Will US Housing Rust or Bust?

Until early this year, I steadfastly argued that US housing markets were far from bubble territory, supported as they were by strong fundamentals and buttressed by what I saw as attractive valuations,' at least by my admittedly crude metrics.

With fundamentals deteriorating and housing no longer cheap, in February I switched to the froth' camp, arguing that house price appreciation and housing activity had peaked. In the event, I was too early: house price appreciation is still rapid, and the evidence for a peak is mixed. Although housing starts have declined from February levels, housing demand is still booming, with traditional sales and lending growth at or near records, and condominium sales accelerating. So it's time to mark myself to market. Is the call wrong?

Call me stubborn, but I'm more convinced than ever that the call was right but merely too soon. Housing markets are verging closer to bubble territory, as fundamentals supporting home prices have slipped further and valuations have become extended.

Subscribe to MoneyWeek

Subscribe to MoneyWeek today and get your first six magazine issues absolutely FREE

Get 6 issues free

Sign up to Money Morning

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Sign up

However, in my opinion, fears of a bursting bubble are overblown; prices already seem to be decelerating and I stick to my longstanding view that they will rust, not bust. Likewise, I think the coming decline in housing activity will be gradual, but I'll concede that the longer it is deferred, the sharper the likely decline.

Let's start with the fundamentals. Rapidly rising home prices have made housing less affordable in many metro areas. Not surprisingly, measures of affordability from the National Association of Realtors, by Census region, suggest that it has been declining for the past two years in the high-priced West and Northeast and much of that in a period of declining interest rates and improvement in median family income.

Prices are the culprit: for example, according to Harvard's Joint Center for Housing Studies, the ratio of median home prices to household incomes in California now tops out at 9 to 1. And it's no surprise that affordability issues are most acute in Los Angeles, Miami and New York, where according to the Office of Federal Housing Enterprise Oversight (OFHEO), house prices have risen by 105%, 95%, and 77% in the past five years, compared with 50.5% nationwide.

Moreover, the incipient backup in interest rates will add to the cost of ownership for both would-be homebuyers and homeowners with adjustable-rate mortgages (ARMS), eroding demand and prices. Clearly, the growing use of ARMs and the shift to interest-only mortgages by mid- to lower-income households enables borrowers to reach for otherwise unaffordable housing as rates back up.

Indeed, according to Morgan Stanley's MBS strategy team, ARM originations jumped to about $1.4 trillion in 2004 from just $330 billion in 2001. In addition, some ARMs, such as so-called negative-am' ARMS, which defer the increase in payments for a while when rates go up by adding it to the principal, merely defer the pain of rising rates with interest. For their part, bank regulators are warning lenders about the risks in high loan-to-value, limited documentation and no documentation, interest-only, and third-party generated home-equity loans.

Some bubble enthusiasts thus see an immediate threat from a 1980s-style ARM squeeze,' in which upward resets from teaser' rates will squeeze both home prices and discretionary income. However, the ARM market has changed dramatically in the past five years, shifting to hybrid' loans with longer fixed-rate periods. We estimate that it will be three years or more before the first interest rate reset on most hybrid ARMS, so rising rates won't affect such borrowers for a while. In contrast, in the late 1980s and early 1990, most ARMs had one-year resets.

However, other factors are more immediate negatives for housing demand and prices. First, there are signs that the healthy pent-up demand of the past decade is beginning to wane and give way to more speculative buying. Homeownership rates, especially for households over 45 and with income above the median, have begun to decline for the first time in more than twenty years. As I see it, this suggests pent-up demand is exhausted.

And two indicators the share of home sales that are for investment, and housing turnover hint strongly at a pickup in speculative activity. HMDA (Home Mortgage Disclosure Act) data indicate that 13-15% of sales are for investment. These data show that some local markets have extremely high investor shares (Las Vegas 24%, Orlando 19%) and those venues are clearly hot.'

