More on Spotting Bubbles

In comments on my previous post on bubbles, John McGowan and others point out that you can use the price-to-rent ratio (or price-to-income) as an indicator of a housing bubble. I think this is a partial but not a perfect solution.

The value of a thing should be the net present value of the future cash flows from the thing. In experimental economics, they use securities with absolutely certain cash flow profiles, so when a bubble in prices appears, you have an objective measure of value to compare it to. With individual stocks, on the other hand, the P/E ratio could go up to 100, and you can back an implied growth rate of earnings out of that, but who’s to say the company won’t hit that growth rate? At that point it’s just your opinion against the market’s.

The question is whether housing is like an economics experiment or like stocks. I think it’s pretty clear that one house is more like an individual stock. You could look around at other houses you might rent, but houses tend to be somewhat unique. (This is also one reason why bubbles perpetuate themselves; as Bond Girl pointed out, people are perfectly able to believe that one part of the market is in a housing bubble, but the part they are buying in is not in a bubble.)

Now, as Leigh Caldwell pointed out, it should be easier to spot a bubble when looking at all stocks in aggregate. If the implied growth rate of earnings of the entire market is 20%, then that looks implausible. Similarly, if the price of houses in aggregate is twice the value implied by the rental income they could geenrate, that looks implausible.

I say this approach is partial but not perfect because current potential rental income is a poor proxy for the long-term value of a house. The fact is that most people buying houses aren’t going to be renting them out, and what they care about is the price at which they will be able to sell that house in 10 years, which depends in part on the price at which that buyer will be able to sell the house 10 years after that, and so on. And it’s entirely possible that the relationship between house prices and rents could change over that timeframe due to any number of economic factors – the homeownership ratio; shifts between the exurbs, suburbs, and cities; the aging of the population; and so on. There’s no axiom that says that these ratios – price-to-rent or price-to-income or price-to-earnings for that matter – have to remain constant over the long term. At the least, it’s generally possible to come up with a reasonable argument that things have changed this time. In part, that’s why we have bubbles, but it’s also why it’s hard to be objectively sure you’re in one.


Originally published at The Baseline Scenario and reproduced here with the author’s permission.