“I can calculate the movement of the stars, but NOT the madness of men.”
– Sir Isaac Newton, after losing a fortune (Â£20,000) in the South Sea bubble.
I’ve been meaning to write about the impending real estate crash for quite some time, but until I get around to it, this piece is worth a read. Roubini is a Professor of Economics at the Stern School of Business at New York University, and the work below is taken from his blog. Though the prose is somewhat awkward, anyone who produces “seminal work in international macroeconomics, global macro policies, financial crises in emerging markets and their resolution, and the reform of the international financial architecture” deserves at least a few minutes of your time…
“The Biggest Slump in US Housing
in the Last 40 Years”…or 53 Years?
By Nouriel Roubini
Aug 23, 2006
The Biggest Slump in US Housing in the Last 40 Years: These are not my views but those of the Toll Brothers, the famous luxury McMansions homebuilders, as CNN reported last week. Also, as reported by the WSJ today: In his 40 years as a home builder, Mr. Toll says, he has never seen a slump unfold like the current one. “I’ve never seen a downturn in housing without a downturn in employment or… some macroeconomic nasty condition that took housing down along with other elements of the economy,” he says. “This time, you’ve got low unemployment, you’ve got job creation, you’ve got a stable stock market and relatively low interest rates.”. This followed last week’s CNN headline: “Builder: Oversupply slump worst in 40 years. Toll Brothers slashes outlook on new homes as orders plunge and revenue misses forecasts” Indeed, yesterday’s sharply falling profit results from the Toll Brothers confirmed their view that this is the worst housing slump in decades. Similarly, Angelo Mozilo, the CEO of Countrywide — the country’s largest independent home mortgage lender – recently stated: “I’ve never seen a soft-landing in 53 years, so we have a ways to go before this levels out. I have to prepare the company for the worst that can happen.” So, effectively the only debate now is whether housing conditions are the worst in the last 40 years or in the last 53 years. So much for the bullish soft-landing wishful thinking coming out of Wall Street these days…
Of course, the message from the Toll Brothers and Countrywide is like the proverbial canary in the mine that is reflective of an ongoing rout — calling it slowdown or slump is a misnomer by now — in the US housing market. Every possible indicator of the housing sector that has been coming out in the last few weeks — and I will discuss their details below – suggests that the housing market is in free fall. And today’s figures on existing home sales and unsold homes say it all; as Bloomberg concisely headlined this morning: U.S. Existing-Home Sales Tumble; Unsold Inventory Is Highest in a Decade
At this point there no doubt on whether the housing sector is contracting — real residential investment fell at the annualized rate of 6.4% in Q2. The first derivative of the housing market is clear and negative today and looking ahead for the next few quarters. There is not even a debate about the second derivative of the housing market as any estimate out there suggests that the housing sector will contract at a faster rate in Q3 and Q4 than in Q2. Some official estimates that I have seen suggest that real residential housing will contract at 10% – rather than the Q2 6.4% in the next two quarters. My own estimate — based on a reading of the coming data — is that, actually, the contraction is more likely to be of the order of 12-15% annualized rate in the next several quarters. So, the only remaining scary question is about the third derivative of the housing sector and at which point — in terms of quantities and prices — the housing market will bottom out.
I have also argued before that the effects of housing on US economic growth and the role of housing in tipping the US economy into a recession in early 2007 are more significant than the role that the tech sector bust in 2000 played in tipping the economy into a recession in 2001. There are three reasons:
- The direct effect of the fall in residential investment in aggregate demand will be as high as the effects of the fall in real investment in the 2000-2001episode. Then, real investment fell by about 2% of GDP. This time around the fall in residential investment alone — let alone the role other components of real investment, such as software and equipment, that are already falling in Q2 — will be as large as residential investment could fall from the peak of about 6.2% of GDP (the highest level since the 1950s) to as low as 4% of GDP at the bottom in 2007.
