$/sq foot falling because fewer low end homes sell
Thursday, May 3, 2007 at 12:24PM In response to this post asking about $/sq ft, I want to devote this morning to setting the record straight. Is $/sq ft a good measure of changes in price? I read the Voice of San Diego article that MFMauceri is referring to. The writer does a great job of explaining the shortcomings in the median, namely buyers are getting better homes for their money, and the lower end buyer is priced out or turned away by lenders. So we've got a market where the higher end homes make up a greater percentage of sales. Thus, the median is rising, the average is rising, and guess what folks. Sit down for this one. And the $/sq foot is falling. Yes, the fact that the $/sq ft is falling does not prove at all that prices are falling, but only shows that more of our sales are higher end homes.
In a housing market downturn, the median rises, the average rises, and the $/sq foot falls, as all 3 measures move in the direction just noted when fewer low end homes sell.
In the beginning of a housing downturn, fewer cheap houses sell. That makes the median price go up, the average price go up, and the price/sq ft go down (since cheaper houses cost more per sq ft).
I can sort of see why people are flocking to $/sq ft. Those of us who know that prices are dropping, are really frustrated by the media emphasis on the median and average price, which are still rising. We know prices are falling, but those measure are rising. The median price in all Southern CA counties rose in March over February, while prices were falling. Some folks see a falling trend in $/sq ft, and jump on it. Hey, finally we found a measure that is falling. But they are taken another distribution number, and using it just because it is falling? Hey, let's think about why the $/sq ft is falling.
Why is a larger house less expensive per square foot? Due to the high fixed cost of land, the land is a smaller component of the price in a large house. Due to economies of scale, larger homes cost less per square foot to build.
Let's look at some actual numbers for San Diego County, for homes that sold this year. Notice that the cost/sq foot falls from $417 to $326 to $303 as we go to bigger homes. When we get into the much larger homes, the price goes up again, because those homes are just in a league of their own. They have tennis courts, movie theatres, and other extravagances that raise the price. Those are the homes that are not affected by the credit crunch, and whose prices are sometimes still rising.
Prices in San Diego have dropped 15%, more in some areas, while high end homes are still selling over asking price (I have 3 examples of that happening in the last month). Has anyone seen any measure which shows that drop? Look, the Case-Shiller index is a poor indicator of price too. The market was cooling since April 2004, but the Case-Shiller index was rising until late last year or early this year. Please do not pay any attention to any of the price measures out there.
We simply cannot cull real estate prices into one number. We can't do it by neighborhood either. There is only one way to know what is happening with prices: evaluate the price of a home today versus another time. It's a home-by-home comparison, and I cannot think of any way to aggregate that.
Read below only if you absolutely need the data
Now I will give the data to show the point I am making. This is data from the MLS, for homes sold this year. I did a search using a few square feet on either side of the size I was looking for, and show the number of results.
Size of home..... $/sq ft .....# of homes
1000 sq ft home: $417/sq ft 182
2000 sq ft home: $326/sq ft 172
3000 sq ft home: $303/sq ft 80
4000 sq ft home: this is totally a different league, where we have huge lots, wine cellars, home theatres: $331/sq ft 29
5000 sq ft home: $ 413/sq ft 8
Now, some may think we can compare the $/sq ft of these homes to the same period last year. But we really can't. In a cooling market, only the good homes sell. In 2004, every 1000 sq ft house sold. This year, only the very nice 1000 sq ft homes are selling. The quality of the homes sold has gone up. Now, the house is only selling if it has a good kitchen, a good view, a great school district, a great yard, etc. So we cannot compare the prices of homes just by looking at a bunch of numbers.
However, for the curious among you, let me give you some information on how $/sq stacks up when we compare this year to last year (1/1/06 - 5/3/06, compared to 07). Notice the huge drops in the number of homes sold, and that the $/sq ft has dropped about 5%. Meanwhile, the home prices have fallen 15% or more, again proving that the homes which are selling this year are the ones of higher quality. If we were to sell the exact same homes we sold last year, the $/sq ft would be down 15% in many areas. If you want to come up with a "correct median", calculate the sales price each month of the SAME group of homes that sold in a base year.
The sample of sold homes from this year compared to last year shows that the big homes make up twice as many sales this year as last year, thus raising the median and average, and lowering the $/sq ft.
