In his recently released letter (.pdf) to shareholders, Warren Buffett spends a bit of time discussing the housing market. He begins by admitting that his prediction from last year's letter, "a housing recovery will probably begin within a year or so," was in fact, "dead wrong." He then mentions Berkshire's (BRK.B) five businesses whose results are "significantly" affected by housing before proclaiming that "Housing will come back - you can be sure of that." It is this proclamation that I would like to address.
Specifically, I would like to address the following excerpt from Buffett's shareholder letter:
For a period of years prior to 2008, however, America added more housing units than households. Inevitably, we ended up with far too many units and the bubble popped with a violence that shook the entire economy. That created still another problem for housing: Early in a recession, household formations slow, and in 2009 the decrease was dramatic.
That devastating supply/demand equation is now reversed: Every day we are creating more households than housing units. People may postpone hitching up during uncertain times, but eventually hormones take over. And while 'doubling-up' may be the initial reaction of some during a recession, living with in-laws can quickly lose its allure.
At our current annual pace of 600,000 housing starts - considerably less than the number of new households being formed - buyers and renters are sopping up what's left of the old oversupply.
Regarding Mr. Buffett's claim that the "devastating supply/demand equation" regarding household formation and housing units has reversed and that the current pace of housing starts is "considerably less than the number of new households being formed," let's take a look at data from census.gov. According to data in Table 61, "Households and Persons Per Household by Type of Household" in the "Population" section of The 2012 Statistical Abstract, there were 116.783 million households in 2008. In 2009, this number grew to 117.181 million, and in 2010, it grew to 117.538 million. Data for 2011 has not yet been released.
As you can see from the most recently available data, household formation was 398,000 in 2009 and 357,000 in 2010. Keep in mind that the National Bureau of Economic Research claims the Great Recession ended in late spring/early summer 2009. Therefore, household formation continued to slow after the "official" end to the recession. Unless you believe housing formations more than doubled in 2011 from 2010 levels, it's hard to agree with Buffett's claim that an annual pace of 600,000 housing starts is "considerably less than the number of new households being formed."
At the beginning of 2011, there were high hopes for housing formation. While the Census Bureau has yet to update its information, a November 29, 2011 Businessweek.com article cites IHS Global Insight as determining a 2011 household formation number of 600,000, far from "considerably" outpacing housing starts.
Furthermore, Mr. Buffett's assertion, "while 'doubling-up' may be the initial reaction of some during a recession, living with in-laws can quickly lose its allure," shows he is a bit out of touch with the economic realities tens of millions of American are living under. With this statement, Buffett seems to be implying that many people living with in-laws have chosen to do so not because of economic/financial hardships but for some other reason, perhaps because they simply don't want to spend the money they have or have access to on a home purchase. He seems to be suggesting that once the "allure" of living in one's parents' basement as a 30 year old wears off, people will simply venture out and spend all the money they supposedly had, but chose not to spend before.
Over the past few years, I have met a handful of people who have moved back in with their parents. I've also read numerous accounts of people, sometimes families, forced to move in with relatives. In not one case have I come across someone who did so because he or she wanted to. It was a choice that was made out of economic hardship. To claim this choice is simply an "initial reaction" to a recession and that once the "allure" runs out people will simply pick up and move seems to show a serious lack of understanding regarding the financial pressures under which millions of Americans are currently living.
Next, in the closing paragraph of his discussion on housing, Buffett says,
Wise monetary and fiscal policies play an important role in tempering recessions, but these tools don't create households nor eliminate excess housing units. Fortunately, demographics and our market system will restore the needed balance - probably before long.
While an entire article could be devoted to discussing whether the monetary and fiscal policies used in "tempering recessions," (throwing massive amounts of money, printed and borrowed, at the problem) can ever be described as "wise," I would like to forgo that discussion and focus on whether the needed balance between household formation and housing units will be returning any time soon. Let's again turn to data from census.gov.
When thinking about household formations and what the trend might look like over the coming years, I found it useful to dig into recent trends among the type of households. Again, let's look at data from Table 61 in the "Population" section of the U.S. Census Bureau's The 2012 Statistical Abstract. In 2008, family households made up 77.873 million of the 116.783 million total. This number grew to 78.850 million in 2009 and then shrank in 2010, coming in at 78.833 million. When drilling down even further, we discover that a big reason for the drop was due to a rather large 708,000 household drop in "Married couple" families.
Also of note is the following: In 2010, there was a large rise in the number of male and female households falling under the "family" category with "no spouse present." The number of female family households grew by 2.5% and the number of male family households by 6.25%, versus the 0.3% growth in total households in 2010. The growth rates in female and male family households were significantly faster than the growth rates from 2008 to 2009 of 0.53% and 2.98% respectively.
