Meet the Author

Michael Shenk |

Research Assistant

Michael Shenk

Michael Shenk was formerly a research assistant in the Research Department of the Federal Reserve Bank of Cleveland. His work focused on international topics and housing-market indicators.

11.09.07

Economic Trends

Housing Cycles

Michael Shenk

October’s housing data, which predominately covers what happened in September, did little to change our picture of the housing market.  A look at the recently released data shows that the downturn in housing continued through September. While this is not good news, some consolation can be had by comparing this downturn to others in the past. Looking at the last three housing slumps suggests this one is pretty typical. In fact, the current downturn actually compares favorably in some metrics. But then again, nothing indicates it is ending anytime soon. 

To compare housing downturns across time in a meaningful way, it helps to look at various indicators of the market as indexes as opposed to raw numbers.  By indexing the data of the different series to their respective peaks, we can get a clearer picture of the severity of each downturn over a given time frame.

During the four most recent downturns, single-family housing starts have fallen at fairly comparable rates.  With the exception of the housing slump beginning in 1990, which is noticeably shorter in terms of its effect on starts, the current position of housing starts is very close to what it was in each of the previous cycles. While the pace and size of the decline are very similar in all of these cases, the most recent downturn sticks out in terms of how long it has lasted.  In none of the previous cycles had starts fallen consistently for more than a year and a half without a noticeable uptick.

As for new-single family home sales, their current decline appears to be pretty typical as well. This shouldn’t be particularly surprising, assuming that there is a link between housing starts and new home sales. The current cycle does differ slightly in the pace of the sales decline, which was somewhat slower than the previous cycles through the first year and a half.  And once again, around the 18-month mark, each of the previous cycles bottomed out, which has not yet happened in the current cycle.

The real median sales price for new single-family homes has historically been fairly flat throughout downturns in sales, with only a modest downward trend starting about 12 months after the peak in sales.  Thus far, the current cycle does not seem to be an exception to this trend, but prices do appear to have been somewhat more resilient through the first 18 months of the cycle than in the past.  The resiliency of prices may partially explain why previous cycles saw an uptick in sales around this time and the current cycle has not as of yet. (The median sales price of new homes may not necessarily reflect the actual cost of a new home.  Sellers of new homes often can adjust the actual cost of the home by offering nonprice incentives or discounts that are not reflected in the sales price.)

The market for existing single-family homes, which is considerably larger than the market for new single-family homes, also appears to be behaving typically for a period of decline, although a bit more favorably than in previous slumps.  Through the first 18 months of the current cycle, existing home sales were doing considerably better than all but the cycle beginning in 1990. (And that cycle would have had a much sharper downturn if data from December 1988 to June 1989 had been included, as there was a sharp decline in existing home sales during the period. The period was excluded, however, in order to keep the time periods consistent with the downturns in starts and new home sales.) As was the case with both housing starts and new home sales, sales of existing homes, which in previous cycles began to bottom out around the year and a half mark, have not shown any signs of reaching a trough in this cycle.

Perhaps responsible for the more modest decline in sales is the decline seen in the real median sales price of new homes.  The current cycle’s decline tracks the price movements of the cycle beginning in 1980 fairly closely.  A notable difference between these two cycles is that the real price decline in the 1980 cycle was accompanied by a much steeper decline in sales.  Still, the current decline in prices may be a little disconcerting for the typical homeowner, as homes generally make up a large portion of a household’s wealth.  It also may be a point of concern from a macroeconomic perspective if the price declines spill over into households’ consumption decisions.  Thus far, we have yet to see any signs that the consumer market is being affected by weakness in home prices.

So is this the worst housing slump in the last 30 years?  Although it is too soon to tell, so far it is broadly similar to the ones we have seen before.  Of course, there is one important difference we haven’t yet mentioned: all three of the previous housing slumps were accompanied by recessions, while this one, so far, has not.