Coming out of most economic downturns, homebuilding is one of the key locomotives to power the economy. Housing is sort of the ultimate durable good, where during downturns demand builds up, and then has a powerful upward force on the economy as the pent-up demand is released.

Given that almost all houses are financed rather than bought with cash, the sector is exquisitely interest-rate sensitive. However, with the rate of home ownership falling, residential investment (RI) will not be a very powerful engine this time around.

First, let me present the following graph (from http://www.calculatedriskblog.com) showing that historically what has been one of the key forces behind both going into and coming out of recessions — it shows the history of housing starts over the last 40 years. Notice that housing starts, and thus residential investment, tends to fall off sharply before the start of most recessions (the dot.com bust induced 2001 recession the major exception) but then picks up sharply coming out of a recession, with the bottom hit before the recession is officially over most of the time. Note that single-family homes (red line), which are more likely to be owner occupied, were particularly strong from 1995 to the end of 2006.

This was in large part because the homeownership rate was expanding, as is seen in the second graph (also from http://www.calculatedriskblog.com). From 1995 to 2005 the home ownership rate rose from under 64% to over 69%, or at about 0.5% per year.

With a population of about 300 million, and about 2.4 people per household, that means there are about 125 million households. The population is also growing at about 3 million per year, meaning that if the home ownership rate had stayed constant at 64%, we would have had a demand for about 800,000 new homes a year (3,000,000/2.4 * 0.64), but the increase in the homeownership rate boosted demand by about 500,000 a year to 1.3 million.

Now, as the price of houses is falling and people are walking way or getting foreclosed on, the homeownership rate is falling at about 0.5% per year, thus subtracting about 500,000 a year from demand, leaving it at about 300,000.

If a household does not own its house, it has to rent. If 64% were home owners, then it logically follows that 36% were renters, and that if there were no change in the homeownership rate, then demand for new apartments would have been 450,000 (3,000,000/2.4 * 0.36). By the end of the process the number would have been 375,500, as that 36% shrank to 31%.

However, if the homeownership total is going up by 500,000 a year, that means that the demand for apartments had to be actually falling by 50,000 a year at the start and growing 112,500 by 2005. Indeed, during this period, the total number of rental units did stop growing, as is shown in the third graph (also from http://www.calculatedriskblog.com), but it did not fall at anything close to the 50,000 a year pace, let alone a 112,500 pace.

It’s not just a slowdown in new construction of apartments, but also condo conversions (and the tearing down of some units as well). However, the net result was a rising number of rental vacancies (shown as the pink part of the graph). That is going to put downward pressure on rents. Since rent, both “rent” rent paid to landlords and “owners equivalent” rent, what the government considers you are paying yourself for living in your own home for calculating inflation make up over 30% of the CPI, and almost 40% of the CPI excluding food and energy. It is a key reason that inflation is likely to stay low, particularly at the core level for some time to come.

This is also extremely bad news for the owners of apartments, like the big apartment REITs such as Equity Residential (EQR) and Apartment Investors (AIV). Since the peak of the housing market, the number of rental units has increased dramatically. Is that because people are breaking ground on lots of big apartment buildings?

No. Refer back to the first graph — it’s not like there has been a big increase in the difference between total housing starts (including apartments and condos) and the number of single-family starts. Rather, what we have seen is a rash of condo reconversions, where units that were originally planned to be condos are turned in to rental units. Sometimes this is done by the developer, other times it is the individual condo owner who can’t sell and decides to rent it out (depending on the condo association rules). Also, many of the people who have been buying up the previously foreclosed houses have been investors, who plan to rent them out, rather than live in the houses they buy.

Housing is both an item we consume (housing services) and an asset — for most people, their single most important asset. Well, the value of any asset is the value of all future cash flows from the asset, discounted back to the present. In the case of a house, it is the value of not having to pay rent to live in the place.

One of the most important indicators that we were in a housing bubble was the fact that the ratio of housing prices-to-rents got way out of whack. This is shown in the next graph (yet again from http://www.calculatedriskblog.com, to whom I am greatly indebted for much of this analysis). The crash in housing prices has brought the price to rent ratio back down to near normal levels. However, with pressure on rents from the high vacancy rate, housing prices have a moving (falling) target to be shooting at.

But wait, the news gets worse! This recession has been particularly hard on young people. With their 401-Ks hurt by the fall in the stock market, and the value of their houses way down, people in their 60’s — if they still have their jobs — are not about to quit and retire. They simply cannot afford to. That has meant fewer jobs available for people just getting out of college (and forget about getting a job if you only have a High School Diploma).

If young people can’t get a job, they tend to move back in with their parents, or perhaps double or triple up with their friends. This means that the rate of household formation slows down and the average size of the household grows. If the size of the average household were to increase from 2.4 people to 2.5 people (not a very large increase), it would result in a total of 5 million fewer households in a population of 300 million. This would more than offset the number of new households being formed due to population growth.

Given this, it is hard to see housing starts, and thus residential investment, climbing back to anything close to what we saw in the earlier part of the decade. While it is true that residential investment was only 2.67% of GDP in the second quarter, that is the lowest level on record, and is a far cry from its peak of over 6.3% in the boom. However, I suspect that we will be lucky to see it get back up to its historical average level of about 4.5% of GDP.

It also means that a return to the sort of profitability the big homebuilders like D.R. Horton (DHI) and Lennar (LEN) saw earlier this decade is going to be almost impossible. Yeah, those stocks got slaughtered last year, but have since increased dramatically, some by more than 5x. This increase seems to me to be a triumph of hope over reality.

Homebuilding is much more important than the profits of a few relatively small firms in the S&P 500, though. It is historically what drives the economy out of recessions. While just seeing residential investment halt its decline will be a major boost to the economy (the absence of a negative is a positive) it is not the same thing as an actual rebound. A declining homeownership rate is one of the key reasons that this recovery is going to be exceptionally anemic, but also why inflation is not going to be a major problem for some time to come.
Read the full analyst report on “EQR”
Read the full analyst report on “AIV”
Read the full analyst report on “DHI”
Read the full analyst report on “LEN”
Zacks Investment Research