Economic Notes – October 28, 2013
Now that the government is back open for business, we again have our flurry of economic items.
(0/-) Durable goods orders for September came in mixed, with the headline number gaining +3.7% versus a forecasted +2.3%. Removing transports from the orders group was a negative, as commercial/non-defense airplane orders played a significant role in total orders last month, gaining +58%—resulting in a one-month -0.1% decline, which underwhelmed relative to expected growth of a positive half-percent. Core orders and shipments were also down, which is significant due to the latter’s use as an input into quarterly GDP. Manufacturing inventories rose almost a percent, which happened to be the largest increase in two years and somewhat of a negative as we look ahead (manufactured goods being stockpiled is less productive than those actually ordered and shipped). This is a moderately-watched economic report, and in keeping with other economic data pointing to some recent softness.
(+) The Richmond Fed manufacturing index came in at +1 for October, which slightly surpassed the expected 0 reading. Shipments and new orders declined, but employment improved by +10 points—quite significant. Future expectations declined somewhat, but remain optimistic overall as opposed to pessimistic.
(-) The preliminary Markit PMI survey for October was weaker than expected, at 51.1 versus a forecasted 52.5 (albeit still positive, in expansionary territory). Output fell by five points and new orders dropped slightly, but employment ticked up a point. The mixed result here is largely in line with an up-and-down series of regional Fed and surveys during the last several weeks.
(0) Import prices rose +0.2% in September, which was largely in line with forecast. The bulk of this was due to a +1% increase in petroleum prices, which is a fickle short-term component, as other prices were flat. However, over the full year, import prices are actually down a percent, which is consistent with widespread tempered inflationary pressures.
(+) The U.S. trade deficit for August narrowed somewhat to -$38.8 billion, better than the expected -$39.4 bil. Actual import/export activity was virtually unchanged for the month; the adjustment was due to a downward revision in July’s deficit data. So, the report was largely in line with trend.
(0) Existing home sales fell -1.9% in September, which exceeded expectations of a -3.3% decline; but August’s numbers were also revised downward from near +2% to zero. However, on a year-over-year basis, existing sales remain up +11%. Single-family units were down a percent and a half, while condos/co-op sales fell nearly -5%. These figures, which measure actual completed closings, obviously track the trend of pending home sales from a few months prior at the point of completion. There’s a good chance of there being some carryover from the early summer mortgage rate spike, but overall growth levels from the longer-term basis remain in recovery mode.
(-) The FHFA home price index rose +0.3% for August, which fell a bit short of expectations of a +0.8% gain, but the year-over-year trend remained positive at +8.5%. Of the nine U.S. regions, six experienced a slowdown in appreciation relative to the prior month, while the Mountain and Pacific regions led with roughly +1% gains each. The South Atlantic/Florida region dropped -0.5%, which was the worst performing regional group.
(+) Construction spending rose +0.6% for August, which slightly surpassed the expected +0.6%; at the same time, July’s construction numbers were revised up by nearly a percent. For the August period, residential construction grew by +1.3% while non-residential grew just a few tenths of a percent. Year-over-year gains were +19% for residential, +4.3% for private non-residential and a drop of almost -2% for public/government spending.
(-) The final October University of Mich. Sentiment survey came in at 73.2, which disappointed somewhat relative to the expected 75.0 reading, September’s 77.5 result and this month’s preliminary figure of 75.2. The differences originated from consumer assessments of both current and future conditions, which worsened. Interestingly, but not surprisingly, survey administrators mentioned that the number of consumers mentioning a distaste for the federal government was at the highest level in the 50-year history of the survey. However, now that things are slowly returning to normal, we expect this to improve a bit going forward—as these responses are often news-driven and reactionary, and tend to fade with time. Inflation expectations are down a few ticks from where they were in September, but remain in the upper 2% to near 3% historical range (these figures have changed so little lately, they’re almost an afterthought to bring up). In fact, gasoline prices drive much of those inflation results month to month.
Now, onto the closely-watched jobs data… We say ‘closely-watched’ due to the impact these numbers are likely to have on Fed decision-making and taper plans in coming months.
(0) Initial jobless claims for the Oct. 19 week fell from a revised 362k to 350k—still above consensus forecasts of 340k, so somewhat of a disappointment. California claims are still problematic—in that claims processors are continuing to work off a backlog caused by their often-discussed computer system change (last week showed a 12k spike in state claims, which is admittedly less than the 33k the week before). On top of that, the DOL mentioned that they weren’t able to get a good handle on the claims of non-federal workers impacted by the shutdown. These little things add up, and point to one more report that may have to be thrown out for the time being until the data gets a bit ‘cleaner.’ Speaking of shutdown impacts, it appears federal workers filed 70k and 44k claims for the weeks ended Oct. 5 and Oct. 12, respectively, which obviously results in a significant percentage of overall filings. Continuing claims for the Oct. 12 ending week came in at 2,874k, which was a bit lower than the 2,882k expected and appeared less out of whack than the initial claims numbers have been during the past month.
(+/0) The August government JOLTS (Job Openings and Labor Turnover Survey) report was decent, showing 3,883k openings in August, compared to 3,765k forecast. The hiring rate rose a tenth to 3.3% (from a July revision), but remains below the typical business cycle rate. Layoff and discharge rates were unchanged.
(-) The long-awaited and shutdown-delayed government unemployment situation report was finally released for September, and was a general disappointment. Payrolls came in at +148k, which was below the forecasted +180k. Some of the sting was tempered by revisions for prior months (August up, July down, net gain of +9k) but this wasn’t enough to really move the needle. The September details were focused on growth in state/local government jobs (+22k), construction (+20k, perhaps weather-aided) and health/education services (+14k) while leisure/hospitality dropped (by -13k).
