Will housing boom go bust like 2006? Not likely.
U.S. home prices roared higher following the onset of COVID-19, and although there has been a moderate pullback lately, the median selling price remains 30% higher than in January 2020. An asset boom inevitably raises the specter of an impending bust, especially with the memory of the real estate crash during the Great Recession of 2007 still in our collective memories.
But not so fast. While the sharp acceleration of postCOVID housing costs resembles the rapid runup leading into the financial crisis, the proximate causes of the two spikes appear to be materially different, suggesting that another traumatic decline is much less likely this time around.
Researchers Lara Lowenstein and Jason Meyer at the Federal Reserve Bank of Cleveland examined several key factors driving the residential real estate market and concluded the two housing booms this century trace their origin to distinctly different fundamental drivers. The economists found that the current surge is tied to fundamental factors including structural supply and demand imbalances, the state of the mortgage market and the behavior of rents. The 2000-2006 boom appears to have been driven primarily by expectations of future appreciation: in a word, speculation. Speculative bubbles burst, while fundamental imbalances adjust more slowly and (usually) more sustainably.
Let’s begin with the first housing boom of the millennium. Starting in 1997 and continuing through the peak in 2006, new home construction expanded rapidly to more than
2 million units per year, a level not seen since the mid-1970s. Note that during this building bonanza, average 30-year mortgage rates hovered above 6%, not much lower than the current environment. New construction was dominated by single family homes; over 2/3 of all units under construction in 2006 were stand-alone houses.
Yet demand continued to outstrip the expanding supply, creating incentives to further boost the quantity of new dwellings coming onto the market. However, somewhat counterintuitively, rents did not rise much during this period. In fact, average rents remained relatively stable, growing by only 10% cumulatively between 2000 and 2006, far slower than home prices rose. Meanwhile, the vacancy rate for residential units increased markedly during the same time frame. Supply, it seems, eventually outran demand.
When the banking system broke, the air came whooshing out of the real estate balloon as median selling prices fell nearly 20%, primarily in particularly overheated areas. In fact, most of the rapid increases and sudden declines took place in only a few states including California, Arizona, Nevada, Florida and parts of the Northeast. The magnitude of the boom and bust was much less pronounced over most of the rest of the U.S.
One important source of demand for housing is the rate of new household formation, which had remained relatively constant through the 1990s but slowed between 2000 and 2006 and then dropped precipitously over the next five years, further eroding the demand for homes. All of these factors lend support to the thesis that much of the housing boom of the early 2000s was driven by expectations of future price appreciation; basically, the equivalent of an overvalued growth stock bubble that was pricked by the global financial crisis.
The environment today differs in some important ways. On the supply side, new construction never recovered its pre-financial crisis level. Housing starts are still 30% below the 2006 peak, despite the addition of another 50 million to the U.S. adult population. Additionally, the composition of new housing has shifted dramatically, with multifamily now comprising a record 60% of all units under construction.
On the demand side, household formation has not only recovered but is more robust than any time since at least the 1980s even as fewer units are being built. Vacancy rates, having peaked in 2009, have fallen sharply as well, again suggesting a shortage of supply. Also, quite unlike the earlier boom, rents have risen dramatically during this cycle, up over 20% in only three years. These factors support the proposition that the current cycle of home price appreciation is much more reflective of longerterm structural supplydemand imbalances than speculative fervor.
There are also notable differences in the mortgage market today versus the previous boom.
Most residential mortgage loans are not held by the bank or lander that initiated the loan but are “securitized” or sold off into pooled debt vehicles called mortgage-backed securities, bonds that are backed by the collateral of the home equity. Today, the vast majority are insured against mortgage defaults by quasigovernmental entities like Fannie Mae and Freddie Mac. This differs markedly from the 2000s when a significant share of mortgage-backed securities were issued by uninsured private underwriters, many of which took a drubbing or flamed out spectacularly during the financial crisis.
Borrowers are also in decidedly better shape. Despite the dramatic increase in home values, the average loan-to value ratio of newly originated home loans is lower than at any time since 2000. Furthermore, that debt is much more manageable. The aggregate ratio of mortgage debt to household income is 30% lower than in 2006, while the average FICO score for newly originated mortgages is 30 points higher. Hardly the stuff of bubbles.
Quite unlike the speculative froth of the 2000s, the current boom appears to be driven much more by fundamental factors like a structural shortage of supply and demographic trends that will take some years for markets to rectify but isn’t likely to face plant.