood (“shortage,” “in-balance,” or “over-supply”), and the marketing time for properties (“under three months,” “three to six months,” or “over six months”) as of the effective date of the appraisal. We also expect the appraiser to describe the reasons when the trend of property values is declining, supply is an over-supply, or marketing time is over six months.
The appraiser’s analysis of a property must take into consideration all factors that affect value. Because we purchase mortgages in all markets, this is particularly important for market areas that are experiencing significant fluctuations in property values (including sub-markets for particular types of housing within the market area). Therefore, lenders must take appropriate steps to ensure that the appraisers they use analyze listings and contract sales as well as closed or settled sales, and use the most recent and similar sales available as part of the sales comparison approach, with particular attention to sales or financing concessions in markets that are experiencing either declining property values, an over-supply of properties, or marketing times over six months. (Also see Section 406, Sales Comparison Approach to Value.)

 

[Michigander]

which is exactly why I gave you specifics

Above are specifics from one neighborhood, 2 different models, one subdivision. It happens to mirror the zip code as a general rule. Of course you have different submarkets and if you know that yours are not declining (or increasing) based on your expertise, then support it with market data as shown above. Otherwise a simple trend as shown above (which works in this market due to the general uniformity of the market) will work.

You can support model for model, subdivision for subdivision, size for size, etc. Our MLS lets us break so much of this out that you can set your own perimeters quite easily. You can even do the sale/resale thing as long as you check that there hasn't been any significant changes from one sale to the next (although I would make that broader and do 2 or 3 years back instead of the 1-year recommended elsewhere, simply because that way you will have a much larger pool to choose from and will see the larger picture over time).

 

[Richard Carlsen]

I typically try to define the neighborhood as that area where there is similar housing with similar appeal. However, in a low density market that is often extremely difficult. Even where it is possible, it is nearly impossible to draw defend-able conclusions from the data.

Example: The photo below shows a property that I am working on right now. It's located in a rural sub about 5 miles from the business/economic center of the market in a 40 year old plat with 42 lots, almost all of them improved. In the last 12 months there has been 1 sale in the sub for about $80,000 with 1 property currently listed for $135,000. Now what conclusions can I draw from that data from within a well defined neighborhood? None that I would want to hang my hat on.

So to provide the reader with some data from which they can understand the report and make the lending decision, I expand the market data to included similar size/style/age off water non-acreage properties from the greater market area which tends to be very similar. That is the only way that I see that I can provide data that has any validity or applicability to the report.

 

[Artemis Fowl]

Guess I'm just repeating what has already been said. I should read more slowly. This still doesn't address Richard's problem with low density. 5 sales can give you a median but that's not exactly high confidence data (for my market). You say that the year before had the same number of sales. What was the median for those? Maybe reconcile sale/resale with sub-market medians? How do other low density people do it?

<Edit> Was writing while you were posting...
That seems a valid approach. Another appraiser or reviewer is going to run into the same issues. Perhaps the method is more defensible that the size and scope of the sample...meaning your 5 relevant sales from similar markets vs. someone looking at ALL sales within 20miles.

 

[Tan V. Nguyen]

I heard that Lenders do not want to see in the report Property value trend is "declining." Is that correct?
I am newbie.