I found an interesting tidbit on the Housing Bubble Blog. A researcher looking at historical California inventory levels found that once the supply of homes hits 9 months worth of sales, median prices fall "on a consistent basis." This is a statewide stat, but it at least gives us some general insight as to the location of that line in the sand past which increased inventory starts to push prices south.
In regard to price declines, what’s more important than the overall inventory level is the amount of "distressed" inventory supplied by owners who have to sell at whatever price they can get. But those two numbers tend to trend up and down together.
As I will discuss more in the monthly housing report (which is coming very soon, and will include new data sources to provide a more current read on the market than is supplied by DataQuick) the combined condo and SFR inventory in San Diego is currently a little below 8 months.
I dont know where you are
I dont know where you are getting your numbers but per Sandicor data we are at about 4 months for detached homes (approx 11,700 active and 3,053 pendings) and 5 months for attached homes (approx 7,292 actives and pendings). I’m sure its a little higher because some will fall out and some are in escrow longer than 30 days but I’m not sure how you get 8 months. Could you clarify?
Economist Kleinhenz, who
Economist Kleinhenz, who conducted this study,
oops – continued below
Economist Kleinhenz, who
Economist Kleinhenz, who conducted this study, used the current sales pace as a predictor. But then the author of the LA Times story, Gregoy Wilcox, used pendings, a clear misapplication of the study he was writing about, to arrive at a March statewide inventory of 5.2 months.
What I found interesting is the leverage of inventory on prices. At less than 7 months, prices climb 7% annually. But add only a couple months more of inventory, (9 months), and prices actually decline.
Assuming sales continue declining at the current rate of 18.3% quarter-this-year to quarter-of-prior-year, what inventory level will we need to get to the magical 9 months of inventory and see price declines which actually show up in the median numbers?
(As we already know, prices are declining but not showing up yet in the median.)
650 listings downtown
I was
650 listings downtown
I was under the impression that 650 listings downtown was significantly more than 9 months worth of inventory. Do I have my numbers wrong?
FYI: One of the new-construction condo developers in my part of town dropped his per-unit asking price by 100K in the last two weeks. He currently has at least 5 projects (and I think 7) in progress.
Let’s see, 5 projects times 4 units per project is 20 units times 100K. Ouch! That’s 2 million dollars.
What part of town is that,
What part of town is that, and which development?
Josh
Forgive me for the rookie
Forgive me for the rookie question, but can someone tell me what metrics are used to compute future inventory? I assume it has something to do with the number of properties on the market divided by an average sales rate per month…or something like that. E.g., 1000 properties on the market / 10 properties sold per month = 10 months of inventory. Is this correct?
Right idea but bad math
Right idea but bad math skills. 1000 properties divided by 10 sales per month would be 100 months of inventory.
Oops! Forgot a 0…my bad!
Oops! Forgot a 0…my bad!
Rich, assuming sales
Rich, assuming sales continue declining at the current rate of 18.3% quarter-this-year to quarter-of-prior-year, what inventory level will we need to get to the magical 9 months of inventory and see price declines which actually show up in the median numbers?
Bugs, it’s interesting that demand drops off first in outlying areas. Values are better maintained in desirable areas. Those outlying areas where everyone moved to avoid the high cost of intown living, are really feeling the pressure. Would it be the same for Temecula (a poster started a thread on that).
I can see there could be flies in the ointment, but I’ll still use the ointment. The MLS seems fraught with inaccuracies, because it relies on the accurate participation of thousands of individuals. As such, it’s done pretty well. I for one would like to see a new field added to the database, Total Time on Market since Last Sale, so each relisting has its DOM updated to this field, and we can get a Total DOM.
Yet, despite its imperfection, DataQuick uses the MLS to provide real estate market statistics.
I also bet that the guy who did the original study had to portend with those same issues of pendings/closed, so he had the same noise we have now.
I’m still not getting why my
I’m still not getting why my number and Rich’s is so far off. using the March inventory of 18,261 and dividing it with the March total sale of 4,146, I’m getting 4.4 months. I can’t see April being that far off from that. So where’s my mistake?
rich, you already have the
rich, you already have the April #’s? using the March numbers I only have 4.5 months of supply.
Are there even enough houses
Are there even enough houses downtown to count? It seems that its pretty much all condos.
