Last Downturn

User Forum Topic
Submitted by jg on September 5, 2006 - 1:06pm

I'm trying to get a sense as to when we'll see prices bottom. To do such, I've plotted two sets of data: (1) annuals sales for San Diego resale homes over '88-'03 (source: index from Robert Campbell's book, "Timing the Real Estate Market") against (2) median price for San Diego resale homes from OFHEO.

If I knew how to post graphs, I'd post the graph. What my graph shows is a sharp 49% drop in annual sales volume (moving from a peak of 50K in '89 to 26K in '91), with annual sales volumes remaining at ~25K over '91-'96. Prices peak in '90, and reach their lowest point in '94. Prices remain at this low level +/- 1% over '94-'96. Only in '97 do prices resume their march upward.

Conclusion -- in the last San Diego real estate downturn, only after sales volume reduced from peak levels in '89 to 50% of such in '91, and remained at such low levels through '95, did trough pricing appear over '94-'96, a lag of 5-7 years from peak sales volume to trough pricing. Only after sales volume picked up for two consecutive years, over '96-'97, did prices finally began moving up (+5% in '97). So, if one is a "bottom fisher," one may have one-to-two years of increasing sales volume as a signal that it's time to reenter the market. And, based on the last downturn, given that peak sales volume occurred in the first half of '04, things should get enticing, from a price perspective, in '09.

What do you think?

Submitted by no_such_reality on September 5, 2006 - 1:35pm.

It'll bottom when it bottoms.

If you attempt to time it, if it isn't like the last downturn, you'll miss it or hop in on the slide.

If you want to live in the house, find one you like, the price is right, the loan terms are right and the alternative living options make it right in expense and lifestyle.

If you want it as an investment, then run the ROI and when it hits your number, buy.

Submitted by bob007 on September 5, 2006 - 2:13pm.

i think it is foolish to predict the bottom.

Telecom/wireless/broadband/Internet/cable boom powered the 1990s. Southern California was going to benefit sooner or later.

Unless you have something to power the economy you might not see anything to prop up prices in Southern California.

The only givens for So Cal economy
1. San Diego will be a major naval base in the Pacific given the riseof China.

2. LA will be a major port for Asia commerce.

3. lot of low skilled Mexican immigrants

4. Hollywood will be a major player in global entertainment business

Submitted by VCJIM on September 5, 2006 - 4:29pm.

I think trying to predict when we're at the bottom is the right thing to do. It doesn't have to be at the very bottom of the trough, but somewhere "bottomish" makes sense. As he wrote, the two year period in the mid-90s was pretty flat at the bottom, I see no reason that similar indicators won't exist this time. Of course, I'm biased : ) I'm also trying to predict the bottom.

Submitted by powayseller on September 5, 2006 - 5:16pm.

jg, I just received my copy of Campbell's book today, and just a couple chapters into the book, I am a devout fan. I can't wait to finish it. So far, I think the guy is brilliant.

Anybody who says you can't time the real estate market, needs to read Campbell's book first. I don't blame you for not believing my prior posts, claiming I could time the real estate market, because I have not given you the graphs yet to show my method will work. But Campbell has! I can't believe I didn't get his book last year...

jg, could you post your graph. Use the "add image" option below the Comment box. That takes you to a new screen, where you can give the name of your file.

I drew a graph to picture your description, but I am not sure how the price and sales change on the way up, so a more detailed description, or a graph, would be helpful to me.

On the way down, it sounds like prices lag sales by 2 years.

Submitted by bob007 on September 5, 2006 - 5:52pm.

housing is a social need not a financial asset.

trying to predict the bottom is as silly as the folks trying to buy a house because they will be priced out of the market when prices were increasing.

if prices in san diego fall 30% and you can afford a home at the reduced price you should buy. if it falls another 20-30% you are going to have to suck it up.

Submitted by PD on September 5, 2006 - 6:27pm.

Trying to predict the bottom is not silly. We won't know for sure when we are at the bottom but we should be able to make a good guess as to when we are near it. Buying in the "bottom range" is a smart thing to aim for.

Submitted by lewman on September 5, 2006 - 7:44pm.

