Before I go into the next post about long term no-arbitrage pricing for homes, I want to address the notion of a home as a security.
First, I realize that as individual assets, homes have an extremely localized exposure. Because of the way we own homes, we regard them in this framework, and that works for the purposes of individual wealth management decisions. But, this is really no different than corporate equity. Firms also can have high idiosyncratic behaviors. And, in fact, just as households and home ownership decisions, most work done by financial analysts is concerned with these idiosyncrasies. So, the localized nature of homes does not particularly set them apart from other asset classes.
There is something interesting to think about here, regarding modern portfolio theory and the effects of diversification. Theoretically, the ability to diversify away idiosyncratic risk creates a marketplace of securities with returns that, in the aggregate, reflect only the systemic risks of the asset class itself. We tend to think about the ability to actually hold a good facsimile of the market basket of securities as the process through which this takes place. But, an individual investor can accomplish most of the available risk reduction of diversification with only a handful of securities. And, in the end, the arbitrage of prices to reflect systemic and idiosyncratic risks really comes from a more diffuse and complex set of trades and allocation decisions that don't necessarily rely on any single investor to be fully diversified.
In fact, it appears to me that this process can take place with a seemingly thin network of potential trades. It is hard for us to conceive of the millions of allocation decisions and how each decision creates a network of 1st and 2nd order effects on its potential substitutes.
Thinking about these things in housing is always complicated by the coincidental nature of housing consumption and home ownership when most households supply their own demand. It is important to keep these actions separate for useful analysis, and I find that they almost never are. So, we might think of decisions about moving between cities, population flows, etc. as part of this process, but those changes are changes in housing consumption. They are priors in the market for home ownership. Markets for home ownership are pretty efficient, so, given these priors, just a small set of transactions on the margin pushes prices to account for fundamentals. More often than not, these marginal transactions probably involve landlords who must treat the transaction mathematically.
So, what I find is that returns to home ownership are actually bid down to a return level that reflects few returns, on net, to idiosyncratic risk and reasonable returns on risk adjusted investment. Homes, in the aggregate, have very stable incomes - similar to the income on inflation-protected bonds - and rather than having default risk they have liquidity risks and occasional vacancy risk (which is nearly eliminated for an owner-occupier). And, the return level we see for homes over time reflects this - returns are very similar to inflation protected treasuries. Actually, I find it useful to treat mortgage rates as a proxy for home returns (in nominal terms, with inflation included), and we know that mortgage rates tend to have a relatively small spread of about 1% above long term treasuries.
I hear pushback regarding this idea from financial advisors. They tend to separate home ownership from other portfolio management decisions. People buy homes because of the benefits of owning, not because they offer high ROI, they tell me. It's consumption, not investment. I realize that there are benefits to ownership and that families don't tend to think of their decision to buy homes as portfolio management. That is because homes aren't commodified like bonds, so as individuals, our focus is on the factors that differentiate one house from another. Very few of us are the marginal buyer, and it is the marginal buyer that pushes homes to non-arbitrage prices.
We tend to live a renting lifestyle where the high transaction costs of housing make it a non-starter until we step over some threshold in our life cycle that puts us in the homeowner category. These things tend to happen in regime shifts, so we tend to hop right over the marginal buyer. This gives us the impression that the marginal buyer doesn't exist. But, that's a false impression. There are some portion of owner-occupiers who are at the margin, and landlords and homebuilders are basically always there.
One other piece of confusion that I want to avoid, which I have mentioned before is the idea of thinking of home values as leveraged purchases made possible with mortgages. This, along with the idea that most households are not on the margin, and would be willing to bid home prices up if they needed to, feeds the idea of a housing market given to irrational booms when credit is loose. But, many households have very high levels of equity. Housing generally lacks price discrimination, so it seems to me that it would be difficult to argue that prices respond to buyers with excessive consumer surplus. The forces at the margin of supply and demand set the prices.
So, I tend to analyze home values from a total equity point of view. Pulling in mortgage factors just confuses the matter. Homes have intrinsic values, regardless of their funding. Now, I confuse this by using mortgage rates as a proxy for required returns on homes. But, I am treating mortgages as another form of real estate ownership with a similar level of required yield, not as a source of funding or demand for the buyer.
This might all seem theoretical, but it is confirmed by the empirical data. Home prices over time follow the path you would expect them to follow if they were a part of an efficient market with relatively low systemic risks. I will look at the data, which I think will include some contrarian findings that I might have called surprising about 40 posts ago.