In the comment section of my previous post on housing, my friend Linus speculates on whether or not his area (Jackson, WY) is in a bubble:

Whether or not people think there's a housing bubble, I've noticed that the vast majority of people think it won't hit their area. I've always thought that a place like Jackson Hole will be somewhat safe because, questionable loans or not, demand for housing here will never really decrease. You think that Houston will be fine because fewer people are signing bad mortgages there.

My friend Keith and I both find ourselves in rental situations because we truly can't afford houses in the valley and can't fathom commuting 50 miles over a mountain pass every day. We've both calculated it out and it makes more sense for us to rent unless we were planning on staying for at least 7 years (maybe more with talk of a slowdown in appreciation).

Bubbles usually occur on the backs of investors. These investors may be developers that are building neighborhoods or people hoping to flip houses for a quick buck. In the latter case, people holding on to a house that they don't wish to live in do wish to make as much money off of it as they can before they sell it. So what do they do? They rent it out. So what can you expect to see in cities with abnormally large numbers of investment properties? Houses and condos for rent and a lot of them. Since in this case supply (of houses and condos for rent) does not have much to do with demand (for rented lodging), and making money is not the primary motive and therefore they can be rented out at a loss, you can expect places these places -- with a large percentage of investment properties and an excess of rentable lodging -- to have comparatively low rent.

We would figure that rent in Los Angeles would be anything but cheap, but if a housing bubble exists there then it should be artificially low. So how can you determine whether rent is lower or higher than it should be? We compare it against markets in which we can be relatively sure there is no housing bubble. If we compare the cost-of-living differential of owning compared to renting, we should be able to determine whether or not housing prices are artificially high.

Lukckilly, there are tools that can tell you cost of living differences for owning and renting between two cities. I chose Casper, Wyoming, and Tyler, Texas as two places that it is unlikely that any significant bubble exists. None of the three are experiencing tremendous growth (Tyler grew 16% between 1990 and 2000, Casper grew 8%, and Lake Charles grew 9%), none have significant renter-friendly college populations, and none are near enough to a growing metropolis to be feeding off of that growth.

Keep in mind that we're looking at relative values here. There are a lot of other factors that go in to the cost of living, but those factors are controlled because they are the same for owning or renting. However, the actual differences (either in dollar amount or percentages) don't tell us anything about the relative housing markets between the cities themselves. Also remember that the higher the x or y value, the cheaper it is to rent/buy in the city we're testing.

Cities that I'm testing: Houston (where I'm from), Austin (where I'm living), El Paso (generally regarded as an undervalued housing market), Seattle (generally regarded as a likely bubble), and Los Angeles (ditto).

With the exception of Austin, the cities went almost exactly how you would expect given the hypothesis. The changes were more drastic than I would have even figured.

• The order from most likely to least likely to be a bubble based on buyer/renter cost ratios was: Los Angeles, Seattle, Las Vegas, Austin, Phoenix, Houston, Denver, and El Paso.
• The difference between buying and renting in Los Angeles is night and day. If you rent, cost of living in Los Angeles is actually cheaper than that in Tyler.
• There were significant renter discounts in Seattle and Las Vegas, as well, but they were less pronounced.
• Austin came down in favor of renting rather than buying against all three though I would have figured that it would be more pronounced than it was.
• The most puzzling of the results were Phoenix and Denver. Both had a rather high proportion of bad loans and should thus be renter-friendly rather than this close to the middle. So those markets aren't in nearly as much danger as I had thought or they are exceptions to this general rule for some reason.
• Houston was the only one with mixed results. The differential ran in favor of renting compared to Lake Charles and Tyler, but more heavily in favor of buying compared to Casper.
• As expected, El Paso is very buyer-friendly. I actually have to wonder what makes El Paso so much less desirable for investors than Tuscon
• Originally I was only going to use Jefferson City, Missouri for the comparison, but Jefferson City was so skewed in favor of buying it wasn't helpful. When I got the results comparing Tyler and Los Angeles, I was so suspicious that I decided to run three cities. The results were pretty consistent.

Unfortunately, the specific numbers (that I spent over an hour collecting) were accidentally erased due to some limitations between Nucleus and the webhost settings and my own carelessness in not accounting for said limitations. I need to start using w.bloggar again whenever I'm typing a longer post. You're free to toy around with the calculator and retrieve them for yourself, though.