- The wealth effect of the tech bust was limited to the elite of folks who had stocks in the NASDAQ. The wealth effect of now falling housing prices — yes median prices are starting to fall at the national level – affects every home-owning household: the value of residential real estate has also increased to 48.5% of household wealth in 2006 from from 38.7% in 1996. Also, the link between housing wealth rising, increased home equity withdrawal (HEW) and consumption of durable and non durables is very significant (see RGE’s Christian Menegatti brief on this), much more than the effect of the tech bubbles of the 1990s. Last year, out of the $800 billion of HEW at least $150 or possibly $200 billion was spent on consumption and another good $100 billion plus went into residential investment (i.e. house capital improvements/expansions). It is enough for house price to flatten — as they already did recently — let alone start falling – as they are doing now since they are beginning to fall in major markets — for the wealth effect to disappear, the HEW dribble to low levels and for consumption to sharply fall. Note that this year there will be large increases in the borrowing costs for $1 trillion of ARM’s while this figure for 2007 will be $1.8 trillion. Thus, debt servicing costs for millions of homeowners will sharply increase this year and next.
- The employment effects of housing are serious; up to 30% of the employment growth in the last three years was due directly and indirectly to housing. The direct effects are job lost in construction, building materials, real estate brokers and sales agents, and employees of the mortgage finance industry. The indirect effects imply that the role of housing is even larger than 30%. The housing boom led to a boom in consumer durables spending on home appliances and furniture. Indeed, in Q2 real consumption of such goods was already negative: as you have less new home built and purchased and less old homes refurbished and expanded, you get less purchases of home appliances and furniture. There are also other indirect effects of the housing bust on employment, even on the purchases of motor vehicles. Indeed, the current auto sector slump is not unrelated to the housing slump. As the Financial Times put recently, the sharp fall in the sales of Ford’s pick-up trucks is related to the housing slump as such truck are widely purchased by real estate contractors. And indeed in Q2 real consumer durables (that include both cars, home appliances and furniture all related to housing) already fell, consistent with the view that we have now have a glut in the stock of consumer durables (durables consumption has a investment-like nature to it as such goods last for a long time). Thus, as housing sector slumps, the job and income and wage losses in housing will percolate throughout the economy.
How bad are the signals coming from the housing sector? As a recent news headline clearly put it: it is simply UGLY. Indeed, all the indicators from the housing sectors – including the latest housing starts and the homebuilders (NAHB) forward looking business conditions – indicate a housing sector that is literally in free fall. New home sales started to fall since the beginning of 2006 and in some regions they are down over 30% relative to a year ago. As Bloomberg summarized today the new housing data: “Sales of previously owned homes in the U.S. fell more than expected in July, resulting in the biggest supply of unsold homes in more than a decade, as higher mortgage rates discouraged would-be home buyers.. Purchases declined 4.1 percent last month to an annual rate of 6.33 million, the lowest since January 2004, from 6.6 million in June, the National Association of Realtors said today in Washington. Sales fell 11.2 percent compared with a year earlier.” Indeed, the number of unsold homes and the ratio of unsold homes to new home sales has therefore risen sharply to over 5.5 months of supply. Similarly the ratio of unsold homes to existing home sales has also sharply increased. These are clear indicator of a glut of unsold homes in the market. Housing starts are also sharply down elative to a year ago and expected to fall further over the next few quarters. Note also that, while overall mortgage applications are still up in the latest figures published today, due to sustained refinancing applications, applications for purchase applications have fallen 1.0% during the last week, this being fifth fall in the last six weeks. Moreover, there is a large amount of evidence that suggests increasing cancellation of initial mortgage applications, as the slump in the housing market and in the economy is now scaring households considering buying a home. Thus, the official data on purchase mortgage applications are very likely to exceed actual home sales.
More generally, note that when demand for housing initially falls relative to a glut of supply, the initial market response is not on price, as it is the case of financial market where prices adjust rapidly, but rather on the quantity of unsold homes and on how long unsold homes stay on the market. Housing prices, unlike financial assets, are sluggish. This market inventory adjustment eventually leads to lower prices once sellers realize that demand is low and that waiting is not going to help.