Josh
If we’re counting how many
If we’re counting how many months of inventory for the sales, we need to stick to using closed sales. Pendings usually extend for more than 30 days and a few of them will drop out, especially if/when prices are in decline. Some will drop out as a result of not being able to get financing at that price and a few as a result of buyers getting cold feet.
Two slight flies in the
Two slight flies in the ointment. If you are using closed sales, shouldnt you be looking at the inventory levels when these homes sold (i.e. went into escrow). The pendings are based upon what is selling now vs what is on the market now and thus is a more current measure. Sure things will fall out but when the April sales went into escrow there was less inventory, generally higher asking prices (I’ve seen an increasing pace of price reductions) and less motivated sellers (market times are getting longer). I don’t think either measure is perfect and the truth probably lies somewhere in the middle.
The other fly is that alot of sales get reported as closed well after the house actually closes. Lazy/incompetent agents, discount brokers, Open Listings (FSBO’s posted in the MLS) are a few of the things that contribute to this. Many of the homes showing as pending have actually closed already. There are currently 1,000 homes that are pending which went into escrow between 2/1 and 3/31. How many of these closed in April? If half of them did the closed sales numbers will increase by more than 30% and Months inventory would decline by 1.5 months. It would not be unusual for April Sales to rise 20 to 25% by the end of May. The March numbers were about 20% higher than the current April numbers and my guess is that April will be higher than March when all sales are reported. I’d argue that you are better off using March Numbers against current inventory which would get you down to 6.5.
My point isn’t really to advocate one way or the other, but rather to continue railing against relying too heavily on poorly collected and reported data. The truth is that it’s just too early to use the April Numbers reliably.
The problem with using
The problem with using pendings is that they aren’t all 30-day escrows. There are many that are 3-4 weeks and there are others that stretch on as long as 2 months or more. The ones that runs less than 30 days mess the number up as much as the ones that run longer.
I ran some numbers on some of the north county cities and communities using the April solds and found the inventories vary somewhat relative to the sales. Carlsbad now has 6.5 months, Oceanside has 6.83 months, Encinitas has 7.22 months, Vista has 6.26 months, San Marcos has 7.18 months. Obviously these communities have a ways to go before they hit the 9-month mark. On the other hand Ramona has 10.6 months and Alpine has 9.54 months of inventory – I’d say those outlying communties are already there.
But that’s just one month’s data. A couple of these towns had less than 40 sales during April, and when the numbers get too low the noise factor can distort the picture. I’d feel a little more confident in the picture the numbers are showing if we used larger samples covering a longer period of time.
I think this spring’s selling season is a make or break situation for the people who are considering selling any time in the next couple years. If those prices actually do show some undeniable losses then I think some sellers are going to head for the exit and that’s when the price cutting will start in earnest. We may not need 9 months worth of inventory this time to trigger a price war.
I still think the ready
I still think the ready availability to the masses of data and the tools to analyze that data might affect where the trigger point is for both decreases and future increases. But that theory has yet to be demonstrated, so for now the 9-month point seems to make sense.
I think at this stage of the game the 9-month point is more likely to occur as a result of declining sales than increasing inventory. I think it will come in fits and spurts during the course any given year.
When prices were on the increase, the majority of the increase would occur during certain periods of the year and in tandem with restrictions in inventory. The rest of the time there were still some increases but the rate of increase was much slower. I think the reverse will be true for any declines. As volume picks up relative to the inventory the pricing should remain relatively stable, but as volume drops off the rate of decrease may quicken. The degree to which financing rates change would definitely have an impact on the rate of change in pricing, too.
This is why I don’t think we’re that close to a free fall at this point, unless the current crop of pendings are all closing much lower than expected. If the trend is at all inconclusive or the losses are only minimal I don’t think that would be enough to trigger a panic.
The thing that bears watching is that what we are referring to as “the market” is actually an array of little localized micro markets. Even within a local area, the different price ranges all have their own dynamics. It’s overly simplistic to expect all these market segments to act in unison – some will decline faster than others. During an increasing market swing the areas to watch for the trends are the highly desireable areas that have good access to employment and services and good schools. During a declining market the areas to watch will be the less desireable areas that are located farther from freeway access, employment and services, and some of the urban areas that have higher crime rates and a lower level of services. The primary exception to this is the downtown condo market, but that’s the direct result of developer speculation.