I appreciate the post and I don't think the important point here is to time the exact bottom. Rather, there are perhaps signs that one could use to see if we're getting close to the bottom, or the worst decline is over, or are starting to emerge from the bottom, and therefore hopefully get a better price. Rich has a chart that plots income vs price, and I also think that would be a good set of numbers to look at. And volume, as suggested here, seems to be a good one too.

Looking forward to seeing the chart.

Submitted by jg on September 5, 2006 - 7:52pm.

PS, we look forward to your thoughts after you read Campbell's book.

Will do, tomorrow, post the graph of the data.

I will buy a particular property (1) only when the fundamentals make sense (i.e., reasonable relationship between imputed rent and price) and (2) may delay a purchase if it appears that the market is going to overcorrect and I can buy something comparable cheaper, later. Why would I pay more, sooner if I don't have to and if my rental house is perfectly comfortable during the interim?

And, I will especially use both fundamentals and timing on a purchase of any rental properties (if my gold mining company mutual funds come through big time!).

Submitted by SD Realtor on September 5, 2006 - 10:39pm.

Very nice post. jg I think that anytime you can present raw data to a speculative subject then I feel you are adding levity and value. While I do agree with the premise that trying to find the exact bottom is tough, I think that your basics are very strong. Waiting for the total number of sales to level off makes sense to all of us. The unknown to all of us (or at least myself) is the lag time for pricing. As far as I know your post is the first one I have seen that attempts to use the hard data.

My non scientific method for purchasing the much needed larger home that my family is pressing me to buy will be as follows:

- wait for total sales to level off and stop decreasing

- wait for inventory to level off and stop increasing

- pray I can hold my family off until then

While I know I may be taking some pricing risk I think the downside will be seriously minimized if the first two factors are in place.

Submitted by CAwireman on September 5, 2006 - 10:59pm.

You apparently have access to data I don't have.

Thanks much for the post. I agree with you and am asking my self the same question: When will the market bottom out
so that I don't feel like I'm paying top dollar for a property (which I did before and don't want to do again).

I'll definitely look for the graph.

Thanks again.

Submitted by powayseller on September 5, 2006 - 11:02pm.

I admire Campbell for being one of the only people to write about selling real estate. You can find hundreds of books about buying RE, but how many are about selling it?

He writes about the cycles of real estate from his own experience, since he lived through 3 cycles himself. Campbell emphasises timing is more important thatn location.

Unfortunately, his index is not quite reliable, for it generates sell/buy signals at wrong times, and did not generate a buy signal for the current bubble until January 2003. He uses what he calls 5 Vital Signs: housing sales, building permits, loan defaults, foreclosures, and interest rates (weakest correlation).

He doesn't use the metrics I discussed before, and doesn't even mention them, so I don't know if he tried them and found them useless or just didn't have access to them.

He makes a convincing case of real estate's cyclical nature, and by the time you finish the book, you realize that real estate cycles are as much a fact as night and day, winter and summer, rain and sun. You realize that people do follow the crowd, and that you make money by going against the crowd. Timing, not location, is the most important.

Campbell is certainly a trend setter in a field usually filled with rah-rah cheerleaders.

Submitted by sdrealtor on September 5, 2006 - 11:23pm.

While timing is very important (maybe most important) dont belittle the value of location. I have a client that unloaded a piece of RE in Summer 2004 which is down at least 10% if not more. He now owns something walking distance to the beach that is up at least 15% (if not 20%) since he bought it in the Summer of 2004. Very few people that bought in 2004 can claim that. He'd say location is pretty darn important also.

With that said, any idiot could have bought any piece of RE in 2000 and made a killing within 4 years. Even a lousy property with a lousy location. Those idiots make a pretty compelling case for timing.

Submitted by carlislematthew on September 6, 2006 - 7:39am.

Unfortunately, his index is not quite reliable, for it generates sell/buy signals at wrong times, and did not generate a buy signal for the current bubble until January 2003.

