Housing Analysis: You read it here first

I relgiously read two of John Mauldin's weekly newsletters, Thoughts from the Frontline and the Outside the Box E-Letter. Last Friday, in FrontLine, he mentioned two deep analyses of the housing market, both Powerpoint™ presentations. Based on his recommendation I read them both. One slide, from John Burns Real Estate Consulting made me particularly proud, because of this chart:

A few weeks ago, in my own reading of the housing tea leaves I produced and published this chart:

When I published this analysis, readers of this blog were quick to comment on my view that it is necessary to see a significant overshoot to the downside in order to offset the long and excessive overshoot to the upside. This follows from the very definition of a trend line. A trend line is defined using a statistical process known as ordinary least squares (OLS). What this does is draw a line through the data set such that minimizes the total distance between the square of the actual data (the Y-value) and the trend line, for every X-value in the data set. Squares are used to eliminate negative values. In other words, we try to find the line that has the least “error” – that one which does the best job of estimating what the Y -should be for any X-value.

In a good data series with a strong trend, we can often “free-hand draw” a trend line through a data series because visually we see that the line must be roughly “in the middle” that is – that error of underestimation (where the actual data is above the trend line) must roughly equal the amount of overestimation (where the actual data fall below the trend line). That is, the point at which there is the least “error” in estimation.

Drawing a trend line through a data set is easy (actually, we can draw multiple trend “lines” as many trends are better described by curves, such as compound interest projections). The question is – does the trend line matter: that is, is the relationship a significant one.

In the case of home sales and their growth over time, the answer is yes. Home sales volume, over the long term, is driven by household formation which itself is remarkably stable, though it certainly has gone through periods of faster and slower expansion, thus residential real estate sales growth over time is remarkably stable (note that the growth rate is the slope of the trend line). Home sales growth is a geometric series with respect to time, however, so there should be a curve sloping upwards, the problem is, the growth rate itself is so small (At less than 1% per year) that even over long periods of time the series may appear to be linear.

Nevertheless, not to get carried away, what is important to realize is in order for housing to bottom, sales must remain below trend for some time. This is not what policymakers are suggesting. Rather, they are trying to find ways to support bubble-era prices in a post-bubble environment.

One suggestion has been for the Federal Government to purchase and demolish undesired homes so as to reduce inventory. Incidentally, this policy is not without precedent in major cities that have fallen on hard times. Philly and Buffalo both have crews who perform demolitions, though sadly, there are so many homes that the crews have multi-year backlogs.

While I think this to be a good idea in cities that clearly have far too many homes to support the population that actually wants to live there, I think it should be done at the local level, since it was local planning boards who approved the building of the homes in the first place. Local taxpayers will suffer the demolition costs, but they are also the ones who will benefit from the presumably higher prices that derive from having fewer homes for sale (and fewer abandoned homes in the neighbourhood). I see no reason to make residents of Omaha pay for demolition to support home prices in Las Vegas.

But attempts to manage inventory is trying to fix the wrong problem. The issue with sales (and ultimately prices) is that there are relatively few buyers. Very few people have anything like an incentive to purchase a home, given that buying a home represents a huge premium to renting. Moreover, very few people can come up with a substantial down payment required to purchase a home in the absence of home price appreciation. But reducing inventory will likely make the problem worse.

Reducing inventory is an attempt to maintain artificially high prices – but who would or could pay those prices? Even the people who “purchased” the homes, using questionable financing let us not forget, could not afford those homes. At best we will exacerbate the problem of poor affordability at worst we could reignite bubble lending and more building.

Do you see, as our politicians do not, that keeping prices high means that the only option for purchasing is to continue to rely on low-down payment financing and hope for future appreciation (i.e. bubble thinking)? The higher prices remain, the larger the down payment required to obtain a conventional mortgage. Why do we want to make it harder for people to purchase a home?

The best option is to see a reduction in prices. Only a revaluation of owned real estate with respect to rented real estate can encourage buyers to make the switch from renting to owning or to purchase an additional property with the intent of renting it. This could, begin to help us lift that trend line.

Conclusion

Housing sales will be down for years. The fundamentals are against home purchase, though there is some potential in trade up now, if you can sell your own home. First time buyers are still priced out of the market and will be discouraged by falling prices to risk hard-saved equity. Moreover, lenders will be insistent on having lower LTV so as to ensure adequate protection for their own loans. The lack of buyers cannot be materially improved by policy moves, though it can be materially impaired by them.

I also recommend that you sign up for John Mauldin’s newsletters here and that you read the presentations here and here.

Just remember, you read it at the Strategic Investor first.