More MOI data vs. price direction

All data taken from MRIS database.

Two more graphs (20171 - Herndon/Oak Hill & 22124 Oakton) to add to the graphs below (Manassas & McLean). To make a long story short, I looked at the data every which way and there is no evidence of months of inventory predicting price direction.

Median home Prices (red line and left axis)

Months of Inventory (blue line and right axis)

Does Months of Inventory predict price direction?

A while back I went through the MRIS database and looked at months of inventory for several zip codes to see if they predicted prices changes. Going against conventional wisdom, I found no relationship. Here is one of the graphs (20111). Notice the broad range of MOI, so there is no problem of range restriction.

EDIT: added time series below

Spring sizzle or fizzle?

So, how did Spring 2009 turn out? First off, let me state that I'm reporting on the areas I've been following closely (see below). But, this spring seems to be marked by two interesting, and traditionally contradictory trends: (1) increased sales volume w/decreased days on market, and (2) decreasing prices.

Interestingly enough, I've seen quite a few houses sell within one week, if not the first day! This is quite a turn-around from the recent past, and might remind some of the days of bidding wars and such. But, this doesn't seem to lead to increasing prices. So, what's up?

My hunch is that affordability is driving sales volume. Affordability, in the case, is due to declining prices, unusually low mortgage rates, and maybe that $8k tax credit (although I'm not sure how many qualify). Anecdotedly, I've seen an unusual number of contracts fall through, only to come back on the market later.

Enough blabber, time for data.

Types of houses:

First let me define the types of houses I'm looking at and the geographic areas. Houses are 4+ bedrooms, single-family homes, no condos or townhouses, 2000+ sq ft (this is above ground, does not include basement), and a price range from $400k-$1.25 million. Data has been gathered from the Washington Post, Fairfax County database, RedFin, and Frankly MLS. All data confirmed against the Fairfax County database. Does not include seller subsidies.

Geographic Areas

"Franklin Farm" Perhaps misnamed in the graph, but this areas represents zip code 20171, and may overlap with a few other zip codes. This includes Franklin Farm, Oak Hill, parts of Herndon, and the western edge of Oakton.

"Oakton" This mostly includes 22124, and only includes homes that go to Oakton Highschool (sorry Reston and South Lakes!). This also includes the western part of Vienna near Hunter Mill and the eastern part of Herndon (20171) near Fox Mill.

Normalization Method

I control for 'quality' by comparing the prices to the 2006 tax assessment. Some can quibble with this, but on the macro level I can not think of a reason why this would interact with the mix of houses sold. Short sales and foreclosures included are included in the data, but not the 'sale' back to the bank. I exclude houses that are blatantly mis-assessed (i.e. a house with an addition that doubles the square footage, but is not reported). I also exclude tear-downs or houses that are badly in need of repair.

Sales Volume

First off, because both of these graphs are 3-month moving averages, and there is some latency in data reporting, so the last two data points should be ignored.

The tentative conclusion is that, although volumes have increased, we really only saw a spring bounce for Oak Hill and not Oakton. Volumes also seem to be down compared to last year.

Why didn't Oakton see the same bounce as Oak Hill? The two areas are very different. Oakton has large treed lots (.5 acre or more), and many of the houses back up to parkland. Oak Hill is suburban, with smaller lots of 1/5th to 1/4 acre size. Oakton is also a little closer to DC and metro stops. Typically, the same house will sell for much less in Oak Hill (25% discount? just a guess).

Prices

How did prices fair in both areas? In a nut shell, they are down, with Oak Hill declining at a quicker rate than Oakton (not surprising). All graphs can be clicked on to enlarge them in a separate window.

In Conclusion

Despite all of the incentives (cheap mortgage rates, etc.), this spring has been a bit of a fizzle. Price and affordability drives home sales. Keep in mind that the cheap mortage rates were not available for conforming jumbo (loans > $417k and < $7xx k), which undoubtedly hurt Oakton. Notice in this next graph that houses > $800k are selling for a greater discount (data from FranklyMLS, and includes seller subsidies):

[after 10pm, my writing goes to hell, so I hope this makes sense]

Recently, there have been quite a few questions about what Fairfax county assessments represent, and whether they drive the market or not. Fairfax county assessments are supposed to be based on sold prices. There are two important words here: "sold" and "supposed".

I'll tackle the last one first. When I mean "supposed", that implies that the assessor does not always get the comps right. For example, a house sold in 2008, and it was clear from the assessment that the comps had not been updated. The house was originally a 2000 sq/ft colonial with an attached 2 car garage. The garage was extended to 3 cars, and the second floor extended over the garage. There was also an addition to the ground floor behind the garage. All told, the square footage increased roughly 50%, yet the assessment was comparing it to identical models on the same street that had not been updated.

The second word was "sold". The assessment is based on the sale prices of comps, and these, naturally, have occurred in the past. The comps may or may not represent the value of the home today. If the market is flat, then they will align. If values are increasing, the assessment will be less, whereas if values are decreasing, the assessment will be more.

So, how far back do the assessments lag? The first graph below is data taken from 10 neighborhoods taken from the Oakton wedge (my name for the region). This area is bordered to the north by the Dulles Toll Road, to the south by 66, to the west by Fairfax County Parkway, and to the east by 123. So, this area encompasses parts of Vienna, Oakton, Reston, and Herndon (Oak Hill?).

