This chart appeared in an article written by Doug Kass on TheStreet.com.Doug is a short seller and so tends to be more of a glass half empty guy.
What I find interesting is what appears to be absolutely no correlation at the left of the chart, evolving into a very tight correlation.
So the question then becomes are we back to no correlation? Kass believes the correlation will stick.
10 comments:
Well I do think that we are in for a decline but I sure wouldn't base my view on that kind of analysis.
I don't think he is basing his entire analysis on this.
I take it as an interesting nugget of information.
Odd. What could explain that? Homebuilders are a pretty small part of the stock market. They couldn't be tugging the S&P along with them.
Well, different things (more than just sectors) rotate in and out of importance over time. This has repeated over and over.
For example the NFP used to be less important. There was one huge number that printed in what I think was 1994 that seemed to give the number more visibility which it has maintained.
In the late 1990s analysts upgrades and down grades were far more important than they are today.
The same can be true of sectors and sub-sectors. Things related to housing have been very important in the last few years. How important could be debated but the visibility for this part of the economy is more than it used to be.
The question posed in the chart, I think, is will that continue as is or will it diminish.
Sectors do rotate in and out of favor and, lord knows, that has been true in spades for the US equity markets the past couple cycles to the point where I think the argument that we have actually been in a rolling, secular bear with embedded bull cycles since 1998 has some credence.
But whether one buys that or not one fairly simple hypothesis might be that most home building in the past was done by private contractors and that the growth of large, publicly held home builders is of more recent vintage; that is, they were such a tiny piece of the market before the mid-90's they were largely off investors' radar screens. Don't know how true that actually is but it seems plausible.
Assuming it is plausible then we could make that hypothesis a bit more complex by adding that mortgage lending and securitizing has become a bigger market so the home builders may have become more closely correlated with financial stocks which are a far more significant market segment. The inverse of that is that securitized mortgages are part of the bond market, presumably in some competition with equities for investment dollars but there I confess I actually sold a bond fund because it had a significant percentage of MBS's and I think there are too many marginal loans embedded in those these days.
Regardless, add the home improvement, metal and lumber suppliers, and other related segments and the home builders could be a reasonable proxy for enough market sectors to suggest the correlation will continue although I suspect it will weaken as the normal cycle moves on, different sectors take over, and demand for equities arises from other sources.
Still, with 'only' four+ cycles in view on the chart (as I read it) I'm not sure sample size is sufficient to take a strong stance one way or the other. It's interesting though and quite suggestive given other analyses I've read that indicate real-estate driven market cycles tend to have poorer (and more drawn out) endings than cycles driven by other market forces.
RW
If what you say is true then the change in the market place that you cite from a few years ago means that there is not sufficient data to to craft a forward looking analysis on whether the corelation sticks or breaks?
Roger, I do think the sample size is too small to make a firm bet on the continuance of strong correlation but that uncertainty is compounded by my inability to tease the secular and cyclical factors apart. That is, (a) has a larger publicly held home builder sector become a reasonable proxy for cyclicals, late cyclicals in particular, as a simple fact of interrelationship or (b) does the fact that real-estate and home equity have been important drivers in equity markets the last couple cycles strengthened what would otherwise be a weaker correlation of the home builders to the overall market.
Since the S&P500 is a capital weighted index I would expect it to reflect both factors but if the degree to which it does that is more dependent upon (b) than (a) then, assuming the next cycle is not RE driven, I would expect the correlation to weaken.
I don't have enough data to push the logic much further than that.
I think that it is just a coincidence that in the recent period the other variables involved in the model just happened to cancel each other out so that there is apparently a tight correlation.
PS - I didn't notice before but very interested that there does seem to be a 4 year cycle in that housing series!
Part of the de-linking in the late 80's was the change in the Tax`Code towards second homes ... That was a huge blow to housing and the collapse ensued right to the early 1990's
Post a Comment