While $/sq ft is higher for those very large homes, the number involved is so small, that they are not skewing the $/sq ft up. So the $/sq ft is falling because the 1000 sq ft homes (high $/sq ft) make up only less of our sales this year (38.6% versus 41.4%). The 5000 sq ft homes were .09% of all sales last year, but are 1.7% this year. Both years they are only 8 homes, but those 8 homes have a bigger effect in a year of less sales. This happens because the rich people are not affected by the credit crunch, and the mood for real estate is not yet pessimistic. The rich are still buying.
For those interested in the data, it's right here:
Below, the size, 2007, and 2006 information . ..................................................% of total sales in 2007, in 2006
1000 sq ft, 182 sales, $417/sq ft.........355 sales, $440/sq ft . Sales down 49%, price down 5.2%....38.6%, 41.4%
2000 sq ft, 172 sales, $326/sq ft.........303 sales, $342/sq ft. Sales down 43%, price down 5%........36.5%, 35.4%.
3000 sq ft, 80 sales, $303/sq ft...........150 sales, $329/sq ft . Sales down 47%, price down 8%.......17%, 17.5%
4000 sq ft, 29 sales, $331/sq ft............ 40 sales, $350/sq ft. Sales down 28%, price down 5%..........6.2%, 4.7%
5000 sq ft, 8 sales, $413/sq ft.................8 sales,.$ 577/sq ft Sales flat, price down 28%................1.7%, .09%
Total: 471 sales in 2007................856 sales in 2006
Reader Comments (14)
Many years ago a pioneer named Jeff Sagarin (google him) came up with predictive models for sports betting. He was so successful that his models are used in determining which college goes to which bowl game in the BCS .
Seems real estate is in desperate need for an accurate pricing model. Who wants to be famous and create such a formula?
You'd need to create an index, that is manually prepared, not off recorder's office sales.
A realtor/appraiser would have to appraise specific homes every month, or quarter, and would need access inside those homes, to adjust for improvements made to them.
It would be expensive, and who is the target customer that would pay for it?
Would you, MAS, pay for it, if I provided this service for a subset of 100 homes in San Diego?
Remember, Case-Shiller index is free because S&P500/Macro Markets is promoting their index trading. OFHEO index is used by the government for various things, so the taxpayer pays for that, it's just another government data series. Dataquick is paid to put out the median...not sure if the newspapers have to pay to get that data, or if Dataquick gives the price data free, hoping to get more business for selling their other data. So they all have a reason to give out their data, and not a single one of them benefits by being accurate.
We need someone who makes money off being accurate. That was my idea in starting this forecast. You get accurate information, but I've made so much of it available for free, hoping that people will send this pricing misinformation out to everyone they know. We need to get the word out. But if I were to create a pricing index, I'd have to charge for that. It would be expensive. Would you pay for it? How much is it worth?
For me, I don't need to know how much prices are dropping. I only need to know that they are dropping. So months supply is all I need.
But I agree it would be interesting data.
How about if we looked at just the listing prices and sold prices? If we are just tracking the Price, then the perception of what the seller believes should be all that matters?
I don't think we would need appraisers because all they do is look at recent sales and apply it to a specific homes.
Can't we just assume that with a large sample size, in aggregate, we can assume the sellers are selling at the "right price". Then we compare the same house to itself over months/years. Since it's a comparison of the same house, it's all relative to itself which elminates a lot of variables.
We just need a large sample size and with a lot of people keeping track of a few homes we would be able to do it with relative little trouble.
It would be lovely to have an accurate predictive model for both real estate and the stock market... woo hoo, we're all rich! However, there are just way too many variables. I believe Zillow.com has legions of folks working on this... and they're wayyyyy off on their numbers.
I think the best route is to pick the three or four metrics and then use some horse sense and gut instinct (something Schahrzad has evidenced by this post) to tie it all together.
wbbgjr - I think we could do it, except that listing prices are still too high. Jim Klinge wrote on his website that we would sell more homes, if sellers would lower their prices by 5-10%. That means the homes that are not selling, are overpriced. The demand is there, and they would buy if the price was lower by just a tad.
Let's compare that to the bottom of the downturn, when demand is just dried up, due to fear of prices falling further. We're not there yet.
We know the market value of a home from offers made, and sales prices. If there were a way for someone to say how much a house is worth, such as an appraiser, we accomplish the same thing. Except even appraisers can come in too high.
I'd suggest using sales prices of a large set of homes. These sales prices would need to be adjusted for changes to that home over time, especially that large remodeling boom we had, with all the room additions, kitchen and bathroom remodels. We'd have to adjust for that in the higher price of that house today.
So it's a manual process, and it can't be done via a database. We need people on the ground, reporting. People following their own neighborhoods, perhaps.