There’s something strange
There’s something strange going on with the current crop of pendings. When the realtors log onto their MLS, the Sandicor message flashes repeatedly, asking them to clear out their pendings. Sandicor is checking into all pendings now, since they have been increasing in number over the past months, and are now at 5,000!
Why? Did these fall out of escrow, and Pending status was not removed? Did they expire? These cannot be closed sales, since there are just too many.
By the way, so far we have 2500 closed sales for April. YTD, closings are down 30%, and the rate of decrease is growing.
YTD, 25% of all listings are expired, cancelled, withdrawn. It costs so much now to list a house, that some realtors might decide, “why bother”?
Questionable Pendings
Last
Questionable Pendings
Last year a 20 unit condo conversion came on the market in my area with prices in the high 400’s. No sales for 4 months and then 8 units went into escrow on the same day.
Maybe they really sold 8 units over the weekend and opened all the escrows monday morning but I doubt it. Units continued to be available in this project for another 5 months and were finally reduced to the low 400’s before selling.
My point is that a condo converter or builder can open up fraudulent escrows and have ‘pending’ sales. These escrows will never close so they can be done in the brother-in-law’s name or the dog’s name, etc.
The escrow company doesn’t know these are dummy escrows and as long as they get their fees they aren’t going to care. If the escrow company did know that the escrows were dummies, they could charge a reduced fee since the dummy escrows won’t require any work other than an eventual cancellation. Of course the escrow company would then be involved in fraud …
These ‘pending’ sales allow the converter/builder to tell a potential buyer that he has similar units in escrow at whatever price point he wants.
A really slimy salesperson might say something like, “Well, Mr. and Mrs. Smith, if you’re serious you’d better place your security deposit today. I just put three of these units into escrow at $429K!”
Another potential reason for fraudulent escrows is that lenders can be picky about lending on small condo projects. Pending listings might give an underwriter a false sense of security about completing a loan.
For example, imagine a 4-unit condo project and the first buyer is trying to get a loan. Most lenders have a guideline about how much of a given project can be owned by any single owner – typically this number is 25% – ie, no owner can own more than 1 unit in a 4 unit project. But at the point our buyer is trying to close his loan, the condo developer still owns 75% of the project. The developer might be able to mitigate this issue by having some of the other units ‘in escrow’.
Anyway, I believe there is some level of fraudulent ‘pending’ listings in our market. These fraudulent listings accomplish at least two things:
1. allow condo developers to simulate demand for their projects at whatever price point they decide
2. facilitate the process of obtaining loans on small condo projects
Closed escrows are the only meaningful escrows!
Another reason a developer
Another reason a developer may have for using straw escrows is to influence the appraisal used to underwrite the mortgage. The average appraiser would absolutely take those transactions into consideration and would probably give them more weight than the closed sales from other projects.
That would be a dangerous game for a developer to play, though, because sooner or later all those units need to be legitimately sold. Having a revolving door of different escrows in the project would probably show up on a lender’s radar (if they were looking).
I haven’t run into any situations like this and I haven’t heard of any so far, but it is possible. Until I did start seeing it I would operate off the assumption that they’re all legitimate. Those few that do fall out of escrow will do so for legitimate reasons like credit/income or not appraising out and I wouldn’t anticpate those to be very many in number at all.
In the last week I have seen
In the last week I have seen in two different properties listed for sale the note that the property would not be available for possession until sometime in 2007! These were for two houses in the same very affluent and desirable area. One was listed for 1.7M and the other for 3.75M. I have never seen this before. I track the market there in a very informal way and would say that inventories are definitely way up. Is the long delay in possession considered a positive market sign. Has anyone seen this before?
Are they under construction?
Are they under construction?
Sorry, I should have made
Sorry, I should have made that point clear. These are both existing (at least 25 years old) houses.One the 3.7M is advertised as being available for possession in April 07 and the other for Feb 07. I happen to know that the seller of the more expensive house is not doing anything (ie remodel) to the house and the other doesn’t appear to be doing anything either.The more expensive house is being sold by owners who have lived in the house for at least the last 15 years and are preparing to downsize. Don’t know about the other house.