Ignore the Noise

unced. While the data was disappointing, I am still optimistic that the economy turned the corner beginning in June. I steadfastly believe that this recovery will fall somewhere between the normal growth rate after major recessions (6%-7% improvement over several quarters) and the current consensus for sluggish growth (2%-3%). Despite last week's setbacks, the magnitude of the past declines and the potential positives are consistent with growth of 3%-4%, though the data indicates that improvements could come a little slower than I had anticipated.

So is this run of bad news the start of a new trend, or just noise? This whole recovery has been characterized by mixed indicators that periodically give everyone a bad scare--and at least a small chance to get back in the market. As I look back over my weekly columns, it seems that about every fourth column, I spend some time explaining away one bad number or another. Just look at some of the headlines I have written:

Growing Pains for the Green Shoots (5/18/2009)
A Bumpy Road to the Recovery (7/13/2009)
Consumers Lay an Egg (8/15/2009)

So while many individuals are suffering, the overall trend in big picture indicators has, on balance, gotten better despite monthly setbacks here or there.



E: Morningstar Estimates from Robert Johnson



All the confusing variability comes for a variety of reasons. First, the economy is made up of four key segments: the consumer, government, fixed investment, and net exports. The economic cycle is generally led by the consumer. This time around, I believe the consumer hit bottom in late 2008, while industrial production, related most closely to investment spending, didn't bottom until June.

Then there is some statistical noise that some market participants confuse with real trend changes. As I wrote about separately last week, the ISM Purchasing Managers Survey slipped from 52.9 to 52.6, while the index generally ranges from 30 to 70. Because the data is based on a sample, it is really hard to conclude that August was very different than July when the contraction from month to month is 0.3; the difference is just too small. Just a few more people being on vacation in August than September could have made the difference.

Some of the same thing happened with the jobs report for September. Keep in mind that we are talking about 260,000 job losses in a workforce of over 150 million. The number is impacted by small changes in seasonal adjustment factors and assumptions about business formation. Given the margin of error that is introduced by only surveying a fraction of all firms, it is hard to get excited about a number that isn't that far off trend. The June jobs report also fell off the wagon so to speak, before recovering in July. Though I don't publish them, last weeks rail numbers were affected by flooding conditions in the Southeast. These types of things flow through the data sets each week, making trend determinations even harder.

One other thing to keep in mind is that once the economy gets moving in a new direction, it tends to stay moving in that direction for some period of time unless shocked by a large outside force. Firms are reluctant to change employment and capital investments each and every month based on a new piece of economic data. In fact, on long lead time capital equipment, companies can't change their minds. So this is why identifying an economic turn is so important. While it is not unusual to have one bad quarter in an economic recovery, it is highly unusual to have a series of bad quarters after a turn. The 1980 recession followed by the 1982 recession was perhaps the biggest exception to this rule.

So what could cause a double-dip this time? Some type of foreign policy action, possibly involving Iran, is probably my biggest concern for the economy. My second big concern would be a major unanticipated bank or other financial institutional failure that once again shakes everyone's confidence. The third big concern is a sudden withdrawal of stimulus by either the Fed or from fiscal policy. So far Fed Chairman Ben Bernanke seems to be well aware of the issue of pulling stimulus too soon (he has written about the effects of monetary and fiscal policy contributing to the second down-leg of the economy in 1937).

The bottom line is, despite the setbacks I explain below, I think the economy is getting better, and without an outside jolt, it is unlikely to fall back into the abyss.

Last Week's Economic Data
The employment data last week was a disappointment as we noted above. We all like nice straight lines, and I was hopeful that we could string together some more months of improvement on the jobs front. Unfortunately, we lost another 263,000 jobs during September, up from 201,000 in August but still far better than the peak of 741,000 in January. I had predicted that the number could have been as good as a minimal loss of 160,000.

So what went wrong? Three sectors contributed to the surprise: construction, retail, and government. Most shocking was that despite all the stimulus money, we lost 53,000 governments jobs. Some were on the federal level, but most job losses were from local governments. Given poor tax receipts, it is possible that the sector could be more negative than I anticipated for a few more months. The retail sector was partially affected by layoffs at auto dealerships that may have been related to the end of the "Cash for Clunkers" program. For some odd reason, employment at gas stations, which is counted in retail sales, was also weak. On the construction side, commercial jobs continued to decline sharply while residential construction barely lost any jobs this month after many months of pretty horrific losses.

Like the June negative surprise on the headline number, individual components showed some meaningful improvement even if the headline number was weak. Eight of the 12 categories I track showed improvement in trend. The information services sector actually managed to effectively show no job losses.

On the downside, the average hourly work week slipped back to 33.0 from 33.1. This number has been in long-term secular decline for more than 50 years. However, there is generally some improvement in this number during the early stages of a recovery. This was perhaps the single most troubling number in the release.

On the positive side, the Case-Shiller Index (announced on Monday) showed home prices actually gained 1.3% for July building on a gain of 0.9% in June. So the market is now up 2.2% from its May low. For the full year, it looks like the index could be almost flat year on year by December (year to date, we are at -4.7% for the 10-city portion of the index). This compares to a loss of 14% that is the base case in the government stress test of banks and well off the 22% worst case scenario. Those scenarios were set up just this spring, so this improvement is really quite notable, despite being largely ignored by Wall Street last week

I commented on the ISM Purchasing Managers survey in a separate piece but generally viewed the data as positive as the index remained well above 50 despite the Street's negative perception.

See More Articles by Robert Johnson

Robert Johnson, CFA, is associate director of economic analysis with Morningstar.