|
November 4, 2005
Long-term rates have risen in a straight line since Labor Day, now more than one-half percent. The 10-year T-note, 4.67% today, has taken out several technical stops without pause, and 30-year low-fee mortgages are at the doorstep of 6.50%.
Bonds were hurt by some fairly strong economic data, and by inflation discussions shifting to a how-bad-is-it competition, but the real damage to rates is coming from the Fed. The light has dawned that wherever neutral was before oil hit sixty bucks, it is higher now. Further, the location of neutral is passing into historical curiosity as the Fed will likely have to tighten past neutral, going as far as necessary to intercept inflation pressure... going until the economy visibly slows.
The Fed’s fingerprints are on the crucial Treasury 2-year to 10-year spread: it’s been steady at a very narrow .20% for four months. Long-term rates are rising because the Fed is bulldozing the whole rate structure upward from underneath; if the bond market thought the Fed wasn’t tough enough, 10s would be pulling away from 2s. However, this .20% 2s-to-10s spread says nothing about high the Fed will ultimately go. The stop signal would be 2s converging on 10s, or rising above.
The first data from October show an economy if anything accelerating from the pre-Katrina pace. The twin reports from the purchasing managers association show manufacturing running hot, above the 59 level, and the service sector in a strong October rebound, to 60 from 53 in September. Same-store retail sales (the measure removes distortion from new openings) soared 4.4% in October.
There is no sign whatever that consumers have been hurt by high energy prices, rising interest rates, or by anything else.
In modest good news for inflation, labor productivity in the third quarter increased at double the forecast pace. However, news this morning that price pressure is moving into wages overwhelmed everything else, including weak growth in the big-company payroll survey.
The fingerprints may be the Fed’s, but mortgages are doing the heavy lifting. The big selling in the bond market this week was mortgage-related: the proud owners of $5.5 trillion-worth of fixed-rate mortgages have to go short bonds to hedge their positions. As one result, this was one of the rare weeks in which mortgage rates rose more than long Treasurys’.
Fixed mortgage rates are now about even with the highs in each year since 2001. How high they will have to go to cut into home purchases... we won’t know until we are there. Classically, a 2.50% rise is necessary to clamp the housing market: in recent examples, from 9.00% to 11.50% in 1989, and 7.00% to 9.50% in 1994. In the Fed’s last cycle, ’99-’00, mortgages went from 7.00% to 8.50%, but the technology bubble-burst collapsed the economy before housing broke.
So far, mortgages are only 1.25% percent above the cycle low, and on theory might have to rise to 7.75%. I doubt it. The anomaly in this cycle (among many) is the role of the adjustable-rate mortgage. In recent real estate expansion phases, ARMs were not important housing-market propellants: short-to-long rate spreads were narrow, and ARMs were no great advantage over fixed loans.
Contrariwise, in the four-year super-cycle now concluded, short-to-long spreads were the widest in a half-century, and ARMs for home buyers bordered on free money. No longer. Rates on five-year hybrid ARMs have almost doubled in eighteen months. Construction money has doubled. Piggy-back seconds, essential to the low-down payment market, have gone from five-something to eight-something.
Especially in the hottest housing markets, these adjustable and innovative products have accounted for more than half of all purchases, and suddenly have no utility at all.
|
