U.S. Markets closed

3Q GDP Too Good to Be True?

Robert Johnson, CFA

The Fed finally ended the tapering speculation on Wednesday. The market cheered an end to the guessing game and a promise of continued low short-term rates--but not necessarily the tapering itself.

The economic news was decidedly mixed, with government furlough gremlins seriously messing with some of the data points. Industrial production was both free of gremlins and healthy at last. Inflation was nonexistent in November, surprising even the Fed. Housing starts were surprisingly and artificially strong even as permits, a leading indicator, faltered. Existing home sales continued to weaken, which might hurt growth more than some suspect. Unemployment claims remained elevated for the second week in a row.

However, the week ended on high note with an upward revision of third-quarter GDP and consumption growth. Even with all these gyrations, I am sticking with my GDP forecast of 2.0% to 2.5% growth, low inflation, and modestly higher interest rates for 2014. That's not so bad. Happy holidays!

The Fed Makes Its Move
Probably the biggest economic news of the week was that the U.S. Federal Reserve would begin to taper its large $85 billion bond purchase program. It will reduce bond purchases to $75 billion, or by $10 billion per month. That is still a considerable sum, but it could be reduced in further $10 billion increments as forthcoming economic data dictates. The Fed also pledged to keep short-term rates near zero until unemployment was well below 6.5%, a lower standard than previously issued reports, which greatly cheered markets overall.

I was glad to see this program begin a slow and measured close. The bond purchase program was a truly extraordinary measure to purchase long-term bonds and mortgages in an attempt to directly control long-term interest rates. Never before has the Fed reacted so boldly and so beyond its sphere of short-term interest rates.

Given extraordinarily tight fiscal measures and a slow growth economy, the program was both helpful and necessary. With the economy at least a little better and an easing of the fiscal tensions, it was probably time to begin winding down such an extraordinary program. Markets had already anticipated the tapering last spring, and interest rates had previously made their move up. This should blunt some of the impact of the actual tapering announcement, but rates still may move up further.

GDP Revised Up Sharply Again, This Time to Annualized Growth of 4.1%
Third-quarter GDP growth was revised upward from an already strong 2.8% on the first estimate and an even better 3.6% growth rate on the second estimate to a final estimate of 4.1% on Friday morning. We haven't seen such large revisions in some time.

Unlike the second revision, this one was of higher quality, with most of the revision centered on consumption in general, and health care and gasoline in particular. The second revision was related almost entirely to inventory building. Inventories still accounted for about 40% of the GDP growth rate. However, consumption in the third quarter is now higher than it was in the second quarter, while consumer income data was little changed in the revision. This partially explains the disconnect between nicely improving consumer incomes and lackluster consumption data that was originally reported.

That said, it's a bit of mystery of why health care would make a sudden move up. I am more than a little suspicious that maybe it's one of those too-good-to-be-true numbers that gets reversed in the fourth quarter. I am not so sure that increased health-care spending is a great help to the economy, either. It sounds more like some kind of accounting shift. Still, I am not complaining--I'll take 4.1% growth any way I can get it.

Falling existing-home sales and slower inventory growth in the fourth quarter is likely to bring the GDP growth rate down to a range of 1.5%-2.0%.

Housing Starts and Permits Data a Mixed Bag
Overall, housing starts looked fantastic when reading just the headline numbers for starts. However, the underlying data and the permits data were quite troubling.

First the good news: Housing starts were up dramatically in November with stunning gains in both the month-to-month and year-over-year data. On a single-month basis starts were up 22.6% month over month and 29.7% year over year. Total starts at 1.09 million units annualized were the best levels of the recovery. Multifamily did better than the more important single-family sector, but not enough to really quibble about.

I do have three worries relative to the November report: Data collection issues due to the furlough probably pushed starts into November that might have normally counted in September and October. November looked great because September and October were relative disasters. Also, because November is usually not a great month for starts anyway, huge seasonal factors multiplied the effects of shifting those starts. My year-over-year analysis will remove some, but not all, of those issues. I usually don't publish just month-to-month numbers, but that is the only way to make the furlough effects clear.

Even if a million units are started in December, full-year starts would come in at about 920,000 units, which would be considerably behind midyear forecasts of more than 1 million starts in 2013. For comparison purposes 781,000 units were sold in 2012.

I suppose the good news about the lower 2013 numbers is that it means a higher rate of growth in 2014, off of this lower base. It also probably adds yet another year to the housing recovery. Recall that there were more than 2.2 million starts at the peak during the housing bubble, and 1.5 million is considered a normal housing year based on population growth and natural deterioration.

The Averaged Starts Data and Permits Paint a Bleaker Picture
Averaging the starts over three months and comparing that figure with a year ago probably provides a better view of the data.

Starts data had been decelerating since March, tumbling from a high of 34% growth to just under 8% in October. That's relatively dramatic, and certainly means that housing, at least the new home portion, will be a smaller contributor to GDP than it has been. The good news is that November may have arrested the downward trend (although even that might be more of a data collection issue and not a true halt to the decline).

Also troubling is the permits data, which almost always leads the starts data. Permit growth has been declining for even longer than starts and continues to slow down, as shown above. The single-point year-over-year change for November was even worse at about 7%. Additionally, November permits were about 3% below October's level and well below the number of starts.

