U.S. CONSUMER OUTLOOK FOR THE HOLIDAYS AND 2008

With Thanksgiving come and gone and the first surge of holiday shoppers having made their way home from the malls and internet cafés, it is a timely occasion to assess the outlook for American consumer spending over the final weeks of 2007 and into the New Year. The kick-off to this year’s holiday season has shown consumers’ willingness to come out for a bargain. The number of people out shopping over the weekend following Thanksgiving was up 4.8% according to estimates by the National Retail Federation, but the amount spent by each customer was down from last year, in part reflecting discounting by retailers in order to draw customers in. Higher gasoline prices and restrained credit have combined with a severe housing market contraction to put the squeeze on American’s wallets heading into the New Year. Given the number of headwinds facing consumers, we expect a slower pace of growth of holiday retail sales over the final weeks of the year. In 2008, resetting mortgage rates and still falling home prices will continue to put the pinch on consumers, pushing expenditure growth to its slowest pace since 1991.

A season to spend

The holiday shopping season unofficially began on Black Friday, the day after Thanksgiving and traditionally the day when retailers begin to turn a profit. The Monday following the Thanksgiving weekend has been named Cyber Monday, reflecting the increasing importance of the internet in driving holiday sales. The holiday season accounts for approximately 20% of retailer’s annual profits and is for many retailers the focal point of their year. In 2006, over 22% of department store type merchandise sales occurred in the months of November and December, with 13% of annual sales in December alone.(1) Stores that cater most directly to gift buyers are the most dependent on the holiday season. Jewelry stores make close to a third of annual sales in the last two-months of the year. Men’s and women’s clothing are also popular gifts, with men’s clothing stores making close to 15% of annual sales in December and women’s clothing stores only slightly behind at 13%. In contrast, stores that cater most directly to new homebuyers, such as building materials and hardware stores, are the least dependent on the holiday shopping season with less than 7% of sales in December.

Our outlook this holiday season is tempered by the ongoing correction in the housing market, the impact of the credit crunch and indications that after months of bad news, consumer confidence is finally starting to wane. Existing home prices, which were down 5.1% on a year-over-year basis in October, already appear to be having an impact on consumer purchases of big-ticket home related items. As one might expect, sales at building materials and supply stores have been the worst hit and are down 3.1% on a year-to-date basis. Sales at furniture and home furnishing stores, which had held up fairly well in the first half of the year, have also shown weakness recently, falling in each of the last three months and up only 1.9% in the first ten months of 2007, compared to 9% growth in the same period in 2006. Similarly, sales of household appliances have weakened significantly over the course of the year and were down nearly 5% on a year-over-year basis in the third quarter.

In addition to falling home prices, the credit crunch has withstood three rounds of Fed rate cuts and continues to tighten the screws on consumers. The spread between government T-bills and the 3-month London Inter-Bank Offer Rate (LIBOR) has expanded to its highest level since the initial subprime financial shock in August. LIBOR has risen more than 25 basis points since the beginning of November, even as T-bill yields have fallen. The risk-aversion of financial intermediaries has increased along with financial losses and has cut into bank’s willingness to lend. As lending standards have tightened, the ability of consumers to finance their holiday purchases has fallen, which will impact most strongly on purchases of big-ticket items.

The credit crunch and increased uncertainty about the U.S. economic outlook has not escaped consumers and indexes of consumer expectations have fallen significantly in the last two months. Declining home prices and restrained consumer credit are likely to result in a slowdown in retail sales growth through the remaining weeks of 2007. We expect nominal December retail sales excluding non-seasonal purchases to grow by 3.0% from last year, its slowest pace in five years (see graph above). Real personal consumption expenditures are expected to grow by around 2% on a year-over-year basis in the fourth quarter, the slowest rate since 2002.