Meanwhile, housing turnover, proxied by sales of existing homes in relation to the housing stock, approached 5% recently, or double the share a decade ago, when home price appreciation began to quicken. Somewhat ominously, that turnover ratio peaked at a similar level in 1978, and that marked the crest of the 1970s housing boom.

Moreover, revised data suggest that valuations are more extended than I thought three months ago. My valuation metric a crude price-earnings' ratio for housing has until recently suggested that valuations were well below their 1980s peaks. It compares home values with owners' rental income to approximate how much homeowners are paying for a dollar of such income. I proxy home values as the product of the OFHEO home price index, the number of single family owner-occupied houses, and $64,000 the value of a typical house in 1980, the base year for the price index.

I proxy the earnings' from housing just as the Commerce Department does, using rental income of persons a measure that nets from gross rents such charges as mortgage interest and property taxes. This denominator is far from perfect, because it contains imputations and makes assumptions about the effective interest rate or outstanding mortgages and transactions costs. Reflecting new data on space rent, and repair and refinancing costs, in the just-released revision to the National Income and Product Accounts, the Bureau of Economic Analysis (BEA) revised rental income down by $40 billion, which boosted the P/E ratio by 35%. In addition, our MBS strategy team calculates that the BEA's average effective interest rate may be too high by about 44 bp, trimming another $36 billion from rental income and further boosting valuations close to their all-time peak.

Whether or not you adjust for that apparent discrepancy in calculating interest expense, incorporating these new data in my P-E ratio shows that valuations have moved up significantly and there are strong hints of overvaluation, although the ratio is still below past peaks. Using adjusted data for rental income, the ratio has risen from 100% to 137% in the past three years (the level of the ratio seems high, but the units of measurement are somewhat arbitrary). That's 51% above the level of the mid-1990s, when housing by this metric was undervalued, but well below the 250-300% level of the 1980s a level that clearly foreshadowed the housing bust of 1989-93.

All this suggests that house price growth is likely to decelerate swiftly, perhaps to 2 - 4% in the next two years, from 12.5% in the year ended in the first quarter. (It's worth noting that even so-called matched-sample' price gauges such as the OFHEO measure may overstate home price appreciation, and some pricing metrics are meaningless.)

Such a deceleration will mean less support for spending from rising housing wealth, consistent with my view that personal saving will begin to rise. While the empirical evidence isn't conclusive, some relationships suggest that changes in housing wealth have an immediate effect on spending.

Financial market participants can't seem to decide what frothy housing markets imply for interest rates and monetary policy. Some think the Fed must tighten to deflate a feared housing bubble, while others think that vulnerable housing markets will stay the Fed's hand.

In fact, while Fed officials are obviously concerned about speculation in housing markets and risky lending practices, they have made it clear that monetary policy isn't directly aimed at pushing home prices in either direction. To quote Fed Governor Don Kohn: "direct influence [of house price movements] on monetary policy is limited. They are important to us in so far as they affect the macro economy." A housing boom that fuels inflation expectations is obviously a policy concern.

But financial stability is also of concern to policymakers, and they may well take comfort from anecdotal evidence that imbalances in housing markets are beginning to cool off. For example, to quote the latest Beige Book: "Residential real estate activity remained robust overall but showed a few signs of cooling in some Districtshousing activity and home price appreciation in Massachusetts moved from hot' to normal', housing inflation slowed in New Jersey, and the condo market in Manhattan was less frenzied than in the spring." That home prices are now decelerating in hot' markets like San Diego, Boston and New York gives me encouragement that the rust scenario is underway.

That's not to say that housing and housing finance are on a smooth glide path for a soft landing.' Indeed, in my view, mortgage lenders and investors are more at risk from rusting home prices than are consumers. They face the loss of income and possibly principal as markets cool and credit quality slowly deteriorates. The acid test will come next year as rates continue to rise. Consumers aren't immune, but the chance of a serious ARM squeeze' currently seems small.

By Richard Berner, Morgan Stanley Economist, as published on the Global Economic Foru