A couple years worth of data do not make a trend. However, if you are attempting to predict the future of housing, it is important to spend a bit of time thinking about the potential ramifications of an ongoing decline or even just extremely slow growth in the number of families, specifically married couples, as well as the implications of growth in households with "no spouse present." There may be differing opinions on this, but in the United States, who is most likely to buy a home: a single person, a married couple, a married couple with children, or a single parent with children?
If the growth of household formation begins to shift in the favor of single parent households and/or nonfamily households, it could have serious implications for the housing market going forward. At a minimum, it could affect the size of houses being built, which, in turn, would affect the amount of stuff people buy for their homes.
According to data from census.gov, in 1975, the average square feet of floor area in new single-family houses was 1,645. This grew to 2,521 square feet in 2007. In 2008, the number dropped to 2,519; in 2009, it dropped to 2,438; and in 2010, it dropped to 2,392. For investors interested in predicting not just a recovery in home sales, but all the derivative effects of housing on the economy (construction jobs, sales of home building materials, sales of things people put in their homes), the square footage of new homes being built will be important to follow.
Also, if the recent decline in square footage indicates a trend in the type of homes people are favoring nowadays, this may be an omen for the future of the McMansion portion of the housing market. Moreover, when combining smaller square footage with tighter credit standards and a decline/slowing of growth not just in family households but also in married couple family households (lots of two income earners families there), it's hard to imagine the part of the housing market that many people in the "middle class" entered into over the past decade coming into balance (on a national scale) "before long."
Before addressing the implications for investors of recent household formation and square footage data, I would like to mention the following: If you believe that family households and specifically married couple family households are more likely than other types of households to purchase homes, thinking through trends in the education system in the United States is important.
Over the past few decades, the number of people graduating from college each year has grown strongly. In 1980, 929,417 Bachelor's degrees were awarded. In 2000, the number was 1,237,875. By 2009, the number of Bachelor's degrees awarded each year was still growing strongly, reaching 1,601,368. Likewise, there has been significant growth in the number of Masters and Doctoral degrees awarded each year.
When thinking about the future of housing, think about how long people are spending in school nowadays, the amount of debt they leave school with, and whether the economy will be able to create enough well-paying jobs for all those college graduates. Furthermore, think about what this means for graduates in terms of wanting/needing to find a job and its implications on family household formation (delaying it). This is yet another topic about which an entire article and even numerous studies could be dedicated. In the meantime, for investors attempting to gauge the future of housing, put aside political and religious beliefs, and just think about it from the dollars and cents perspective of a 20-something college graduate starting his or her life in the "real world."
In closing, let's look at things from an investor's point of view. In my article, "Why The Market Is Not Necessarily Cheap," I spend some time discussing various things that might prevent investors from paying higher price-to-earnings ratios for stocks. If you are thinking about putting money to work in the S&P 500 (SPY), Dow Jones Industrial Average (DIA), or even in housing-related equities (XHB), slower growth in household formation and declining square footage of newly built homes could be headwinds for index-wide valuations going forward. Investors don't pay up for slow growth, nor do they pay up for an economy with a housing market in, as Mr. Buffett put it in his recent letter to shareholders, a "depression."
If slower growth in household formation, specifically family household formation, and declining square footage of new homes become a trend over the coming years, it will likely slow the potential gains for new home construction and cause investors to adjust downward expectations of future sales in household-related items.
Does this mean you should sell housing-related stocks like Lennar (LEN), KB Home (KBH), or Masco (MAS), which doubled off their October 2011 lows and are still trading relatively close to those highs? It is true that equities often bottom before the fundamentals of an industry/economy clearly indicate things are improving, and in recent months, the rally in many housing-related equities has been breathtaking. If I were trading these stocks from the long side, I would be taking profits, selling some now, and selling some into rallies. For buy-and-hold investors, keep an eye on the bonds of these companies. When investing in the stocks of companies with junk-rated debt (as many well-known housing-related companies have), the debt will often paint a more sober picture than the stocks will of the financial realities of those types of companies.
How ever you decide to shape your views of the housing market and its implications for investors, keep in mind that even the best investors in the world (Buffett among them) can be wrong sometimes. Buffett maintains that a balance in housing will occur "before long." I don't dismiss that possibility. However, when attempting to determine the likelihood of this happening, investors should look beyond headline numbers and beyond assumptions people make about future household formation.
Instead, focus on underlying trends in data, think about the way younger people are choosing to live their lives, and even spend some time working through all the derivative effects of a housing market that might look different going forward from what society became accustomed to over the previous decades.
Disclosure: I am long Masco's senior unsecured note, CUSIP: 574599AT3