Politics, Technology, and Language

If thought corrupts language, language can also corrupt thought — George Orwell

The housing market: everything you know is wrong

Posted by metaphorical on 4 March 2007

The power of graphical representations to drive complex relationships straight into the understanding is often nothing short of amazing. I’ve mentioned Strange Maps before; last month’s selection included a terrific map showing right-handed vs left-handed driving. Almost in passing it gives a U.N.-level representation of the decline of the British Empire and the coincident American Century.


Dark blue: drives on left (mainly British ex-colonies).
Light blue: used to drive on right, now on left (Namibia).
Purple: used to have mixed system, now drives on right.
Light red: used to drive on left, now on right.
Dark red: drives on right.

The countries still in blue are all former colonies, as are a couple of the purples—most prominently Canada, which I didn’t know ever drove on the left, though after thinking about it for a moment; of course they did! And it certainly makes you want to run to Wikipedia and see what the hell Namibia’s story is. (Apparently it started out as a German colony and ended up under South Africa’s control after WWI. And indeed, Wikipedia gives 1918 as the year it started driving on the left.)

Not all graphical representations are maps. Some are, well, graphs.

For about a week I’ve been looking at “A History of Home Values.” In a single wavy line, it calls into question most of what I thought I knew about residential real estate as an investment.

For example, “A home is a good investment.” Uh, no. It wasn’t for most years between 1890 and the end of WWII. Since then, it’s kept up with inflation, but every spike has been met with a decline. I always thought the effect of the up and down was always a net up—not so until then early 1990s. Then, home prices started a steep upward movement that never wavered.

Looking at the graph you see that the current housing boom is so unconditionally aberrantly anomalous that anything you thought you know about the market is surely wrong, or at least, unsupported by anything like historical antecedent.

6 Responses to “The housing market: everything you know is wrong”

  1. JoAnne said

    The graph does use the “lying with statistics” trick of starting the graph’s vertical axis at 60 instead of 0.

    If it were it would look like this instead:

  2. Good catch, but does it matter a lot here? 100 is a baseline, so it could have been zero, and then everything below the line would be a negative number. And then there wouldn’t be any harm in omitting the yet-more-negative values that are below the lowest point in the graph.

  3. Another thing I get is that “housing prices are always anomalous”.

    Some observations:

    – In the 100-year period before the big housing boom, prices generally declined for the first 50 years, and generally rose for the last 50 (there are relatively few years between the baseline and the end of WWII in which prices are above the starting point; there are no years after WWII in which they were below it). Something fundamental about the market changed in that period; prices don’t track inflation before the war, or even before the Depression; they generally do thereafter.

    – Almost a third of the timeline (37 out of 116 years) evidences sharp booms or busts attributed to anomalies (WWI, Depression, WWII, the “booms”). But something that happens one-third of the time is not an anomaly – it’s a normal condition.

    – There is only one extended period – 1950-1970 – in which the market average stays within 10% of its starting value. There are 12-15 times in which it gained or lost over 10% of the average price in just one or two years – not counting the post-90s boom; there are several points at which the average price rose or fell by a third in that time. Although average housing prices showed a net increase of only 10% from 1890 to 1998, they swing from about 25% above to about 33% below baseline repeatedly during that period. Thus, housing prices are vastly more volatile than the long-term growth rate would indicate.

    – The current boom has shown an average price increase of about 80% in 9 or 10 years; the post-1942 boom showed a price increase of over 60% in just 4 years, and about 70% in 10. The current boom is thus slower than that already experienced once (though it’s true it is longer-lasting, and the previous one resulted from a sudden population spurt in a depressed market).

    – Most of the increase even in the post-war period comes from “booms”; the post-boom retractions in each case reach to within a few percent of the starting price. In the majority of the post-war years up to 1998, buying a house and re-selling it within 3-5 years would bring a loss, not a profit; if you bought a house in the mid 50s and wanted to sell it at a profit, there are only 9 years, in two brief windows, before the turn of the century in which you could break even after inflation. Profiting from the “rising” housing market depends almost entirely on “market timing” – picking the exact moment in a fairly narrow window in which to sell; a “buy and hold” strategy, predicated on the assumption that the rising market will make most purchases profitable at almost any time thereafter, is unworkable. Housing is thus more like speculative short-term stock trading than it is “value” investing.

    – The conclusion of “little net long-term price increase” is partly an artifact of the timeline: the graph appears to show very moderate net price gains over the first 100 years, but only because of the precise basline year chosen. If the graph had begun at 1894 instead of 1890, it would show almost unbroken losses through WWII, and a total net loss over 100 years; if it had begun at 1891 instead of 1890, it would show generally flat prices until WWI, then lower, but flat prices until WWII, and a very considerable net gain over 100 years. So which year is the outlier: 1890, 1891, or 1894? The short-term volatility of the graph makes even the notion of a “baseline” problematic.

    A couple of summary conclusions:

    – There is no way to characterize “the housing market”. Housing as a product and a part of our lifestyles has changed hugely over the years; even though prices have been stable over the period shown in the graph, they are not stable over any likely period during which someone is going to hold and then sell a house; short-term volatility vastly overrides long-term stability; “anomalous” events have significant impact on prices – and impact which can shift the market baseline for decades at a time – but occur almost every generation. In short, we just can’t predict what the market is going to do on the basis of the last 100 years.

    – The post-war “anomalies” have all been “booms” – which is just another way of saying “the prices went up”. Only the 1940s boom seems to have permanently shifted prices; the next two lost as much value as they created, and we don’t know what’s going to happen to the current one. And the 1940s boom is the only one tied to an obvious “anomaly”: WWII. So the market since 1975 appears not only to be highly volatile but inexplicably volatile – or at least it shows high volatility in reaction to events nothing like the world-shaking anomalies that triggered booms and busts up through WWII.

    – The “moving baseline window” – the fact that the net trend changes so dramatically depending upon which specific 10-year, or 20-year, or 50-year, or 100-year window you look at – coupled with the huge differences in how people have lived over those periods, makes meaningful long-term comparisons rather questionable.

    – Given the trend toward bigger houses, more amenities, and better fuel efficiency, and the almost-stagnant trend in adjusted prices, people are getting better deals on houses, and better houses in general. But that has to be its own reward, from the buyer’s perspective; profit appears to be a risky motivation for buying.

    – The current boom could be (though you cannot say from this graph) the result of an anomaly of a new type: the fact that all the good places to live are now taken. During the WWII housing boom, big cities and their suburbs were the places that were booming. Now, there simply isn’t any room for added housing in New York, LA, San Francisco, etc., and their surrounding megalopolises are pretty much wall-to-wall also. There is cheap land in the interior, but . . . there’s a reason they call it “flyover country”. So prices are screaming in the impacted markets where affluent people want to live – as market theory tells us they will – and that can’t change, because it’s physically impossible to create more of the product in demand (single-family homes on separate lots).

  4. […] in March, Meta had a post, “The housing market: everything you know is wrong” (by which he meant, everything he thought he knew was wrong) that looked at a surprising graph […]

  5. Maria said

    The facts are simple, where there is money to be spent there will be housing to be bought.
    As long as a quick profit is not the motive to purchase now is the time to buy.

  6. Mikeharvey said

    Hey, from Toronto, Canada

    Just a quick hello from as I’m new to the board. I’ve seen some interesting posts so far.

    To be honest I’m new to forums and computers in general :)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: