Lies, Damned Lies, and the Case-Shiller Index

After the 9:00 AM release of the Case-Shiller 20-city index of home prices, I observed the following three headlines (emphasis mine):

Home prices extend win streak in January: S&P/Case-Shiller – CNBC

US home prices dip in Jan. for 3rd straight month – Yahoo! Finance

Home Prices in 20 U.S. Cities Rose 13.2% in Year to January – Bloomberg

Well, which is it?   A win, a dip, or a rise?

No worries, though.   The Dow Jones Industrials took the news as good, and promptly shot up over 100 points.   The S&P 500, not to be outdone, tacked on over 10.

So what’s with the bi-polar market launching headlines?

I’ll tell you.   It’s my nemesis: the Seasonal Adjustment.

Seasonal Adjustment can be defined as a statistical technique designed to even out periodic swings in statistics or movements in supply and demand related to changing seasons.   More accurately, it is a fudge factor designed to manipulate actual data into something that really isn’t true but fits neatly into an economist’s outlook.   So, rather than actually measure the genuine price of homes sold – you know… the one that you (as an actual person) would get (in terms of actual money) when you sold (your actual home)… they introduce an additive or multiplier to the regression to come up with some other number.

And don’t get all Frisch-Waugh-Lovell on me.   I’m well aware that using projection matrices to make the explanatory variables orthogonal to each other will lead to the same results as running the regression with all non-orthogonal explanators included.   Duh.

But it still doesn’t reconcile with me that, rather than using facts, we use results generated by a unicorn working with its imaginary abacus in Xanadu.

So here are the facts.   The Not Seasonally Adjusted Case-Shiller month over month change in home prices was a decline of -0.08% in January.   They call these “Not Seasonally Adjusted” rather that “Real” numbers because, well, calling them “Not Seasonally Adjusted” psychologically anchors you to the concept that “Seasonally Adjusted” is the base case.

The Seasonally Adjusted Case-Shiller month over month change showed a rise of 0.85% — smashing the expectation of 0.60%.

Not Seasonally Adjusted numbers fell for the third consecutive month, while Seasonally Adjusted numbers posted their 24th consecutive gain.

Despite the nearly 1% unicorn wag between the NSA and SA numbers this month, the year-over-year Seasonally Adjusted Index still fell short of expectations; rising 13.24% versus expectations of 13.42%.   Maybe the unicorn accidentally transposed the last two digits.

Either way, you can clearly see that this is the most bullish indicator ever.                                       

As for the headlines from the infotainment sources cited at the beginning of this blog: If we haven’t learned to take the “facts” distributed by these gaggles of people who have enough spare time to nerble on all afternoon on poorly rated shows and websites with a GIANT grain of salt we’ve learned nothing at all.

I wonder if the headline scrubbing algos learned anything.

/rant off

– LL

images