Is this a national "buy signal" or is this for the San Diego market specifically? I ask because I bought a property in Seattle in Sep of 2003 and sold in Sep of 2005. Two years, and price of house went up about 30%. In-laws bought in Dec of 2003 and sold early this year. Their house went up about 40%, although they did spend about 5-10K in upgrades (windows, etc).

Submitted by jg on September 6, 2006 - 7:51am.

When I attempt to upload my Excel graph, I get the error message, "Uploaded file is not a valid image." What to do?

Submitted by speedingpullet on September 6, 2006 - 8:35am.

Probably because the Excel file is saved as an *.xls file.

If you have Word, import the file into a new document and save it as a *.doc file. Word and Excel are sister software packages, and so support each other's file types.

Submitted by jg on September 6, 2006 - 10:37am.

Thanks for the tip, SP, but that didn't work for me: I pasted the chart as a picture in Word, saved the document as a Word document, and attempted to upload it; no joy.

Submitted by ocbuyer on September 6, 2006 - 11:06am.

Interesting discussion. I feel that being comfortable with a "buy range" or overall bottoming out or leveling off is what people should try to predict.

Here's why - they need to know how much cash they should have on hand. I think that's going to be the toughest part of this downturn. Getting cash.

Laugh if you want, say "duh, that's the toughest part of ANY recession." But i want you to think about this:

Credit Debt = all time high
Home Ownership = all time high
Cost of Living = all time high
Consumer Expectations = all time high
Graduate Debt = all time high

The institutional lenders are going to get their asses kicked by the international markets and the government regulators (while losing their tail on hard money RE investments too). At the same time the interest rates will begin to rise again to stave off inflation. All the while, people own more homes, have more debt and are less liquid than they were during the last recession.

(BTW If you want a real comparison to what we have now i.e. gas and economic climate, international relations, etc. Look to the 70's and see what sort of mess we became when the gvmt. got it's hands involved. which consumers will demand again)

CIF as my pops calls it Cash In Fist - will rule the day.

Will you be ready?

Submitted by PerryChase on September 6, 2006 - 12:45pm.

jp, you need to create a .jpg file out of graph. Do alt-PrtScn then paste onto an image application, save the file at .jpg then upload.

Submitted by no_such_reality on September 6, 2006 - 1:14pm.

The institutional lenders are going to get their asses kicked by the international markets and the government regulators

Really, right now, the international market can't buy enough subprime mortgages.

Submitted by rocketman on September 6, 2006 - 1:45pm.

I prefer to use Snagit by Techsmith Link for transfering any image off your desktop to a .jpg file (charts, excel files, etc). I've used it here on several posts. You can even capture video. I recommend it very highly. There is a Trial copy you can download.

Submitted by powayseller on September 6, 2006 - 1:57pm.

The charts I referred to in Campbell's book were for the San Diego market.

Submitted by jg on September 6, 2006 - 3:01pm.

Very helpful, Perry and Rocketman; thanks!

Very helpful, Perry and Rocketman! Here's the belated graph. I look forward to your read of the data. Pink is annual sales volume of San Diego resale homes from Robert Campbell's book. Blue is the OFHEO price index data for San Diego homes.

Submitted by technovelist on September 6, 2006 - 3:10pm.

As the guy who was falling off a tall building said when he passed the 30th floor, "Everything's fine so far."

Submitted by powayseller on September 6, 2006 - 3:30pm.

This is what I think: Campbell is missing the inventory part of this. He's got sales, he's got price, but he doesn't have inventory. So he generates a buy signal in January 1996, but then generates a false Sell signal in 12/01, which turns again into a Buy signal in January 03. However, I can't criticize his approach too much, because it's the only published buy-sell method I've seen, so he sure gets credit for coming up with a system.

I'm surprised the OFHEO price index is fairly flat. In Campbell's book, he says prices rose 150% from 1982 to 1990, and fell 30-40% untkil 1996. But the blue line doesn't show that.

