Thursday, February 28, 2008
real housing declines and price-to-income
but, in the meantime, the transition period has offered increasing inflationary pressure. and the effect on the housing decline in real terms is noted by calculated risk.
In nominal terms, the index is off 8.9% over the last year, and 10.2% from the peak.
However, in real terms, the index has declined 12.9% during the last year, and is off 14.6% from the peak.
Inflation is helping significantly in lowering real house prices. If prices will eventually fall 30% in nominal terms, then we are only about 1/3 of the way there. But if the eventual decline is 30% in real terms, then we are about half way there.
better than taking guesses at static targets, i think what we're witnessing is a normalization of the financing behind housing. house prices are of course about supply and demand. early in this decade, a combination of negative real rates and wall street initiative to expand securitization into the mortgage market changed the credit dynamic behind demand -- more buyers found they had more resources at their command for the same debt service payment/income level. increased demand drove up prices. higher prices signalled the supply side -- that is, homebuilders -- to expand operations, which they did.
i think one can say the mania took hold when expanded supply could not meet increasing demand -- prices continued to rise as financing and underwriting standards fell through the floor in an effort to keep the securitization machine growing and the fee revenue coming in to the banks, creating ever greater housing demand. speculation took hold in many markets, creating a new class of real estate "investor" who serviced multiple houses on the skimpy financing vehicles being offered by banks. many confused reasons for the surging prices were contrived, but in the final analysis this is 95% of what happened to home prices between 2001 and 2007.
the whole mechanism seized beginning in february 2007, when the first investors began to realize that default rates on mortgages originated in 2005 were not behaving normally -- the first sign that the shoddy underwriting standards of the boom had included in mortgage pools a significant slice of mortgagors who were not going to pay back their loans. as end investors began to shun securitized mortgages, banks found that they could no longer sell the mortgages they were making. without room on overstuffed balance sheets to take on the mortgages themselves (as they would have in the days before mortgage securitization), they essentially quit making mortgages of any kind that could not be sold on to fannie mae and freddie mac, quit making loans to lower quality mortgagors, quit making loans with payment-reducing features like negative amortization and teaser rates. with as many as 9 in 10 mortgage applicants that would have been approved in 2006 now being declined in some areas, demand fell precipitously and prices began to decline.
this is what i mean about the normalization of the financing behind housing. the tricky loans and blind underwriting that fed the financial engineering of the boom allowed a certain amount of income to support more house than it ever had before. the longstanding national relationship of median house price to median income since the advent of the 30-year mortgage and the 20% downpayment expanded by something like 30% during this period. now, with the loans no longer being offered and underwriting standards returning, that relationship is re-establishing itself at its old level.
what is that old level? it depends on a number of factors -- average home size, mortgage rates, average down payment required -- but traditionally the price of a home on a national view has been something like 2.7 times gross household income per goldman sachs' jan hatzius, with some variation. (this can vary by metropolitan area, but happens to be approximately true for chicagoland.)
worse, for so long as banks remain under terrible balance sheet stress, the likelihood is that the relationship of price-to-income will undershoot the mean of the old relationship and find bottom somewhere near the low levels seen in other recent times of bank distress -- something like 2.2 times income.
moreover, this has caught the homebuilders out on the thinnest of branches. with the debts of forthcoming massive developments already incurred, they have had little choice but continue to build in an effort to liquidate assets in an effort to at least pay off as much of their financing as they can. new home supply has, therefore, expanded spectacularly. and with the collapse in demand, there have not been enough buyers to absorb the existing houses being put up for sale by their current owners (many of them recently bought on variable or fraudulent terms and now under duress of impending foreclosure), compounding supply problems. calculated risk has estimated that homes for sale in the united states will rise to more than a year's worth of sales at some point this summer, a post-depression record.
so what should one expect in terms of prices, then? median household income in the united states is currently about $48,500. the median price for a single-family home in the united states in 4q2007 according to the national association of realtors was $206,000. (condos are reported separately and are more expensive at $221k.) this ratio of 4.25 implies that housing on a national scale in the united states is as much as 36% overvalued as compared to a historical ratio of 2.7. compared to a trough ratio to income of 2.2, they could be as much as 48% overvalued.
in chicago, the 4q average price of an SFR was $261,000. median income in the metropolitan statistical area (MSA) is probably about $74,000. that is a ratio of 3.53, or some 24% overvalued from a long-term ratio of 2.7 -- and 38% above a trough ratio of 2.2.
listing times indicate that house prices in chicago haven't fallen nearly far enough to start resolving inventory issues, i.e. bring supply in line with traditional-finance demand to work off the overhang of unsold homes.