That’s today’s headline in the Arizona Daily Star. The gist of the story is that there are record number of unsold homes flooding the market, and that the explosion of new houses and condo conversions have contributed to the glut of unsold homes on the market.
Let’s take the research a little further, and see if the paper’s results hold up to actual numbers.
The Star Says: Residential listings averaged 9925 at the end of January, February, and march, nearly triple the average for the same quarter in 2005, the peak year for sales.
It’s very true that our number of listings has increased dramatically. However, looking at years 2002 through 2004, we see a pretty stable number of listings. In 2005, the investors hit and our inventory dropped, which led to a massive decrease of supply for a large increase in demand. Since 2006, we’ve seen number of listings rise pretty steadily. Number of listings on the market right now is well over average.
The Star Says: Sales for the quarter (2007 Q1) were down 18 percent from same quarter 2005.
Here’s a chart of sold homes and townhomes/condos, show per quarter. I’ve also plotted a smoothed version of the data. You can see our seasonality in all the bumps in these lines! From this chart, we see that sales were rising at a nice pace, started increasing around 2004-2005, and are now decreasing since 2006. The question we’re all waiting to get the answer: will sales continue to decrease, or will we level out?
The Star Says: Average number of days on market is 67, up 27 days from same time 2005.
Why do they keep comparing to 2005? 2005 was an exceptional year for Tucson, in every way. Here’s what the days on market looks like when we consider a larger amount of time.
Average days on market was pretty steady, staying between 50 and 60 days for quite a bit. In 2005, days on market took a nosedive, as the market sped up to a frenzy: low inventory + lots of buyers = low days on market. Since then, we’ve been coming back up to more normal levels. Have we peaked or will DOM keep rising? Only time will tell. It’s a shifting market right now.
The Star Says: The median price for the past quarter (2007 Q1), which is roughly $220k, remained unchanged from same time 2006 and is up 20% from 2005.
Median sales price has been steadily rising for the past 5 years. We had one small dip at the end of 2006, but median prices are back on the way up. In this graph, I’ve separated single family house prices from condo/townhome prices. They follow the same pattern.
The Star Says: Condos are a glut on the market.
Here’s a chart showing the number of single family home listings and condo/townhome listings per quarter over the past 5 years. While it is true that the number of condos for sale has doubled, so has the number of single family homes. It’s hard to place the blame solely on a “condo glut” due to all of the condo conversions available in town.
Let’s look at how many of those listings are selling. Here’s single family homes:
It seems there’s a glut of single family homes on the market in Tucson! The number of sales doesn’t seem to be keeping pace with the number of listings.
Here’s the condos and townhomes:
Well, this doesn’t look much different than the single family houses chart, does it? I think it’s fair to say that condo conversions may contribute to the glut, but certainly aren’t the only cause.
There you have it. The Star isn’t that far off in their conclusions. Right now, the real estate market in Tucson is providing more questions, than answers. Stay tuned for regular updates!
May 8th, 2007 at 11:32 pm
“Why do they keep comparing to 2005? 2005 was an exceptional year for Tucson, in every way.”
They do the exact same thing up here in Phoenix — and it was exceptional here too. It drives me bonkers. Comparisons to a year full of aberrations like 2005 simply isn’t right, or “good statistical practice”.
Nice analysis! Nice blog too, I added you to my feed reader.
Thanks for stopping by my blog!
May 8th, 2007 at 11:38 pm
You had me at “good statistical practice.” I never trust a statistic I didn’t generate myself.
Thanks for the comment and the kind words! Woo-hoo!
May 9th, 2007 at 5:04 pm
The reason the comparisons are being made to 2005 is purely for reasons of sensationalism. It seems to be more typical to compare real estate numbers to the previous year, so as to minimize the seasonal fluctuations. One way to make your analysis more robust would be to fit the pre-2005 data to an exponential. Then plot the exponential, obtained with the pre-2005 data, to the most current data. If you want to really do this right, correct the house prices for inflation before you do the fit.
A plot of inflation adjusted house prices that is, somewhat, famous can be found at
http://www.biggerpockets.com/images/blog/shillerbig.gif
The plot was created by Robert Shiller. He is a well respected economist that predicted the stock market bubble.
You are right to question the analysis of other people. However, it is not possible to do everything by one self. Also, you have to be careful about your own personal bias when looking at data.
May 9th, 2007 at 8:47 pm
Hi M Dugger! Thanks for weighing in. You make some good points there. As agents, we’re fed a lot of blather and marketing disguised as actual information and statistics, so I tend to be skeptical of other’s data and analysis. However, I know I’ve forgotten a good portion of the statistics knowledge I had back in college! Can I ask why you would plot the exponential?
May 10th, 2007 at 5:59 pm
The reason that the data should be fit to an exponential is not due to statistics. The reason for the exponential has to do with steady price appreciation. This is just like what happens with compounded interest on any investment. There are a few important cases where exponential should be used. In fact, the exponential is the function to use for population growth, stock market gains, inflation, just to name a few applications.
Why:
Unrealistic example:
If I receive 1% interest per day and start with 100 dollars, then the amount I have after one day is 101 dollars.
The amount of money I gain -For that one day- is 1 dollar. To determine the amount gained for the second day, I have to take my -Now- 101 dollars and multiply it by 0.01 (1%) to get 1.01 dollars gained for that day. For the third day, I multiply -my now- 102.01 dollars by 0.01 to obtain a gain of 1.0201 for that day. The amount is now 103.0301.
This goes on and on.
This can be written in a compact way:
gain per day = interest*amount
If I write
gain per day = (difference in amount)/(difference in time)
and please allow me the notation:
Delta = “difference in”
then I can make the differential equation:
Delta(amount)/Delta(time) = interest*amount.
The solution to this differential equation is an exponential. In particular,
amount = (initial amount)*exp(interest*time).