(By Robert Johnson, CFA) This week, Hurricane Sandy's destructive powers swept their way through a series of reports, including new home sales, personal income, November same-store sales, and October consumption figures, rendering them nearly useless. Further complicating matters, personal income data was revised sharply downward all the way back to April, along with consumption data for August-September. Then there is the violent up-and-down swings in auto production (at least, according to government data) that took 0.5% of third-quarter results. Did I mention that gasoline prices have been acting like a giant yo-yo, too?
Unfortunately, the data is likely to remain muddled for at least another couple of months. The next piece of storm-addled data is likely to be next week's employment report for November, which could show some significant effects from Sandy. In fact, one can't rule out the possibility that we might potentially lose jobs in November, and October numbers could be subject to downward revisions.
Don't despair. The recent data suggest that going into the Sandy situation, the economy was performing better than most economists believed, as third-quarter GDP growth rates were revised sharply upward. Also coming out of the storm, some of the weekly data, including shopping center sales and initial unemployment, show some pretty sharp improvement. Next week, industry experts are expecting a hard rebound in auto sales, which should also help lift some of the gloom and give us some concrete evidence that the storm-related declines are not permanent. Auto sales were the first storm-affected piece of data to be hit hard, and I suspect that they will be the first to rebound. Mercifully, gasoline prices are also continuing to decline, putting more money in consumers' pockets during the crucial holiday season.
In this week's economic news the GDP report was indeed revised sharply higher to a growth rate of a very respectable 2.7%, up from the initial read of 2.0%. Unfortunately, changes in inventory levels represented a bigger portion of the revision than I had hoped.
The personal income and consumption report was disappointing, but included large, negative manual adjustment for storm effects that took almost 0.2% off of the income report.
Three of the four real estate related reports were quite a bit better than expected, while new home sales were disappointing, but clearly affected by Sandy. Thursday's same-store sales report was substandard on the surface, falling well short of expectations. However, several chains noted that the storm had a bigger-than-anticipated influence on the first few weeks of the quarter, while the last two weeks of the quarter were surprisingly strong.
On the manufacturing front, durable goods orders were positive and better than expected. Meanwhile, the Chicago purchasing managers' report moved back over 50 (more firms saw growth rather than contraction) after spending two months below 50. The employment part of the report was particularly strong while the new order indicator was disappointing.
The Fiscal Cliff Dominates Investors' Thinking
Meanwhile, the market continued its obsession with fiscal cliff issues, with the markets seemingly expecting a better outcome than just two weeks ago. While I still think this will get resolved peacefully and by year-end, I was disappointed to see that the most recent Democratic position seemed to step up demands on both the spending and taxing side of the equation. More stimulus spending and an extension of 2% payroll tax and the 99 weeks of unemployment (both thought to be sure to disappear) crept into the most recent proposal. However, both sides seem to be willing to be a little more flexible on the exact top rate and the level at which exemptions might be capped.
GDP Revised Higher
Overall, the GDP report for the third quarter was revised up from 2.0% to 2.7%. While exports were a key to the revision, inventories were an even bigger help, which is generally not a good thing.
The consumer spending category saw softer gains than originally announced, but were still in an acceptable range while business investments were moved from a small gain to a small loss. Government spending, which was a much bigger contributor than usual, was basically unchanged in the revisions.
The table below shows the overall contribution by category, which represents a combination of the relative size of the category and the magnitude of the gain.
The contribution data clearly shows that the consumer remains in the driver's seat. Despite some recent improvement, the housing industry has yet to be a very big contributor to the overall GDP numbers. While better this quarter, government has been an overall detractor from the recovery, an event that is almost unprecedented in the postwar era.
Looking ahead, the consensus is that GDP growth will dip to 1.5% in the fourth quarter (and probably going lower after Friday's consumption report). I have now seen a few of the more updated forecasts, and they all look closer to 1% than the 2% level that I am using. I think a lot of people missed how much new autos hurt the third quarter (they subtracted 0.5%) and how likely it is that they will be at least a small contributor to the fourth quarter. That would just about offset the 0.7% positive contribution from government in the third quarter that is unlikely to recur in the fourth quarter. Farm inventories, another huge detractor in the third quarter, is unlikely to recur. However, this week's personal income and consumption report got the fourth quarter off to a really bad start on the all-important consumer sector.
Sandy Hits Consumer Income and Expenditure Report Hard
Most economists had expected some pretty dismal numbers for the personal income and expenditures report, but the news was even more disappointing than expected. According to the report there was no income growth in October, and adjusted for inflation, disposable income was down 0.1%. However, I caution that most of the negative effects were likely storm related. Because of a lack of some data and not believing other portions, the government felt compelled to make an additional $18 billion manual adjustment to the personal income report. That equates to about a 0.2% effect on income, which is approximately the amount by which the report missed my estimates. Perhaps even more disconcerting was personal income, primarily the wages category, was adjusted down a stunning $30 billion or 0.3% for the June quarter. That's not a typo. There was an unusually large data adjustment for events that happened seven months ago. The new set of data suggests that consumers dug deeper in their pockets to finance their spending than previously anticipated.