Why do sales decrease from 1999 to 2001? This surprises me,because I bought my house during that time, and the market was HOT. People were making offers on homes without even having seen them. Realtors had to check the MLS every hour. Once, my realtor called me to tell me about a house that had just been listed a few minutes before. I rushed over to see it, and as I arrived, darn it, a couple was already there. I asked them if they were looking at it, and if I could join them and see it too "Oh no, we've already made an offer", they said. That is how it was. Houses were sold within hours of being on the MLS. The house we ended up buying had just been put on the MLS a few hours prior to our offer, and there were several backup offers going late into a Friday evening. It was absolutely crazy back then.

A friend told me the night he listed his house, a couple came over at midnight and woke him, to make an offer on his house; they had not even seen it. He sent them away, and called his agent at 8am ruefully the next morning. "Don't worry," said the agent, "I've got 4 other people on their way over to look at it". So how could sales be down in that time period?

Submitted by jg on September 6, 2006 - 7:28pm.

I put a call into DataQuick to find out the price to purchase:

1. Monthly sales of resale homes and median prices of such for San Diego county.
2. Monthly sales of resale homes and median prices of such for La Jolla.

If it's not too expensive, I'll buy the data and rerun the graphs.

Then, I'm going to attempt to use my dated experience with times-series analysis to see if regressing price on Campbell's dependent variables leads to a simpler model. Graphing some of Campbell's data, it's hard to believe that notices of disclosure and foreclosure sales are not colinear, and that you can use one or the other, instead of both.

Actually, the best approach would be to turn the data over to a UCSD Ph.D. (it's a world-class econometrics school) for analysis. Anybody have any good contacts to a good UCSD econometrician?

Submitted by sdduuuude on September 6, 2006 - 7:49pm.

All this talk of timing and cyclical markets led my mind to this graph, posted in another link:

Historical Prices.

While I see two cyclical events in the 80s and 90s, then major wierdness in 2001, I'm not so sure I see a cyclical pattern here that I can assume will continue perpetually.

Trying to predict the bottom of something that isn't consistently cyclical is a bit of a challenge, methinks.

Look at Japan for the last 13 years, for example, after their last crash. Maybe we will see more stability after the coming bubble deflation, and there won't be a stark "bottom."

Maybe SD data is more cyclical, if one looks back more than two cycles, but I have only seen two cycles plotted.


Submitted by jg on September 6, 2006 - 7:59pm.

PS, the OFHEO data has its weaknesses, as you know: it only tracks the prices of resale homes, not new homes, and is limited to homes financed via conforming mortgages, i.e., relatively inexpensive homes (it does not track homes financed via jumbo mortgages). However, the OFHEO series is particularly valuable as it gives a clean index of price changes, and is less susceptible to reflecting price increases due solely to increases in sizes of homes in recent years.

I'm not sure what price data Campbell used, as he didn't reveal such in his book, per memory; I assume that he used DataQuick median sales price, not the OFHEO price index. And, his data is monthly; I used annual data, which smooths the peaks and valleys (but eliminates the horrid seasonality of the monthly data). So, that could account for the differences.

Submitted by powayseller on September 6, 2006 - 10:08pm.

OFHEO data ignores the tremendous amount of dollars put into kitchen remodels, additions, pools and backyard barbeques, and myriad other contracting work on the house. A kitchen remodel typically costs $60K, so it could increase the value of a $600K house by 10%. Thus, the OFHEO data is somewhat overstating the changes in price.

Campbell used DataQuick for Existing Home Sales, Construction Research Industry Board for New Building Permits, the SD Cty Recorders Office for NODs and Foreclosure Sales. The appendix of his book lists this data from 1988 - 2004 in 4 tables. Table 5 is interest rates.

Submitted by Chris Johnston on September 6, 2006 - 10:18pm.

Chris Johnston

JG - I asked him if he had done any regression analysis and he told he that he had considered it but did not have to time to do it. I was mostly interested in how many std dev's some of these things in his model were from a regression line.

Submitted by SD Realtor on September 6, 2006 - 10:47pm.

jg -

If you go to the website they have information on ordering custom reports. Alternately every month dataquick posts the sales activity by zip code for the entire county. HOWEVER, the San Diego Union actually prints it all out every single month. If you go to a local library you can most likely get the tables for each month for however many years back you want to go. The only problem is that it will be a hardcopy not a soft copy.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.