Housing Debt and the 18-year “Real Estate” Cycle
|February 19, 2014||Posted by Rich Nymoen under Editorials|
When it comes to our ailing housing markets, we need to blaze a different trail from the one we’ve been on. For decades, the primary means of expanding home ownership, first among the middle class and then to those historically excluded from mainstream housing markets, has been through mortgage lending. Federal programs were created such as FHA and Fannie Mae/Freddie Mac to lower interest rates and down payment requirements, expanding access but also creating more indebtedness. Over time, such efforts were extended further by lenders, state and local governments, and non-profit organizations. The approach then culminated, at the peak of the last housing bubble, with the securitized-mortgage crash and foreclosure crisis from which our economy and communities are still recovering.
Unfortunately, what the lending approach has increasingly ignored is that costs of all housing not only include the labor and material expenses that go into building, improving and maintaining structures, but also the cost of the land upon which the houses sit. It is land value that is being referred to in the real estate adage of “location, location, location.” Recognizing the distinction between structure and land costs is important because lending against one has very different economic and community implications than does lending against the other.
The value of land (land being a gift of nature) reflects no labor or material costs. Rather its value is community-created, reflecting the growth in population and wealth of the surrounding community along with the community’s investments in infrastructure and public services such as schools, parks, police and fire protection — at least until a real estate bubble begins to form. Bubbles form when land prices begin to trend upward, leading sellers to demand a speculative premium to sell now rather than at a future higher price. Lending then goes up in order to keep pace with the prices and, as borrowing becomes easier, prices are driven up even further in a vicious spiral. Because labor and material costs are much more stable than land values, lending against the building value they comprise is much less volatile.
Under the current approach, when housing bubbles pop massive foreclosures result. But most foreclosures could be prevented if lending occurred only against buildings and not land. This is demonstrated by the recent experience of Community Land Trusts (CLTs), under which home buyers only finance the cost of the building not the land. After the latest bubble burst, CLTs had dramatically lower delinquency and foreclosure rates than conventional financing approaches, as shown in Figure 1.
Figure 1 – Community land trusts (CLT) vs. mortgages
Thaden, Emily and Greg Rosenberg, Outperforming the Market: Delinquency and Foreclosure Rates in Community Land Trusts (Land Lines Article, October 2010)http://www.lincolninst.edu/pubs/1846_Outperforming-the-Market–Delinquency-and-Foreclosure-Rates-in-Community-Land-Trusts
Historically, real estate bubbles are not rare occurrences. Land economists have identified a regular land price cycle averaging 18 years that consists of a 7 year moderate rise in prices, a slight lull, then a steeper 7 year rise followed by a sharp drop in prices and 4 year down period. As shown in figure 2 below, the 18 year pattern was disrupted only twice over the last two centuries, by the economy’s absorption in WWII during the 1940s (the cycle resumed in the 1950s) and by the recession induced in 1980 via the Federal Reserve’s response to stagflation. Recessions, or declines in GDP, can occur for a variety of reasons. But all GDP declines accompanied by financial industry collapses have occurred after land values peak. Indeed, based on this data, land prices are the leading economic indicator.
There are eight distinct phases of the land price cycle, as shown in Figure 3 below, and indications are that we have gone through the 4-year down period of the cycle and are currently in the cycle’s 2-year first phase that sees rising rents leading to higher prices for older buildings. If the pattern continues to hold we would expect another crash in about 2026.
The key to dampening the effects of this cycle, instead of amplifying it as we did last time, is to separate as much as possible the land market from the financial markets. There are several ways this can be done:
1) Fifty years ago, it was not uncommon that a down payment of 30% was required to finance the purchase of a home. Because 1/3 of the price of a home is typically the value of the underlying land, paying 30% up front essentially meant only the structure cost was being financed, the components of which are much less volatile;
2) Currently, the growth in community land trusts – in which the homeowner has title to and finances only the house while a non-profit organization keeps title to the land – also helps to keep land and financial markets separate;
3) Finally, because the price of land – and therefore how much must be borrowed to buy it – is affected by how much bidders have to pay in taxes for it, any land value left un-taxed usually gets pledged to banks as a mortgage. If someday the majority of public revenue came from land value taxes, there would then be much less land rent to be pledged to banks; if the tax collected the full annual rental value of land, then homeowners would in effect be renting their home sites from the community.
In sum, lending approaches are appropriate when helping to finance housing’s structure costs. But lending against land values contributes to inflating housing bubbles, resulting in foreclosures that harm the very people the lending approach is trying to help. It’s time to start using our imaginations to help keep land mortgage-free.
Figure 2 - Past Real Estate and Business Cycles in U.S.
|Peaks in Land Value||Interval||Peaks in Construction||Interval||Peaks in Business Cycle||Interval|
Foldvary, Fred, The Depression of 2008, September 18, 2007, http://www.foldvary.net/works/dep08.pdf
Figure 3 – Phases of Real Estate Cycle (phase periods are approximate)
|Phase||Last Cycle||Current Cycle (if pattern holds)|
|rising rents = higher prices for older buildings (Years 1-2)||1995-96||2013-14|
|vacancy rate declines = building profitability = fast growth in new building (Years 3-4)||1997-98||2015-16|
|construction takes off, easy credit begins to appear (Years 5-6 )||1999-00||2017-18|
|increased building activity absorbs vacant land = mid-cycle slow down (Years 7-8)||2001-02||2019-20|
|real estate boom = frenetic activity at the peak (Years 9-12)||2003-06||2021-24|
|activity slackens but confidence remains high, foreclosures increase (Years 13-14)||2007-08||2025-26|
|vacancies increase, tight credit begins, economic activity stalls (Years 15-16)||2009-10||2027-28|
|debts reduced, wreckage cleared away during recession (Years 17-18)||2011-12||2029-30|
Anderson, Phillip J. The Secret Life of Real Estate and Banking, p. 360,