Ringing in the New Year

Taking a look beyond the holiday season, the slumping housing market presents the key risk to the American consumer in 2008. The contraction in U.S. housing has followed 14 years of nearly uninterrupted growth. Besides small downturns in 1995 and 2000, annual average housing starts rose in every year from 1991 to 2005, leading to a doubling in the level of starts from 1.0 million in 1991 to over 2.0 million in 2005. The expanding housing market contributed strongly to rising household wealth. Between 1991 and 2005, home prices rose by close to 140% and households real estate assets rose by 185%. Gains in housing wealth helped to offset financial losses following the tech-bust of the late 1990s and pushed even higher following the drop in interest rates after the 2001 recession. Through this period, the importance of the American consumer to the U.S. economy also grew, rising from 67% of real GDP in 1991 to 71% by 2005.

Economists have tried to quantify the impact of housing wealth on consumer spending and have come to varying conclusions. Estimates of the impact of a one dollar rise in housing wealth on consumption range from 2 cents additional spending to up to 10 cents in each period following the increase in wealth. A mid-range estimate of 5 cents more spending per 1 dollar rise in wealth is cited widely by economists. So far this year real existing home prices have fallen 3.5% from the same period in 2006. Home prices are likely to get worse before they get any better. At the current pace of sales and with no new additional listings it will take 10.5 months to clear the inventory of existing homes on the market. The inventory build up has been swift and has more than doubled from 5.1 months of supply in January of 2006. Even if new housing completions remain at current levels, it will require a substantial pick up in home sales to drawdown the build up in inventory. With mortgage payment resets and foreclosures not set to peak until mid-2008, home prices through 2008 could suffer real year-over-year declines in the range of 5-10%. While there is some uncertainty surrounding the lag time between changes in housing wealth and changes in consumer spending and the total impact of falling prices may take place over more than one year, the negative wealth effect is likely to take a noticeable bite out of consumer spending over the course of 2008.

Financial market innovation during the long period of housing market growth resulted in an expansion of choices for households wanting to take advantage of the increased value of their homes. Withdrawals from home equity can be made when a home is sold and capital gains are realized, or through mortgage refinancing and home-equity lines of credit. The latter two forms of home equity withdrawal are commonly referred to as active Mortgage Equity Withdrawal (MEW). There is much debate among economists as to the amount by which MEW in itself supports increased consumer spending. On the high end are estimates that suggest that every dollar of active MEW supports an additional 50-62 cents of consumer spending. Active MEW, as calculated by analysts at the Federal Reserve Board has fallen by over $100 billion from 2005 to the first half of 2007. With total consumer spending of $9.7 trillion, this drop in MEW is equivalent to reducing consumer spending growth by 1.3 percentage points over this period. Assuming that MEW will continue to fall along with home prices, this presents a significant downside risk to the outlook for consumer spending in 2008.

On the other side of the debate are economists that contend that mortgage-equity withdrawal is simply a vehicle used by households to fund the extra consumption they are afforded by their rising level of wealth. For these analysts, it is not MEW that causes spending but a desire to increase spending (through the wealth effect) that increases MEW. In other words, the rise in MEW is explained by a growing number of consumers who wish to increase spending as a result of rising wealth and for whom cashing out some of the equity from their homes is the best way to do so. Under these assumptions, the impact of housing price changes on consumption is less significant. A decline in housing wealth due to falling home prices will still result in a reduction in spending, but the impact is felt completely through the wealth effect.

While it is easy to buy into the idea that access to MEW simply made it too easy for consumers to use their homes as cash machines (essentially burning a hole in their pockets), conventional economic theory leads us to believe that what is more important to determining spending is a consumer’s real net worth. As a result, we believe that the decline in consumer spending due to the falling home prices is unlikely to be as high as the most pessimistic estimates would assume. Nonetheless, it does give support to the higher estimate of the wealth effect at 10 cents. Adding up the impact of negative wealth effects and falling MEW on consumer spending, we expect declining home prices to take close to half a percentage point off consumer spending growth over the next year.

The prevalence of adjustable rate mortgages (ARMs) is yet another factor that has turned on consumers as the housing market collapsed. According to estimates by Global Insight, the share of adjustable-rate mortgages in new mortgage originations increased from 17% in 2002-03 to over 31% in 2006. Moreover, the vast majority (over 75% by most estimates) of ARMs in 2006 were issued to subprime borrowers. Many of these loans are now in the process of resetting to higher interest rates. Since many adjustable-rate mortgages were given with an initial 1 or 2 year “teaser” interest rate that was lower than the market rate, the combined impact of the rise in mortgage rates since the mortgages were issued and the reset to market interest rates means that some loans could face resets of more than 4 percentage points.

