What's Driving Home Prices (Part 2)

Building a properly supply-focused housing sector POV

In the previous posts, I discussed how the increase in home prices has been systematically the result of localized rising rents. Rents have become increasingly important as determinants of home values as they have become increasingly excessive in selected locations - locations that happen to limit new home building in historically unprecedented ways.

Here, I will take a look at long term trends. Figure 1 is a measure of owner-occupied home values as a % of GDP. This figure has basically doubled since the mid-1970s, after having been relatively flat during the preceding decades.

Figure 1

This is a big deal, and the question it raises is: Why? To what extent is this increase in values due to increased demand or constrained supply? Are we living large in subsidized McMansions or signing inflated rent checks to oligopolistic landlords and inflated purchase contracts to existing homeowners?

One way to parse this is to break Real Estate Value/GDP into 2 components: Real Estate Value/Rental Value x Rental Value/GDP.

Figure 2

It looks like the doubling of real estate values comes from two factors of roughly equal scale: Rising rental values that amount to nearly 50% since the 1970s and very volatile rising valuations that amount to nearly 50% since the 1970s. As a first estimation, you might concede that rising rents (which could be either from greater consumption or from rising prices) account for half the rise, but that at least half is from demand side factors that drive up prices (low interest rates, federal subsidies, etc.)

But, one could argue that valuations have been inflated through various forms of national subsidization, etc., and that has led to more housing (i.e. more rental value as % of GDP).

Note that, as I posted previously, rising aggregate values have been correlated with rising rents. The cycles in price/rent have been highly correlated with rents. When rents become more important, price/rent ratios increase, but they mostly increase where rents are high. So, even the cyclical part could be driven by rents.

Figure 3

So, we really are left with two extremes. Either increasingly stimulative federal policies keep driving up valuations and spending on housing, which is expressed as both rising rents and rising valuations. Or, supply constraints are driving up rents, and rising rents are driving rising valuations.

If we take these causal stories to their inevitable conclusions, there is little potential middle ground. It’s either all driven, fundamentally, by rising rents (constrained supply) or it’s all driven, fundamentally, by excessive demand.

One additional way to look at the evidence is to break down the rent/GDP measure over time into a real measure and an inflation measure. In figure 4, relative rental values (using owner-occupied properties to maintain an apples-to-apples comparison) have risen entirely because of inflation (red). Real rental value, adjusted for inflation, has declined relative to other forms of consumption (blue).

Figure 4

I would argue that if there is any period that portrays a market dominated by federal subsidies and stimulus, it is the pre-1980 period where housing values remained stable, rent inflation was low relative to other prices, and relative real housing consumption was growing. If we needed a confirmation that the main result of generous federally facilitated mortgage lending is to lower rents, we got it in the 2010s when mortgage lending was sharply curtailed, rents skyrocketed, and real housing consumption collapsed.

This is not the picture of an overstimulated market. It is a supply deprived market.

Another hypothesis could be that construction has seen poor productivity growth over the past 50 years, and that is why inflation is high and real housing consumption is low. But, if that was the case, total rental values/GDP would be declining rather than rising. Consider horses as transportation. Productivity of horse-powered transportation trailed productivity gains of other forms of transportation. So, I expect that inflation in horse maintenance has been above average over the last century, but our total spending on horse transportation has grown at a slower rate than other forms of transportation because of its low productivity growth. It would be very strange for Americans to increase their relative total spending on housing by nearly 50% because housing was generally more expensive to build.

Secondly, if that hypothesis was the primary mover, then home values and rents would be increasing in most places equally. But, as the earlier posts established, housing costs are highly localized.

Another facet one might apply to salvage a demand-side view is that short-term housing supply is inelastic. So, these cyclical upheavals could be the result of wrong-headed stimulus and subsidy, which, when it heats up, runs up against short term inelastic supply and leads to price spikes. Certainly, that is a factor right now. And, it could be an explanation for each individual cyclical price boom. But, it can’t explain the long term continuation of excessive valuations. It seems to me to be an example of the old saw that eventually when you add up a bunch of short-terms, you have a long-term. Has anyone seriously tried to tackle this problem?

Actually, this is one of the harms that has come from the negative credit shock after 2008, which drove valuations down in a way that had nothing to do with long-term supply elasticity. It was a demand shock that produced (on a shallow level) the results one would expect to see when more elastic long-term supply replaced the inelastic short-term supply. Clearly that isn’t what happened, and it is a result that could only be mistakenly applied once.

I must contend that this is basically, entirely, a rent story - a constrained supply story - and there is little room for compromise. There may be forms of subsidy or stimulus. There may be a lack of productivity in construction. The price of beans in China may have also increased. In spite of strong intuitions that might favor the former hypotheses over the latter one, none of them, it turns out, work very well as an explanation for the evolution of US housing markets.

Finally, I will note that there is no indication in the BEA data (figure 4 blue line) on real relative rental values of cyclical increases in real housing growth. In the late 1980s and the mid-2000s, the blue line just chugged along sideways. And, of course, in the valuation boom of the 2010s, the blue line (real rental values) declined more sharply than it had over any post-WW II period. The 2010s really should put the nail in the coffin of demand-driven explanations for high home values.

Now, of course, every asset value is, as a matter of course, always dually determined by supply and demand. And the 2010s were the result of trying to bring those values down with demand-side levers. Since those were the wrong levers, they made the valuation problem worse.

Additionally, I would point you to the rise in relative rental values from 2007 to 2009. One of the themes that comes out of the demand-side explanation for the 2000s valuation boom is that we weren’t as wealthy as we thought we were. Various forms of stimulus had driven home values unsustainably high, and we had to face the inevitable reckoning.

In reality, we were not as wealthy as we should have been, and what we sorely needed to acquire the wealth we had coming to us was more urban housing. Instead of bridging that gap with more well-placed housing, we bridged it by reducing our non-housing wealth. The rise in relative real housing consumption from 2007 to 2009 was the result. We made ourselves temporarily poor enough to match with our lethargic housing stock. That language (“not as wealthy as we thought”) literally describes the poverty-begging result of misunderstanding the role of inadequate supply as a source of inflated real estate values.

I can only assume that this will become increasingly ridiculous until it becomes too obvious to ignore. In 2030, how many of us will look back at 2 decades with no per capita increase in real housing and perennial 5%+ rent inflation, and continue to push for tighter monetary policy, lower incomes, tighter lending standards, or anti-investor or anti-landlord regulations in an effort to solve it?

This has been a regime shift, so it is understandable that it has taken a while to understand. At this point, it is a matter of how ridiculous it will have to become in order to create a plurality of support for aggregate growth and local land use deregulation. In the meantime, from a pure cash flow point of view, those with the means to own will benefit from the ridiculousness, though, of course, a policy regime bent on bringing down asset values by constraining liquidity may not be the owner’s best friend from from a mark-to-market point of view. I don’t have to tell this to anyone who was around in 2008. That error, and its effect on market values, will become increasingly difficult to make, however.

Slowly, these posts will build a paradigm for a new way of understanding recent trends in American residential real estate. I’m happy to receive feedback. Am I taking this too slow? Is it repetitive?

As always, please consider supporting the work with a subscription and sharing it with anyone who might benefit from thinking about it.