Legendary investor Jeremy Grantham recently had his research staff compile info on all the historical stock, currency, and commodity bubbles they could find. A "bubble" was defined here as a two-sigma event: an asset price that got more than two standard deviations away from its historical trend. Grantham and company identified 28 financial bubbles, of which every one—every single one—saw prices revert back to the mean.Since I already had the base data shown here, I decided to perform the professor's analysis for Marin County (see that previous post for an explanation of how this data was collected and calculated). I normalized all of the Marin data to that of 1969 and calculated a best-fitting trend line (green) to the data minima:
(Click on the image for a larger view.)I then calculated the percent deviation of each of the blue data points in the above graph from the base trend line. These percent deviations are shown in the following graph (along with the average, and both the first and second standard deviations):
(Click on the image for a larger view.)What does this mean? If Jeremy Grantham and his team are correct, it probably means no so-called "soft landing" for Marin County.
And in the words of the professor:
Well, we'll see... Based on the data shown here, things aren't as extreme in Marin as they seem to be in San Diego.Of course, there is more to any story than deviations from the mean, but the point here is that history is telling us something. People can make all the excuses and rationalizations they want, but the fact is that markets revert to the mean. Bubbles burst.
People continue to tell themselves (and anyone else who will listen) that San Diego [and Marin County] real estate will be the history's first asset to rise so far, so fast, and never come back down to earth. They will eventually be relieved of this misconception. For now, though, faith in the Soft Landing holds sway.
11 comments:
"Church of the Soft Landing"
It does appear some adherents may be facing economics lessons in the near future.
Good for them.
Great data again, as usual - and the more often that facts and figures like this are available, all the better.
Thank you.
Isn't Califorina different?
The only difference I see in California is that real estates prices are more vulnerable for big correction than prices in other inland states. The higher it climbs, the deeper the fall it would encounter. Eventually there will be self-correcting process.
If the previous two cycles repeat, then it would seem that we have another 5-7 years to wait before prices return to the mean. That qualifies as a soft landing in my book.
Anonymous said...
If the previous two cycles repeat, then it would seem that we have another 5-7 years to wait before prices return to the mean.
Question is whether prices would fall below the mean by 2 standard deviations?
There are differences between this runup in prices and previous runups as well as the expectatins driving them. Whether those differences will matter is anyone's guess as there is not much history to go by. And for me, "bust" or "pop" or whatever does not mean all in one year or all in a few months.
is it me or a couple of months ago most of the realtor-speak focused on this being a "balanced" market, and now we have gone from that "balanced" market to a "soft landing".
Maybe in another few months... "accelerated soft landing"?
Anyone see a correlation between real estate and presidential political party? I have an unproven theory that "real" assets such as oil, and real estate perform well during Republican president reign, while technology companies do better during Democratic period.
During Regan's time, real estate investment did quite well. Later real estate had a sever correction, especially in Southern California (lots layoffs at defense companies) when Clinton became the president because hot money was transferred and invested into technology stocks.
If this theory is proven correct, watch out for the next presidential election in 2008.
So as bettor the risk/reward favors selling to buying.
In order to sell there must be a buyer. Let's just hope they have enough sense to buy low.
Grantham has another report that he published in Oct. Worth a read.
Do you have the reference (preferably a URL)? Thanks.
Your choice of data minima as the basis upon which to fit a line you call the historical trend is odd. A more reasoned choice for the historical trend would be a linear fit to all the data available. Then the dataset does not exhibit the 2sigma bubble property.
will,
That's not actually true because although the variation will be half, so are the standard deviations. Anyway, just to put the issue to rest, here is a URL to the analysis you suggested:
http://tinyurl.com/9tzkh
Keep in mind what I published is a replication of the other experts analyses. I wanted to be able to compare results for Marin to those of other areas.
Post a Comment