The unemployment rate declined to 7.2% (from the simple fact that the published number is rounded to the nearest tenth—to be more precise, the actual drop was only four basis points from 7.277% to 7.235%—which minimizes the excitement). The labor force participation rate, which is being especially monitored lately as an important factor in this calculation, held steady.
Average hourly earnings rose +0.1% for September, which fell below the median forecast of +0.2% and brought the year-over-year rate of change to 2.1%—well in line with current inflation figures. The average hourly workweek was unchanged at 34.5. For the third quarter, the aggregate weekly hours index (which measures the number of workers and hours per worker) only rose at a one percent annualized rate.
Finally, now that the government is working again, how much damage did this shutdown really do? Economically, not much on a net basis (perhaps a trim of 0.1-0.2%off GDP according to some estimates) considering back pay was reinstated. So, we might see a drop in some of the October economic numbers due to lower activity during this time, but the figures for later in the quarter may well return to more normal levels reflecting the bounceback. The real damage seemed to appear from a sentiment standpoint; however, this, too, seems to have been mitigated by a general public frustration with government and a numbing of sensitivities to these political shenanigans.
For this report and those going forward, labor force participation and other composition components will be closely watched—particularly due to Janet Yellen’s high level of academic and practical attention on that segment of the economy.
Essentially, the unemployment rate is the number of sampled unemployed persons divided by the ‘participating’ labor pool of certain-age workers in the denominator. The broader U-6 report adds in the underemployed and frustrated folks in the numerator, but keeps the same pool in the denominator (the U-6 figure has generally been about two times the size of the standard unemployment rate—last month a bit smaller than 2x, at 13.6%—the lowest reading of the recovery so far). The percentage of unemployed looking for work longer than six months has also fallen, to 36.9%, which is the second-lowest reading of the current recovery period beginning June 2009.
These long-term unemployed continue to be a problem, as the data indicates. While the percentage rate has fallen, it remains a significant population of 4.3 million Americans compared to the 20-year average of 1.2 million (which traditionally represented 15% of the labor pool—less than half what it is now). The problem is really being felt and discussed at the edges. For example, there is a debate among economists right now about the size of the population caught in the frictional group—not unemployed because of normal business cycle factors (which are getting better, albeit slowly), but increasingly due to structural issues involving the inability to sell a home to relocate for better work, deterioration of skills (or lack of skills to begin with), etc. There is no easy way to measure the margins here.
What does this all mean? Likely, if this lackluster and meandering trend of employment data continues, the Fed may take it as a reason/excuse to push back tapering of QE, especially as it’s been Yellen’s bias that cyclical factors are the predominant factor. We’ll have another note to discuss this after the intermediate Fed meeting this coming week, but that is the general theme. But is more/extended QE the elixir that solves the employment problem?
It was a rare week with both stocks and bonds moving higher. Stocks continued upward to record highs with stronger-than-expected earnings numbers, while bond yields ticked lower on weaker jobs data.
In the S&P, industrials and consumer discretionary stocks led on the week, while energy and financials lagged. So far, in the 3rd quarter earnings season, 76% of reporting firms have beaten earnings estimates, while 54% have surpassed on the revenue side. Earnings growth for the index stands at an estimated 12% for the quarter, while revenue is growing at 5%. The general consensus estimate for 2013 earnings is $107, which, with the S&P at 1760 this morning, puts the trailing P/E at 16.4. Estimates for 2014 put earnings at $116, which put the forward P/E at 15.2. Both remain near long-term historical averages and large-caps are near ‘fair value’ by a variety of metrics.
We haven’t discussed shorter-term sentiment lately, but the AAII survey that measures these levels each weekly has turned much more bullish (49% of respondents were bulls, relative to 33% neutral and only 18% bears). These readings are, of course, reactionary. The better the stock market performs, the more bullish retail investors get, and vice versa.
Foreign developed market stocks lagged U.S. equities, with Europe gaining 0.75% tempered by Japan’s -2% week (most of which occurred on Friday in response to a 7.3 earthquake off the coast). China was the worst-performing emerging market, down nearly -4%, as decent manufacturing survey data was overwhelmed by more money market credit problems—as the central bank chose not to inject cash to support the system. As a result, lending rates shot up from 3% to 4% in just a few days, on speculation the government may take additional steps to rein in real estate lending that has contributed to extremely buoyant housing price increases and inflation.
Bonds fared decently with rates falling a few basis points. Long bonds gained about a percentage point, as did municipals and developed market international debt. Floating rate and short-duration bonds were flat to just slightly positive. The good news about the bond rally for many bond investors is the improvement in year-to-date return from lows of -4% to just -1% today. Much of the recent recovery rally has been due to a re-slowing of economic data, government shutdown/debt ceiling negativity and postponement of tapering—which benefitted risk-off assets. Should the data again pick up dramatically and/or tapering commences, we think yields might be poised to shoot higher once again (short duration portfolios should perform better than conventional index-like bond portfolios in that environment).
Commodity returns were led by precious metals—both gold and silver—the former rising 3% due to the beginning of festival season in India and probably elongated stimulus from the Fed. By contrast, crude oil fell 3% as a result of stronger inventories globally. U.S. production has risen to almost 7.9 million barrels/day, which is the highest rate since 1989, in no small part to the fracking/shale oil boom in recent years. While a welcome development for employment in the Central Plains states and prices domestically, it is important to note that shale oil is relatively expensive oil, from the standpoint of it being more difficult and energy-intensive to process. If world oil prices fall from their current near-$100/barrel levels towards $50 or even $75, the economic viability of many of these projects may look a little less rosy. Just something to keep in mind as we watch oil prices and this U.S. production boom.
Have a great week!