The housing market has thus followed so far the predicted various stages of adjustment to cycle driven by the initial housing bubble: initially a glut of supply of new homes as high prices (driven in part by speculative demand) led to high and excessive production of new homes; then a fall in demand as speculative high prices and rising rates made the purchases of housing less affordable to many; then, the ensuing inventory adjustment — an increase in unsold homes. Then, the reduction in the production of new homes — lower housing starts — as homebuilders with falling revenues and profits and lower expected demand finally reacted to the growing glut of unsold inventories. Indeed, the value of home builders’ shares on the NYSE has fallen by almost 50% relative to a year ago. Finally, we have now a price adjustment in two directions: a) an increase in rents as housing affordability fell since more and more households could not afford to pay the speculative prices of existing and new homes; this increase in rents is now correctly jacking up owner equivalent rent and increasing headline CPI inflation; b) the beginning of a fall in actual housing prices as the glut of unsold homes is now putting downward pressure on actual prices. (for more on recent indicators of the housing bust see the RGE Monitor cluster of readings on housing indicators)
The evidence on falling home prices is now becoming clearer. Since the end of World War II, there has never been a year on year fall in housing prices. There have been instead several quarters in which housing prices declined. Of course in some regions where there were housing busts prices declined for a while: in Texas during the housing bust of the mid 1980s that led to the S&L crisis; in California in the early 1990s following the recession in that state; in Boston in 1990. Those episodes were all associated with the housing bust that was related to the 1990-1991 recession So, you do not need a persistent year-on-year fall in median housing prices to have a housing bust; such bust can occur even if prices are flattening or falling in some regions, but not nationally. Moreover, such regional bust can be associated with national recession, as in the 1990-91 episode. So, the fact that the latest housing bubble was concentrated on the two coasts (North East all the way to Florida; and West Coast, especially California) does not mean that the coming housing bust in these regions will not have national macro effects. For one thing, the value of the housing stock in those two regions is close to 50% of the total housing stock given the bubble of recent years. Thus, a housing bust in the two coasts can and will have macro effects.
Indeed, today the National Association of Realtors reported today that the median price of an existing home rose only 0.9 percent in July from a year ago. So, housing prices are practically flat at the national level. Worse, relative to a year ago housing prices have already fallen in the North East (-2.1%), Mid-West (-0.6%) and the West (-0.3%). So, not only housing prices are falling in the bubbly two coast; they are also starting to fall in the Mid-West, the region where the conventional wisdom was that there was no housing bubble. The fact that home prices are falling in the Mid-West where prices did not skyrocket in the bubble years is a scary signal of how much the housing bust and glut in supply will lead to a sharp fall in housing prices in the quarters ahead with painful effects on the wealth, and thus consumption, of households. You can expect falling median housing prices, on a year-on-year basis, at the national level starting this month of August: indeed, today’s figures on the glut of unsold homes – much larger than in the housing bust of the early 1990s – are only consistent with a highly likely actual fall in home prices in the months ahead and throughout most of 2007. Note also that, on an inflation adjusted basis, real home prices (relative to the CPI index) are already falling at a 4% plus rate.
Also, as noted by Dean Baker: “current house price indices are failing to pick up the full decline in prices because they miss the various concessions (seller paid closing costs, buyer-side realtor bonuses, and seller subsidized mortgages) that sellers often use to move their houses.”
Even more ominously, futures markets now expect that house prices will fall during 2007. Following the lead and prodding of Robert Shiller — the maverick Yale professor who predicted the 2000 stock bust and is now predicting a housing bust – the Chicago Mercantile Exchange opened this spring a new futures market for house prices in ten U.S. cities. While this market is very new and still relatively illiquid, it is now predicting that U.S. house prices will fall in 2007 at the national average level, for the first time in over fifty years. The index of this futures’ market for the entire US is projecting a 5% price fall in 2007. And the futures contracts for individual cities show expected declines in housing prices even larger than 5% for Miami, New York, Boston, San Francisco, Boston, San Diego and Las Vegas.
