DAILY FINANCIAL MARKET COMMENT 3/14/04 Goldman Sachs Economics
* The Fed's flow-of-funds tables show a further pickup in mortgage equity withdrawal (MEW) to a record $640 billion, or 7.4% of personal disposable income, in 2004. Although cash- out refinancing was down somewhat, this was more than offset by a near-doubling of home equity borrowing as well as further strength in housing turnover. We believe MEW has been one key reason for the strength of household spending. As interest rates rise, we expect this boost to turn into a drag.
Mortgage Equity Withdrawal Picks Up Further in 2004
According to the Fed's flow of funds tables, US households stepped up their extraction of home equity in 2004 even further. We estimate that our measure of total mortgage equity withdrawal rose to $640 billion, or 7.4% of personal disposable income, from $528 billion in 2003. We define MEW as the flow of new borrowing secured on existing homes and calculate it as total mortgage borrowing less the amount required to finance the construction of new homes, which we estimate by the dollar volume of new home sales multiplied by the average loan-to-value ratio on new homes.
The table below breaks down our MEW estimate into three components: (1) cash-out refinancing, as estimated by Freddie Mac, (2) home equity borrowing, as estimated in the Fed's flow- of-funds tables, and (3) equity extracted in the housing turnover process, which we calculate as a residual.
Mortgage Equity Withdrawal ($, billions) Year Total Cash-Out Home Equity Turnover- Refinancing Borrowing Related MEW 1993 $79 $39 $-8 $47 1994 88 29 13 46 1995 74 22 17 35 1996 113 34 31 47 1997 110 39 41 30 1998 175 72 29 73 1999 242 71 40 131 2000 229 60 90 79 2001 317 136 26 156 2002 452 171 65 217 2003 528 224 102 202 2004 640 181 196 262 Source: Federal Reserve Board. Freddie Mac. Our calculations.
The table shows that the increase in MEW comes despite a significant decline in cash-out mortgage refinancing volumes as estimated by Freddie Mac. This decline is consistent with a sharp drop in the Mortgage Bankers Association's refinancing applications index to 2379 in 2004 from 4982 in 2003. It probably reflects the fact that mortgage rates, while still very low, were no longer declining in 2004, and the refinancing decision depends on the change in interest rates than their level.
However, slower refinancing was more than offset by a near- doubling in home equity loans and another pickup in turnover- related equity extraction. Although the latter series is estimated as a residual, the increase is quite plausible given the acceleration in both existing home sales and house price inflation. Existing single-family home sales rose to 5.91 million from 5.44 million, and home price inflation climbed to 10.2% from 9.0%. (The home price inflation data refer to the Office of Housing Enterprise Oversight's series for purchase transactions only, on a fourth-quarter/fourth-quarter basis.)
We view MEW as an important variable because it provides a convenient way to track the housing wealth effect. Of course, this assumes that a significant proportion of MEW goes toward financing personal consumption and home improvement spending. This assumption is consistent with the fact that, over the last 30 years, there has been a strong positive correlation between MEW and household spending (both measured relative to disposable personal income). The only major exception occurred during the buildup and subsequent bursting of the 1990s equity market bubble, which drove spending growth well beyond -- and subsequently well below -- the level implied by the change in MEW. Combined with the big pickup in MEW, this correlation suggests that much of the recent strength in consumer spending has been due to a powerful housing wealth effect.
As interest rates rise, we expect the boost from MEW to turn into a substantial drag. Even if house price inflation merely slows -- but does not turn negative -- in response to higher rates, homeowners will experience far slower growth in their wealth than they have in recent years. Assuming that there are institutional and psychological limits on the proportion of newly created home equity that households can and will extract, this would imply a substantially lower pace of MEW. Unless it is offset by other factors such as a big stock market rally, such a decline in MEW would likely put substantial upward pressure on the personal saving rate. We believe the impact would be gradual, but nonetheless significant.
There are both upside and downside risks to this view of a gradual deceleration. The downside risk is that the housing market is looking more and more like an outright bubble. House prices in the country as a whole are probably still at the upper end of the reasonable range. Our back-of-the-envelope model of median home prices as a function of median household incomes and mortgage rates currently suggests an overvaluation of about 10%, a moderate number by the standards of countries such as the UK or Australia and in view of the imprecision of any fair-value estimate for house prices. However, this estimate is based on an average mortgage rate of about 6%, and we expect rates to rise. If rates were to increase 100 basis points, this would reduce fair value for house prices by about 8% according to our model. At an estimated overvaluation of 18%, the risks of a sharper downturn would increase significantly.
On the other hand, there is an upside risk -- at least for the near term -- that stems from the fact that the housing sector has a long history of surprising on the upside and a bullish mentality now seems to be well entrenched among homebuyers. If the upside surprises continue, the risk would be that the consumer stays stronger than expected in the near term. Given the strength currently seen in the corporate sector, rates might then have to rise significantly further than currently discounted, in order to keep the economy from overheating.
Jan Hatzius