In Depth Analysis: CalculatedRisk Newsletter on Real Estate (Ad Free) Read it here.

Monday, July 09, 2007

Housing and Consumption

by Calculated Risk on 7/09/2007 12:07:00 AM

The "unexpected" sharp slowdown in Q2 Personal Consumption Expenditures (PCE) will likely restart the discussion of the impact of the housing slump - and less Mortgage Equity Withdrawal (MEW) - on PCE.

Some variation of the following graph will probably be part of the discussion.

Mortgage Equity Withdrawal vs. Personal Consumption ExpendituresClick on graph for larger image.

This graph shows MEW(1) as a percent of Disposable Personal Income vs. the Year-over-year change in Real PCE. An example of the use of this graph is presented as Exhibit 5 in this paper from Wells Fargo: Housing Bust Without A Consumption Bust???. The author writes:

Exhibit 5 shows virtually no relationship between real consumption growth and MEW. As MEW exploded between 1998 and 2003, consumption growth declined from over 5 percent to about 2 percent. As MEW has collapsed since 2005 from about 10 percent to about 3 percent, consumption growth has remained remarkably healthy at about 3.5 percent. Despite the widespread belief that MEW is a primary driver of consumer spending and the fear that any decline in MEW would certainly cause a collapse in consumption, this exhibit suggest consumer spending has been and continues to be little impacted by MEW.
First, although the author is correct that there is no clear correlation between MEW and the change in real PCE during the period presented, perhaps he has it backwards; maybe the reason PCE was in the normal historical range for the last few years was because of MEW.

Of course the graph is flawed too. The author uses the YoY change for PCE (introducing a lag in PCE) and compares it to the current quarterly value for MEW (even though studies have shown MEW is spent over a number of quarters following extraction). This error in graphing led the author to this conclusion:
As MEW has collapsed since 2005 from about 10 percent to about 3 percent, consumption growth has remained remarkably healthy at about 3.5 percent.
If declining MEW was going to have an impact on PCE, we wouldn't expect it to happen until several quarters after MEW declined. And since the YoY real PCE presentation introduces a mathematical lag, we would expect the graph to show a decline in PCE significantly after MEW declines - so the author's conclusion regarding "remarkably healthy" PCE while MEW is "collapsing" is, at best, premature.

Another perspective comes from Dallas Federal Reserve senior economist John V. Duca (written last November): Making Sense of the U.S. Housing Slowdown
The limited U.S. econometric evidence indicates that the strong pace of MEW may have boosted annual consumption growth by 1 to 3 percentage points in the first half of the present decade.[8] This implies that a slowing of home-price appreciation into the low single digits might shave 1 to 2 percentage points off consumption growth and 0.75 to 1.5 percentage points from GDP growth for a few years.

While these estimates provide an idea of housing’s potential economic impact, considerable uncertainty exists about how much a slowdown in MEW might restrain consumption growth. Key issues include how much home-price appreciation might slow, how much the deceleration would affect MEW and how much a slowdown in MEW would restrain consumer spending.
Consumption Share of GDP
Dr. Duca presents this graph (as chart 4). Duca suggests that the recent increase in PCE as a percent of GDP might have been driven by MEW - and declining MEW might "shave 1% to 2%" off consumption growth for "a few years".

Duca's paper also includes references to much of the recent research on the impact of MEW.

Now we just have to wait for the "unexpected" decline in Q2 PCE!

(1) MEW calculations courtesy of James Kennedy, and are based on the mortgage system presented in "Estimates of Home Mortgage Originations, Repayments, and Debt On One-to-Four-Family Residences," Alan Greenspan and James Kennedy, Federal Reserve Board FEDS working paper no. 2005-41.