patrick.net

I get a kick out of how so many people price houses at $499K, $599K, etc. It's clearly visible in the price distribution graph.

Met too, clearly this has nothing to do with the real value of the house.

Why does the time series oscillate up and down?

Just noise I think.

I can see a time say in 2013-15 that the density of that chart moves to the left (lower is I am reading it correctly) about 3/4 of inch.

Have you looked at data smoothing for either or both of these plots?

Bucketing price data into 10k or 25k blocks should remove a fair amount of the small-range variability.

I'd think that 7-14 day smoothing of the listing prices should also make trends much more visible (generally some multiple of a week is best for various forms of economic data).

As far as Case-Schiller -- does it take into account changes in prices for equivalent housing? If the underlying housing mix changes, the aggregate sales price will move even if the price change for components of that mix is doing something different. E.g.: if more large houses are listed, but at a cheaper per-square-foot cost, the aggregate listing price will increase, though the actual market trend is down.

A very close comp to my relatives' house just sold for $200k less than what they paid near the peak in October 2006. This represents 21% as their original purchase price was $950k. They've rented money from a bank for four years now, and are down approx. $200k (which happens to match their down payment, which was partially borrowed from the mother-in-law). This is ugly all the way around. $950k was way too much to pay for their house (in a highly desired peninsula zip code) and was entirely more than they could afford. Are they in default? No. Will they end up losing their house some day? Maybe. But they are slaves to their very large and poor financial decision. All of this because they "had to own" and tried to keep up with the Joneses.

karsten says

As far as Case-Schiller — does it take into account changes in prices for equivalent housing? If the underlying housing mix changes, the aggregate sales price will move even if the price change for components of that mix is doing something different.

Welcome to Karsten. The methodology for Case-Shiller is documentet in

http://www2.standardandpoors.com/spf/pdf/index/SPCS_MetroArea_HomePrices_Methodology.pdf

Case-Shiller is an INDEX (not an average or median or anything selling price), and is much less affected by mix than medians and averages because of the specific methodology used. Most importantly:

(1) It looks at repeat sales (meaning each house sold is compare with it's own previous sales)

(2) It has 3 tiers low-med-high to account for possible different general rates of change in different tiers of the market.

There is much more detail in the above referenced pdf file, enjoy,

shrekgrinch says

The Graph of the Denial Pushers.

Yes, let us not forget that these are ASKING/WISHING prices and not closing prices.

Wasn't it also publicized just a few weeks ago that the numbers may be being nontrivially skewed upward by recent misreporting of sales prices upward to include, e.g., money subsequently kicked back to the buyer, and/or misreporting sales at asking price that were in fact much lower?

JimAtLaw says

Wasn’t it also publicized just a few weeks ago that the numbers may be being nontrivially skewed upward by recent misreporting of sales prices upward to include, e.g., money subsequently kicked back to the buyer, and/or misreporting sales at asking price that were in fact much lower?

Good Point, which is often overlooked.

APOCALYPSEFUCK says

Look. Look at that graph. It spells APOCALYPSE.
Plant potatoes.

U should change your display name to "Plantpotatoes" :)