Details on the Last Upturn

[admin]

[img_assist|nid=1604|title=
San Diego Resale Homes, ’88-’99: Sales and Prices|desc=|link=node|align=left|width=400|height=267]

[img_assist|nid=1605|title=San Diego Resale Homes, ’88-’99: Sales and NODs|desc=|link=node|align=left|width=400|height=267]

Now that we have some sense of the leading indicators — sales and notices of default (NODs) — I think that it makes sense to look in detail at the last upturn and downturn.

In the last upturn, sales peaked in June 1988 (3,860), and prices peaked 3 years, 2 months later, in August 1991 ($187,500). NODs troughed in November 1988 (474), and prices peaked 2 years, 9 momths later. Thus, peak sales and trough NODs gave approximately 3 years notice of peak prices.

How about this upturn? For resale homes, sales peaked in June 2004 (3,363) and prices peaked in May 2006 ($569,500), a lag of 1 year, 11 months. Thus, in this upturn, peak sales only gave 2 years notice of peak prices, unlike in the last upturn

NODs troughed in November 2003 (329) and prices peaked in May 2006, a lag of 2 years, 6 months. Thus, in this upturn, NODs gave similar advance notice of peak pricing as they did in the last upturn (2 years, 9 months).

[powayseller]

Wow, I am impressed. You’ve just proven that real estate cycles can be predicted. Good stuff!

Let’s remember that median prices lag by 2 years, so in reality your model is only lagging by 1 year on the way up. This would be one more indicator that we can use to confirm the top or trough.

I recall prior posts where we discussed whether real estate cycles can be predicted, and some people insisted that real estate troughs cannot be predicted, since there would be false bounces/rallies.

[SDbear]

Awesome analysis. What was the correlation at the bottom? How much lag did we see between sales and prices?
I hope it is easy for you extend your plot.
Thanks in advance.

[Anonymous]

Did prices really peak in May of 2006? Only the median price did and I believe we have already established that median prices are really lagging the market by 12-24 months depending on the cycle. In reality, the mean price in a given neighborhood was falling sometime between late summer 2004 and early 2005. If we believe that prices did in fact peak in late 2004, how would that affect your analysis?

[CAwireman]

Stupid Question I have to ask…

JG, excellent graphs by the way!

But, if median (despite it not being a great measure of RE
value) was largely flat from 91 to 96, what data suggests as us Piggingtons believe, that prices will drop after this current bubble runs its course?

I know this is blasphemy on this blog, but I had to ask…

[(former)FormerSanDiegan]

But, if median (despite it not being a great measure of RE
value) was largely flat from 91 to 96

The median was not flat, it actually declined by ~ 8-10%. It just looks flat because of the large (2x to 3x) run-ups during the boom times.

[studenteconomist]

One of the interesting conclusions from this data is that the previous runup (1980’s and pre-internet) took over 3 years to go from peak sales to peak prices. This time, with the internet blogs such as this one and electronic tracking, the public (especially the RE watching public) is much better informed and the time from peak sales to peak price was under 2 years. Better data collecting and more rapid dissemination of that data has resulted in a 14 month drop in the lag time between the variables!! This is huge and will have important ramifications for the timing of the bottom of the market. Instead of a 6-7 year wait, we may only have a 4-5 year peak appreciation to bottom of the market. That is why I feel (and now have data to support) that this downturn will be quicker and more dramatic than the last one (and may even hit bottom in the next 12-18 months. Faster (and better) information leads to faster market revaluation and leads to faster market correction.
So keep up the posts Piggingtons, you are helping the RE industry reach toward a more rational and correct market pricing system.

[greekfire]

If you check out NAR’s website, and look under David Lereah’s presentations, he states that the peak was in August 2005. Whether this is peak sales or prices, I am not sure.

[Anonymous]

SDbear, I put details on the bottom under the topic, “Details on the Last Downturn.”

[Anonymous]

Also, in the spirit of full disclosure, the source of my data series on NODs is from Robert Campbell’s book, which only had data through Nov. ’03. So, NODs may (probably, with the ridiculously low rates/easy credit terms?) have trended even lower, troughing in ’04 or ’05.

I need to get monthly/quarterly NOD data from ’88 to present. Once I do, I’ll revise the findings on trough NODs lead to peak prices.

[Anonymous]

Malfred, I only have data on median (not mean) prices for resale (not new) homes (no condos). I could only guess, and I hate guessing, what would happen if I used average, instead of median, prices.

[Anonymous]

Fair enough, boss. I do not actually come with charts and figures to back up my assumptions but I remain curious.

[Anonymous]

Thoughtful observations, studenteconomist.

Are you a Ph.D. candidate? Which school?

[avidsaver]

cawireman, thanks for asking about the “flat” median prices because I had the same “stupid” question. Even at a 8% – 10% decline in the current housing market, I’m basically screwed. Do you all (yeah, bad grammar, I know) think that there will really be a much larger decline this time around?

[CAwireman]

Avidsaver, yes, that’s the overall impression I got.

I still think that we’ll see greater drops than 8 – 10 % this time round. But if we only went with the prior down turn percentage we’re already bottomed out in certain zipcodes.

[Author]

Posts