All the homes are single-family homes, 4br with a basement and 2car garage. All are colonial style, and none are split-levels, split foyers, etc.

Because the neighborhoods are at different price points, all of the prices are normed by dividing them by the 2000 assessment. The pink line is the sales price and the blue line is the assessment. You can see that in 2000 they were selling for roughly 30% more than assessed. Interestingly, if you shift the pink line forward by 18 months (yellow line), it aligns with assessments almost perfectly, at least until 2007-2008, where things break down a bit. But nonetheless, it looks like assessments for a given year more often than not reflect sales prices from 18 months ago.

One caveat about the 2009 sales: only 3 of the 10 neighborhoods have had sales in 2009, so the 2009 price point my not be a very representative sample. But, it is certainly suggesting that houses in 2009 (they probably closed late 2008) are selling for less than the 2009 assessment.

[edit: found error in graph -- assessments changed slightly]

Using median (or mean) house prices as one big disadvantage: they are extremely sensitive to changes in the underlying distribution.

For example, maybe in 2006 60% of the homes were 3br, 30% were 4br, and 10% were 5br. Fast forward to 2007, and let's pretend that 40% of sales were 3br homes, 20% 4br, and 40% was 5br. Undoubtedly, the average selling price would have increased from 2007, but can we say that prices of the individual homes increased? There is no way for us to tell. It could be that the increase in the average sales price was due to the shift towards more expensive 5br homes (that perhaps actually decreased in value).

That's where the Case-Shiller method comes in (although I understand that OFHEO uses a similar method). CS looks at pairs of sales. In other words, it calculates the change in value for a single house over time. The index for a particular time is a function of the average price change, as opposed to the change in the average price (see example in above paragraph).

OK, so it's a great method. But it's a real pain in the ass to mine county records by hand. So, here I'm going to introduce a new method, which norms the data based upon the 2006 county assessment. Why norm to the county assessment? Because, like Case-Shiller, it is gives us an apples-to-apples comparison: I can compare the sale price of a house in 1990 to the 2006 assessment. Likewise, I can compare the sales price in 2010 against the 2006 assessment.

Note on Fairfax assessments: unlike other regions of the country, Fairfax homes are assessed every year based on recent sales of comparable homes. If your neighbor just sold their home for $800k, that means that your identical home will be assessed around $800k next year. Granted, any assessment, like any appraisal, has some error, but it is a reasonable baseline.

Below, I present you the normed sales trends for Oakton and McLean, which are two of the wealthiest areas in the region. But first, a little about the methodology.

The homes are all 4br SFH homes. The list was first gathered using Redfin (www.redfin.com) to gather all homes sold in the last 36 months (back to early 2005 or so). Sales from before that time were gathered from the Washington Post's real-estate sales database.

Tax information was gathered from the Fairfax County online database (http://icare.fairfaxcounty.gov/).

Homes excluded:
Sales to relatives
Teardowns, or other instances where the property was substantially improved
Sales of properties split into multiple parcels
Foreclosure and REO sales to the bank

Homes included:
'Normal' arms-length sales
Relocations
REOs, foreclosure, short sales when the sale is to a new owner. That is, it does not include the sale back to the bank.




As can be seen for both McLean and Oakton, the best fitting straight line doesn't describe the data (R2 ~ .01), whereas the polynomials account for 26-33% of the variation.  You can also see that most homes in those two areas are starting to sell below their 2006 tax assessment, whereas at the peak in 2006 most sold for substantially above the 2006 assessment.

Inspired by the Case-Shiller Index, I decided to make my own index of prices for various Northern Virginia neighborhoods. For those unfamiliar with the Case-Shiller index, unlike the NAR and other data on home prices which use parametric statistics, usually the median, to gauge home prices, the Case-Shiller index looks at repeat sales. Parametric statistics are vulnerable to changes in the underlying distribution (e.g. what if large houses become unfashionable).

For my initial foray, I decided to examine the prices of 4BR (four bedroom) SFH (single family detached homes) with a basement with between 2000-2500 square feet. Note that the square footage excludes the basement. If you include the basements, then square footage would range from roughly 3000-3750 sq ft.

I chose this size of home because it appears to be a popular size throughout the region, making apples-to-apples comparisons easier among neighborhoods. In older neighborhoods, these might be considered larger houses, whereas in newer neighborhoods these might be on the smaller side. Also, these tend to be colonial-style homes: no ranches or split-levels.

I originally chose to sample more neighborhoods (e.g. McLean and Manassas), but the task of finding comps (square footage, taxes, etc.) is a bit tedious, and I ran out of steam. All data is from the Loudoun and Fairfax county tax databases.

In the future I'd like to add a second size bracket (3000-3500 sq ft) and include Manassas and McLean in the mix.



• The 3rd graph, price percentage from peak, does a nice job of illustrating the cascading fall of prices as it works from the outlying areas inward.

• The 2nd graph, percentage of 1998 prices, illustrates the huge run-up in the region.

• The 3rd graph, in case anyone cares, is of the mean values of the houses sampled from the difference areas.