Estimates of the increase in interest payments as a result of ARM resets range from $10 billion to $35 billion sequentially in 2007 and 2008, with the peak in reset volumes not to occur until spring 2008. A $35 billion increase in mortgage payments is equivalent to adding more than 2% to annual household debt service payments and shaving about 0.3% off annual personal disposable income. Already the impact of ARM resets has lead to a rise in delinquency and default levels. The percent of subprime ARMs more than 60 days overdue rose over 2 percentage points in the third quarter of 2007 to 14.3%. Based on past trends, delinquency rates of subprime mortgages are not likely to peak until early to mid 2008. Due to the prevalence of subprime loans in ARM resets also being weighted towards 2008, the delinquency rate of subprime mortgages originated in 2006 could reach rates as high as 18-20%. This rise in delinquencies will put upward pressure on the number of foreclosures. Though only about a quarter of the homes that enter foreclosure proceedings end in a forced sale, this will add to the downward pressure on home prices in 2008 especially in markets that are already depressed.

An initiative by the U.S. Treasury to broker a deal between loan service providers and mortgage investors may lessen the blow of future ARM resets. The deal would allow qualifying households who face a first payment reset between January 2008 and July 2010 to freeze their existing mortgage rate for a period of five years. If the plan is enacted as it is proposed it should forestall some of the pressure on households facing resets. However, as at this point the plan is only a preliminary set of guidelines, it is yet too early to quantify its impact. A worst case scenario would see ARM resets taking a further 0.3% from consumer spending growth. With the negative wealth effect and reduced MEW taking a half percentage point away from consumer spending growth and mortgage resets taking a further 0.3, the declining housing market is likely to take upwards of 0.8 percent from consumer spending growth over the course of 2008.

Employment growth to slow, income growth to continue

With falling home prices across much of the country and the impact of mortgage resets still approaching, any hope for continued consumer spending depends on the prospects for personal income growth. Weakness in the U.S. economy next year will also result in a slowdown in employment growth. Non-farm employment rose by 1.8 million jobs over the first 11 months of 2007, down from 2.5 million over the same period in 2006. Employment in residential construction has fallen by 28,000 jobs so far this year and although the use of illegal immigrant workers distorts the job loss picture somewhat, presuming the 63,000 jobs created in 2005 and 2006 were legal workers leads us to believe that housing-related job losses are still ahead of us. The U.S. economy will likely add fewer than half a million jobs over the course of 2008, down significantly from recent years. Nonetheless, the unemployment rate is currently sitting at 4.7% and while we expect upward movement to 5.3% in 2008 as a result of the slowdown in hiring, labour markets will remain relatively tight, leading to continued wage and income gains. Real personal disposable income growth is expected to grow by close to 2.0% in 2008, down from 3.0% rise in 2007. These trends augur for the slowdown in consumer spending in our forecast but not the consumer-led recession that some are predicting.

As this year winds down and a new one sets in, signs of consumer vulnerability to falling home prices have already started to show in slowing sales of durable household items such as furniture and appliances. The impact of the credit crunch, high gasoline prices and falling home prices is likely to result in a slowdown in the pace of year-over-year holiday retail sales to 3.0% in December of this year. Looking into 2008, the credit crunch, falling home prices and resetting ARMs will continue to squeeze the consumer’s pocket book and slow real consumer spending to a sluggish pace of 1.7%. The slowdown in consumer spending as a result of the housing market’s contraction will likely push GDP growth to a tepid 1.8% over the course of the year, not quite a recession, but it may feel like one to many Americans.

James Marple, Economist
416-982-2557

(1) Using the Census Bureau’s GAFO definition, which includes furniture and home furnishings, electronics and appliances, clothing and accessories, sporting goods, hobby, book, music, general merchandise, office supplies, stationery and gift stores.