The likely fall in median home prices in 2007 may actually turn out to be larger than the 5% priced in the futures markets. In fact, one of the peculiar features of the latest housing cycle has been the presence of a large housing bubble: prices were going up well above economic fundamentals because of the speculative demand coming from expectations of increased housing prices that were feeding further speculative demand: “condo flipping” is the popular term for this speculative demand. Now that the bubble is bursting the fall in prices will be sharper than the one implied by medium term fundamantals as the initial price increase was due to a bubble that is bursting and leading to a fall in speculative demand: with prices now falling homeowners and speculators have no incentive to buy new homes as they expect prices to be lower in the future. So, an expected price fall leads to fall in speculative and fundamental demand and triggers actual larger than otherwise fall in actual prices. The speculative excess of a price bubble will now bring the bust of this price bubble. While the effect will be slower than in asset markets where prices adjust instantaneously (due to the sluggish nature of housing prices and their slow adjustment to increased inventories) eventually this price adjustment will occur – as it is now – and it will be very persistent over time. So, you can expect falling housing prices throughout most of 2007.
So, the simple conclusion from the analysis above is that this is indeed the biggest housing slump in the last four or five decades: every housing indictor is in free fall, including now housing prices. By itself this slump is enough to trigger a US recession: its effects on real residential investment, wealth and consumption, and employment will be more severe than the tech bust that triggered the 2001 recession. And on top of the housing bust, US consumers are facing oil above $70, the delayed effects of rising Fed Fund and long term rates, falling real wages, negative savings, high debt ratios and higher and higher debt servicing ratios. This is the tipping point for the US consumer and the effects will be ugly. Expect the great recession of 2007 to be much nastier, deeper and more protracted than the 2001 recession.
And the housing bust is not going to be only a US phenomenon. As I will discuss in another blog, housing bubbles festered in many other economies including many European ones. Thus, the combination of high oil prices, delayed effects of rising interest rates and slump of housing that is now leading to a US recession is a phenomenon that is common to many other economies, including several European ones. So, expect the same deadly combinations of three ugly bears (slumping housing, high oil prices and rising interest rates) to hammer Goldilocks and sharply hurt Europe and other economies in the world.
Thursday morning update:
This morning’s data on new homes sales, inventories of new homes and prices of new homes fully confirm and reinforce my analysis yesterday that this will be the worst housing bust – calling it slump is too mild – in decades. And since median home prices may actually fall on a year-on-year basis in 2007 – something that has not happened since the Great Depression of the 1930s – this may end up being the biggest housing bust in the last 75 years, not just 40 years as the Toll Brothers argue or 53 years as Countrywide argues. As reported by Bloomberg this morning:
New Home Sales in U.S. Fell to a 1.072 Million Pace by Bob Willis
Aug. 24 (Bloomberg) — New home sales in the U.S. fell last month and inventories rose to a record, raising the risk for the economy that the housing market slowdown will become more pronounced.
Purchases of new homes, which account for about 15 percent of the market, dropped 4.3 percent to an annual pace of 1.072 million, the Commerce Department said in Washington. Sales in the Midwest slumped to the lowest level in almost nine years. The median U.S. home price rose 0.3 percent from July 2005, the smallest rise since December 2003.
The weakening housing market is leaving builders with record inventory and little choice but to reduce prices at a time when profits are declining. Some Federal Reserve policy makers have said a slump in housing is one of their biggest concerns because refinancing, which provided homeowners with extra cash to spend, may dry up as home values decline.
“There is no question we have what could be called a housing recession,” said James O’Sullivan, a senior economist at UBS Securities LLC in Stamford, Connecticut, who forecast a drop to 1.075 million. “I don’t think we’ve seen the full effect of this yet on the overall economy.”
The decline in new home purchases follows yesterday’s report that showed a drop in sales of previously owned houses to the lowest in more than two years. The National Association of Realtors said existing home sales, which make up about 85 percent of all purchases, declined 4.1 percent to an annual pace of 6.33 million in July.