This chart comes from a new paper by Karl Case and Robert Shiller, looking at the results of a survey they’ve been handing out to homebuyers annually since 2003. The idea is a very smart one: if you want to get an idea of the behavioral economics of homebuyers, the best way to understand what they’re thinking is to simply ask them.
And this chart, in particular, is both very elegant and very informative. It’s elegant because you have a very close maturity match: the average duration of a US mortgage, before it’s refinanced or the house is sold, is about 7.5 years, which is close to the ten-year horizon in this question, which Case and Shiller ask every year:
On average over the next 10 years, how much do you expect the value of your property to change each year?
Now the number of homebuyers in America vastly exceeds the number of people who understand the mechanics of compound interest. If you asked instead “how much do you think your home will be worth in ten years”, and then presented that answer as an annualized percentage increase, I suspect that the answers — especially in the peak years of 2004 and 2005 — would be substantially lower. (Put it this way: if you bought a $260,000 home in 2004 and expected its value to rise at 12% a year for 10 years, then by 2014, you’re saying, it would be worth more than $1 million. I suspect the number of people answering 12% or more is going to be greater than the number of people who think the value of their home will quadruple in ten years.)
Still, that’s not particularly important, especially since the question has remained the same for the past decade: the trend here is real. And what’s fascinating is that the big fall in expected long-term home-price appreciation happened before the financial crisis, and that the crisis is actually completely invisible in this chart: expectations continued to deteriorate long after it was over.
And even given the fact that homeowners tend to overestimate annualized percentage returns over 10-year horizons, we’re now at the point at which the expected rise in home values barely exceeds today’s record-low mortgage rates. Over the long term, homebuyers still think it’s a good idea to buy a house. And they might be right about that. But they’re not buying because they think they’ll make a handy profit in ten years’ time.
Which brings me to one of the central themes of the Case-Shiller paper: the idea of a “speculative bubble”. If you look at the situation in the chart circa 2004-5, there was a huge gap between the cost of funds and the long-term expected return. And if people really believed house prices were going to rise that much in future, it made all the sense in the world to lever up, get the biggest mortgage they could find, and buy lots and lots of house. After all, the more levered you are, and the more house you buy, the more money you make.
Case and Shiller have a handy definition of a speculative bubble, in this paper: it’s a bubble with “prices driven up by greed and excessive speculation”. But here’s the thing: people don’t speculate on a ten-year time horizon, and the producers of “Flip This House” weren’t waiting around to see what properties would end up being worth once the kids had gone off to college. A truly speculative bubble, it seems to me, is a function much more of short-term house-price expectations than it is of long-term expectations. If you think you can buy a house today, sell it in a few months’ time, and make tens of thousands of dollars doing so, and if you intend to do precisely that, then you’re clearly part of a speculative bubble. But it turns out that home buyers were actually surprisingly modest in their expectations of one-year price increases — they expected prices to rise less than they ended up rising in reality.
On the other hand, if you buy a house now in the expectation that it’s going to increase in value substantially over the next decade, you might be a buy-and-hold investor, but it’s hard to characterize what you’re doing as speculation.
I’ve been disagreeing with Shiller on the subject of speculative bubbles for five years now, but I think this is important: just because you have a bubble, doesn’t mean you have a speculative bubble. The dot-com bubble was speculative; the rise in house prices in 2000 was not. There was a speculative bubble in Miami condos; there was not a speculative bubble in Manhattan co-ops. If you buy because prices are rising, that might be because you want to flip your property and make money — or it might equally be because you worry that if you don’t buy now, prices are going to run away from you, and you’ll be forced to move out of the neighborhood you love because you can’t afford it any more. It’s still a bubble, but it’s more of a fear bubble than a greed bubble.
Still, bubbles are bad things, and they’re liable to burst either way. And so I take solace in this chart, because it shows me that people are buying, these days, for the right reason — which has nothing to do with expectations of future house prices, and everything to do with simply paying a fair price for the shelter they’re consuming. House prices might not rise much over the next decade. But if they fail to rise, today’s house buyers aren’t going to be disappointed: they will still have lived in their homes while paying a perfectly reasonable sum to do so. Which is a much better state of affairs than bubble-and-bust.
- Chart of the day, housing bubble edition (blogs.reuters.com)
- CR: Shiller on House Prices (woodonfire.wordpress.com)
- Case-Shiller: House Prices increased 1.2% year-over-year in July (woodonfire.wordpress.com)