Stepping back and taking a broader view, starts are likely to increase by about 20% in 2013 following 28% growth in 2012 and 4% in 2011. For 2014, something in the range of 1.05 million to 1.1 million units seems to be the most likely outcome, which equates to 14%-19% growth. That's higher than the current permit data suggests at the moment.

Builders Remain Optimistic, With Some Improvement in November
A lot of the government data is plagued with collection- and furlough-related issues, complicated by seasonal adjustment factors. Still, through thick and thin, builder sentiment has remained on a fairly even keel. In truth, sentiment improved in November tying a record made this summer and is now at its highest point since 2005. Readings on traffic, current sales, and future sales were all improved from the prior reading.

Though the builders have been a little too optimistic in general, I think it's a decent indicator that the housing market is holding together pretty well all things considered (higher mortgage rates, land shortages, appraisal issues, and so on). New home sales were also hurt midyear by competition from existing homes that could be more quickly purchased in a rapidly rising rate environment.

Existing-Home Sales Finally Take Their Expected Tumble
Pending home sales data has been indicating a slowing in the housing market for some time, but didn't immediately show up in the closed sales data (existing-home sales). The push to higher rates this summer caused a quick spurt in sales as fence-sitters were finally moved to action as rates moved higher. Those same quickly rising rates made new homes less competitive, giving the temporary edge to existing-home sales.

That process is now working in reverse. Near desperation on the part of buyers to beat the rate increases also substantially reduced the time between signing the deal (a pending sale) and a closed sale (an existing-home sale). That effectively drained the backlog of homes in process and artificially boosted existing-home sales. Now it is time to pay the piper. Sales decreased about 10% from July's 5.39 million level to 4.9 million units in November. The November figure is even below the year-ago number and is the first time existing-home sales have dipped below 5 million units since May. Assuming December sales look like November's data, the fall in existing-home sales will likely take 0.2% off of fourth-quarter GDP instead of adding about 0.1% in the third quarter.

In units, the averaged total has been falling since September and the average price (a reflection of the mix of homes sold and not changes in the price of comparable units) has been declining since August. The poor November data suggests that the averaged data could continue to decline for a few more months. With higher rates, less affordability, and tight lending constraints, the National Association of Realtors is expecting existing-home sales of 5.1 million units in 2014, basically unchanged from 2013 levels.

The growth rate data isn't much more reassuring. Furthermore, increasing prices and the previous favorable mix shift to higher-end homes has now turned negative and is no longer providing a tailwind to the transaction value data.

Consumer Prices Flat Again
There was no change in overall prices between October and November and only a 1.2% increase November over November. The averaged data continues to show a very benign inflation picture and an even lower year-over-year inflation rate, which should be helping consumers over the holiday season.

The general thought is that inflation will again increase slightly in 2014. The tailwind from the impact of key prescription drug patent expirations and hospital expense reductions due to the sequester will slowly erode in 2014. However, we should hopefully avoid the huge spring gas price hike that we have seen in each of the last three years. Food prices also should be under better control with a bumper U.S. crop of corn and soybeans. Overall, I suspect inflation will grow 1.5%-1.8% in 2014 from this year's estimated 1.2% level.

The category-by-category analysis shows that the low inflation rate this month was driven primarily by energy-related categories, although apparel prices and new vehicles also registered outright declines in price, and food prices other than dairy were little changed.

Energy was not a one-way street this month, either, as fuel oil and electricity registered increases. Restaurant prices were also up big, which is a bit of a surprise because sales were up too, according to the earlier retail sales report. Shelter has also bumped up a bit as higher rents are beginning to work their way into the database. My guess is that this could be an ongoing problem until more apartment buildings come online in 2014. Air fares continue to move up, and that is even before the American/U.S. Air merger, which is likely to raise rates even more. The month-to-month (not even annualized) airline figure was up 3.6%.

A 0.4% increase in auto insurance rates didn't help the transportation services category, either. With the exception of dairy products, most grocery categories were down, with even the pesky meats category finally registering a decline.

Industrial Production Breaks Loose at Last, Gaining 1.1% October to November
Strong purchasing managers' data has suggested a strong manufacturing segment for more than six months with no real, definitive breakout. This month we got that breakout. The manufacturing-only sector, which excludes a big bounce in the mining sector and a weather-related pop in utility sales, still managed a 0.6% month-to-month growth rate. Even the year-over-year data is on the mend but still shy of the ebullient conditions that PMI data suggests.

By category, the pickup was widespread strength. The 3.4% growth in auto production made up almost a quarter of the production gain in the quarter. Wood products and furniture categories were also strong and haven't yet reacted to slowing conditions in the housing market (although we hear a lot of wood products are going to China, cushioning some of the data). Apparel, perennial loser printing, machinery, computers, and aerospace were among the few small losers.

Cut another interesting way: Consumer-related items trounced business equipment, with consumer goods production jumping 1.5% while business equipment (computers, transit, and other) fell 5%. Consumers in general seem to be feeling more confident than businesses. Defense-related products did even worse, falling 0.8%. With some sequestration pressures now easing, this category could lift a little in the months ahead.