The 10-year T-note hovered near 4.35 percent this week, and mortgages have held their improvement to 6.125 percent, the best level since September.
The predominant bond-market bet is still placed on a slowing economy, but until a slowdown actually appears in the data it’s near impossible for long-term rates to do any better than they have. The Fed will on Feb. 1 hike another .25 percent to 4.5 percent, leaving the 10-year T-note badly exposed unless and until slowdown data appear.
Long-term rates moved as low as they have over the holidays because the Fed has wigwagged a near end to its rate hikes, and because of the first, tentative signs of a slowing housing market. Just before Christmas, the November home sales data arrived weaker than a weak forecast: existing-home sales fell 1.7 percent (versus the negative 1.3 percent expected), and new homes were off 11.3 percent (versus minus 8 percent). Inventories of unsold homes rose to a 19-year high, but the time span overstates the actual number. New purchase-loan application numbers tanked in the second half of December, but thin seasonal markets make the indicator questionable.
Data on the rest of the economy could be interpreted to show a slowdown underway, but it’s an eyes-of-the-beholder deal. Today’s employment data for December showed only half the forecast job creation: 108K instead of 213K, but November’s 215K was revised to 305K. Weakness developing, or a wobble along a healthy-growth average? The purchasing-manager indices showed a four-point slip in manufacturing to a still-positive 54.2, but the service sector held a strong 59.
The first flash reports on holiday retail sales are on the weak side, and a pattern you might expect given the ramp-up in gasoline and heating expenses for consumers: luxury retailers had a good season, while the lower-end mass merchants did not. Scrooge suffers less from the price of coal than Cratchit.
The bond-market’s slowdown bet is on display in market prices: without that belief there is no way that long-term yields would be so close to short-term ones. The market speaks for itself.
On the other hand, prognosticators publish. Morgan Stanley’s look into the new year insists that long-term rates will rise, perhaps a lot. Morgan confesses that it had the same forecast in each of the last three years and was mistaken, but the stopped clock has no shame. Persistence is a virtue, but so is learning.
The Wall Street Journal issued its semi-annual compendium by 76 economists, and I guess that most of them buy their booze at the Morgan Stanley saloon. Twenty-six think the 10-year T-note will be above 5 percent by June, and only four think it will yield the same or less than it does today. In the same group last January, only one guy had it right that the 10-year yield would fall below 4 percent.
The group did better at mid-year last year, when 15 of 74 guessed the actual or below for the last six months of ’05. Still, 80 percent wrong on the high side is not a proud credential. In the profession, and among a lot of others in step with Morgan, there is deep disbelief that inflation is a problem of the past, and can remain a non-problem for a long time. It should be obvious to all by now that the deflationary forces flowing from Asia and technology overwhelmed $60 oil; if they could do that, those forces are strong.
The best voice speaking for the bond market is Bill Gross’ newest piece at www.pimco.com, posted Thursday. Well nigh incomprehensible to civilians, it says that long-term rates have now topped out. Further: not much recession risk or yield decline ahead unless the Fed plods blindly upward, fighting the last war; or if the Fed is slow to react to the economic slowdown ahead, slow to cut its rate later this year.
Gross is a bond salesman, and therefore his forecast favorable to bond buyers is suspect. However, trustworthy or not, it’s the best I’ve seen.
Lou Barnes is a mortgage broker and nationally syndicated columnist based in Boulder, Colo. He can be reached at firstname.lastname@example.org.
What’s your opinion? Send your Letter to the Editor to email@example.com.