One of the economic trends that we think investors should focus on over the next few years is the nascent housing recovery. We believe that a normalization of the residential construction industry has the potential to surprise investors with rapid and material improvement in the economic picture of the US, from GDP and unemployment to state and Federal budgets. In Part I below, we’ll discuss the recent history and current state of housing supply and demand, while in Part II, we’ll investigate how foreclosures and excess housing stock affect the market and what could be the possible economic impact of a recovery.
New Housing Supply
Between 2002 and 2006, Census Bureau data show that construction began on more new housing units (e.g. single-family homes, apartment units, condos, etc.) in the US than in any other five-year period since the late 60’s. A housing bubble was underway as loose loan standards and a belief that home values could not fall led to speculative fever that nearly doubled housing prices on average nationwide.
Note, however, that in contrast the five years ending 2011 saw the fewest number of housing starts in the last 40 years. This dramatic shift in the pace of home construction poses an important question: have the last five years of anemic housing construction made up for the excesses of the bubble era?
Examining housing starts on a rolling ten-year basis shows that the decade ending in 2011 had one of the lowest totals during the period shown. In fact, only during the ten years ending 1997 did construction begin on fewer housing units. These data suggest that the excesses of the bubble have indeed been rectified, at least in terms of housing supply.
But what about demand for housing? The Census Bureau defines a household as a person living alone or a group of people living together in a single home or apartment, regardless of whether they constitute an actual family. From 2001 through 2005 (the last full year that home prices consistently increased), an average of ~1.3M households formed in the US, just a little under the 50-year average. In contrast, during the subsequent period of 2006 through 2011 only an average of ~600K households formed, yet the population grew at largely the same rate. Weak household formation is most likely related to high unemployment, as people have delayed moving out of their parents’ homes, getting married, or even getting divorced until their employment situations are secure. Demand for housing appears to therefore be driving the recent low numbers of housing starts: now that the mania of the bubble is over, homes are largely being purchased to be occupied instead of as investments to be flipped.
Even with an unemployment rate that remains stubbornly high, there are recent signs that household formation has begun to normalize. The Census Bureau estimates that in each month of 2012 (through September) there were ~1M more households on average than in the same month of the previous year. This is a significant change from 2011 when the average year-over-year change was ~635k and from the previous four years when the average was ~550k. Household formation in 2012 is therefore approaching more normal levels, close to the historical average of ~1.3M. Over the long-term, household formation is what drives demand for housing and for home construction. The current positive trends in household formation therefore suggest that demand for housing is starting to finally look healthy again.
Housing Market Normalization?
The data show that a marked reduction in home construction has severely curtailed the amount of new supply coming to market over the last few years. Home builders were largely responding to changes in demand as household formation dramatically slowed after home prices peaked. Now that households are being formed at ~1M annual run-rate in 2012, home construction is picking back up, but the annual rate implied by September’s housing starts is only 872k and construction was only started on an estimated 582.5k housing units in the first three quarters of 2012.
One last point to ponder is that the Federal Reserve Bank of Cleveland estimates that there was a household shortfall of 2.6M between 2008 and 2011. In other words, ~2.6M households that were expected to form based on population growth failed to materialize. These “missing” households likely represent pent-up demand for housing, which begs the question (which we’ll delve deeper into in Part II) of whether the housing market could easily absorb this demand if something (like an improving economy) triggered their formation.