The consumption part of the report was even worse with spending down 0.2% (actually 0.3% when adjusted for inflation). Again, falling auto sales and the retail sales report telegraphed that there would likely be an issue, but this report was even worse. This was not the best way to start the fourth quarter.
Bimodal November Retail Data Won't Help November Consumption Numbers Much
The hurricane seems to have managed to hurt two months of data pretty effectively. The consumption report mentioned above covered October, which only reflected part of the storm effects (the storm hit on Oct. 29, and the storm-affected week was Oct. 28-Nov. 3).
This month's same-store sales report suggests an even worse effect for November. Same-store sales, ex-drugstores, were up a miserly 1.7% versus expectations of 3%-4%. Most had believed that a strong Black Friday would offset the earlier-in-the-month effects of the storms. No such luck. However, some retailers are whining that cyber Monday sales will count in December this year versus November last year. And a higher portion of sales this year were on layaway, which aren't recognized until fully paid for.
In reality, everyone (including me) was just too complacent about the effects of the storm, at least in the short run. Previous storms may have been more damaging to property, but this is one of the few times that a storm has made a direct hit on such a highly populated, high-income, and high-spending part of the country. The good news is that the data will bounce back at some point, though the moves in both directions will be bigger than previously expected.
Short-Term Data Suggests That Solid Improvement is on the Way
Weekly same-store sales as reported by the International Council of Shopping Centers showed a very sharp improvement, as shown below. Still, an earlier start to Black Friday might have provided some of that boost.
The seasonally adjusted index was at its highest level since July. In various management commentaries it appears that the Black Friday results generally exceed expectations. The following quote from the Reuters (TRI) website sized up the situation well:
"We would probably have seen just blockbuster numbers had it not been for Sandy and the nor'easter, the back-to-back storms in the Northeast," said Alison Paul, U.S. retail and distribution leader and vice chairman for advisory and consulting firm Deloitte.
The two storms and resulting massive power outages and damage "really wiped out a week or two of any retail activity in the most populated part of the country," she said.
Initial Unemployment Claims Continue to Improve
After a big (and delayed) spike from the storm, initial unemployment claims have settled back to close to normal. Unfortunately, claims were still quite elevated for the week ended Nov. 17, which was the week that employment report was compiled. This likely means bad news for the employment report due next week. On a side note, the spike in claims in California last week was completely reversed this week, which now means almost all the elevated level of claims is in storm-related states.
Real Estate Data Remains Robust
There was a lot of real estate data this week and most of it remained in decisively positive territory, as shown in the chart below.
The pricing data was particularly strong and now suggests that home prices may be up as much as 5% for the year, which should certainly put consumers in a better mood. Pending sales, which are a decent projector of existing home sales, showed some nice month-to-month growth, and year-over-year data remained steady, which should mean more good news to come.
The big jump came despite a storm-related slump in the Northeast and an inventory shortage in the West. The only mildly disappointing news this week in real estate was new home sales, which showed a minimal decline, almost all of it related to sales in the storm-affected South and Northeast regions. A little more bothersome was the downward revision of the prior month's data. New home sales, which primarily reflect single-family tract sales (where the builder owns the land and the house), have not done as well as housing starts, which include multifamily units and single-family homes built by an owner or a general contractor on land owned by the buyer and not the builder.
Storms Will Make Most of Next Week's Data Pretty Worthless
Next week, ISM Purchasing Manager Data, construction reports, auto sales, and the employment report are all due. So far, most of the regional purchasing manager reports that came early in the month (and in the Northeast) all showed a manufacturing economy under pressure, especially from Sandy. The Chicago regional report (which came later in the month and was not in much of the storm's path) moved above 50 after two months of decline. Economists are playing it safe for now, forecasting the national index will stay flat between October and November at 51.7. I wouldn't lose too much sleep over this one even if it comes in a little light. However, an improving auto industry should help the report, at least a little.
Speaking of autos, I suspect that sales are likely to improve to 14.9 million units, which would put the data back to where they were before the storm. Given that some October sales were shifted into November from October and that the seasonal adjustment factor for November is large, an upside surprise seems a real possibility to me. If there is an underlying softness in the economy (as some of this week's reports suggest), auto sales would show little or no improvement from the 14.2 million units sold in October.
Employment Report to See Storm-Related Pressures
I suspect the unemployment report will be near indecipherable due to the storm. The survey week was after the storm hit but before everything in the Northeast was back to normal. Even before the storm, fiscal cliff worries probably held back at least some hiring. And then there is the Hostess strike issue, which involves 18,000 jobs. My best guess is that the liquidation petition probably hit after the data collection period (but it will show up in December's report). But I will be checking the formal report details just in case.
The consensus is for job growth of about 80,000 people, which strikes me as a little high given the elevated level of unemployment claims and a generally weak economic backdrop. Still, the seasonal factors are working in the right direction, and the manufacturing sector seems to be acting a little better. In addition, I am still waiting for home construction workers to show up in the employment report sometime soon. So far, massive gains in housing starts have not translated to additional workers.