Any takers on starting such a measurement?
I am going to be attending an REO auction at the San Diego Convention Center this Saturday. This should give me a pretty good idea as to what the 'bid' is for this real estate market.
Would you report on your assessment of REO auctions? Why are lenders so stubborn on price? They are not in a hurry to sell?
I will definitely do that. Most of these properties have starting bids of 30 to 50% below the Zillow recent sales numbers; which are much lower than the Zestimates, BTW.
The lenders are stubborn because they are not yet under pressure from the mortgage investors to liquidate holdings at any price. However, this too, will slowly change with the passage of time as more and more REOs go to auction, and as the credit markets start to fall apart.
The company that is holding the auction is REDC (Real Estate Disposition Corporation):
[Link]
I am very confident that this is in fact the start of many more auctions that have yet to come.
Someone told me, he knows some investors who are pooling money to buy groups of homes from the lenders at a discount.
There is also the complication that 75% of recent mortgages (purchase and refi) are sliced into MBS, so there is not one person holding each loan. Instead, a bunch of investors hold the rights to the payment streams of groups of loans, and that right is further leveraged and diced somehow. I don't really understand the whole way this is done. I read though that the legal structure of the MBS prevents the FDIC and Fed from issuing any regulations on foreclosure prevention or workouts, and that these loans backed by the investors have to run their course.
So the government intervention and bailouts can only help a fraction of the market, something like 5%.
I am anxious to see how the Saudis, Chinese, European pension funds and insurance companies, and US hedge funds, deal with the losses on their MBS, and how those properties will be liquidated.
Someone told me, he knows some investors who are pooling money to buy groups of homes from the lenders at a discount.
Well, then they will be in for a rude awakening when they try to flip into a market with 20000+ properties for sale and rising.
Anyway, this is the list of properties that will be up for sale at the San Diego Convention Center on Saturday.
[Link]
I've been to a couple of open houses to look at some of properties first hand. The Mira Mesa SFH properties looked like 'fixer-uppers' and the Poway SFH properties looked like they were in 'Move-in' condition. I'm not expecting many bids on the 'fixer-uppers', but I am expecting some bids on the 'Move-in' condition properties.
Anyone who decides to make a bid on these properties, sight-unseen, is definitely taking a huge gamble.
And at this particular auction, all multi-home purchase investors are required to purchase with 15% cash on each property, so they will at least have some skin in the game; which is a good thing, IMHO.
15% cash - you don't need 100% cash?
I checked a Poway and Laguna Beach home on the list, on ForeclosureRadar.com, and both these homes are bank owned. So the banks are using this company to auction their homes?
15% cash - you don't need 100% cash?
Yes
So the banks are using this company to auction their homes?
Yes. Except that it didn't turn out very nicely for them.
There were a number of properties that sold at or near market value based on near sale comparables. However, only half the sales were able to finalize at the auction. There were many times in which the real estate buyer could not qualify for a loan in order to finalize the transaction; let alone come up with the required 5% downpayment.
When other participants started to get wind of this, the maximum bid would always keep going lower than the previous failed bid(s).
Thus, the final winning sale would almost always be at some discount to the market (usually by more than 10%). However, there were quite a few winning bids that were as much as 30% below the market. This was especially the case for properties that required a much larger down payment (i.e. 30% down) and for all the condos that went up for sale. There were very few people that were willing to step up to the plate and overpay for a condo, and as a result, the condos got creamed.
So in this case, the auction company tried to rah-rah the real estate buyers into overbidding against each other, but this tactic failed miserably.
...
"Do I hear 340K? Do I hear $345K? Oh you didn't say 340K? O.K. you withdrew your bid. Do I hear $340K? $340K? $330K? Are you o.k. with $335K?? Look folks.. This is supposed to be an up escalator for prices."
And from the winning bidder's point of view.. "O.K. I won the bid. What do you mean that I don't qualify for financing on the spot? Yes. I know that I don't qualify for any kind of real estate loan. Is that a problem?.."
Ladies and Gentlemen,
SD080 just went back up on the auction block. Now starting back at the original minimum bid..
...
Nothing sold for over $300 per square foot.
Thanks for reporting back. I'm really interested in this.
Do you have any examples of homes, and then we can put your comments into a story. Is there a specific address for which you remember the bidding price?
Also, why were people surprised they did not qualify for financing? I guess they didn't screen the people for financing before the auction?
Auctions seem to work best when you have a few qualified people interested in each property, and then they bid the home up to whatever the market value is. Do you think that happened here or not?
WSJ reports on the auction