ADP versus BLS
This week is jobs week in the United States. The ADP Employment Survey is published Wednesday, followed by the official U.S. Bureau of Labor Statistics payroll numbers Friday.
The first, the ADP report, often spooks or cheers the markets when it signals a potential surprise lies ahead in the official numbers. Yet, increasingly, the Automatic Data Processing Inc. report, produced by one of the largest private-sector payroll companies in the U.S., is a problematic predictor of the BLS figure.
Jennifer Lee of BMO Nesbitt Burns Inc. says the ADP report has missed the monthly change in the official number by an average of 92,000 jobs over the last 12 months. While that doesn't sound like much in a work force of more than 100 million, the month-to-month change in payrolls is typically in the tens, or hundreds, of thousands.
For instance, ADP's estimate for September was a decrease of 39,000 jobs, while the Bureau of Labor Statistics reported a 64,000 gain - a difference of 103,000. It was the fourth time in the last 12 months ADP's report differed from the official number by 100,000 or more, including a 200,000 difference in November, 2009.
ADP says it's trying to help the public get a jump on the official numbers, saying "information that helps analysts anticipate monthly changes in employment is valuable."
Alas, says Ms. Lee, it seems to be increasingly less valuable. "One should take the ADP results with lots of salt every month, as it has provided a weaker tone than what the official results told us, at least during 10 of the past 11 months."
GDP
Not that we want to undermine your confidence in economic statistics, but we should point out that last Friday's numbers for U.S. gross domestic product are not the final word, as suggested by the "advance" label placed on it by the Bureau of Economic Analysis.
The number goes through at least two revisions from its initial release; the end figure can be substantially different, particularly in times of upheaval. The preliminary third-quarter GDP growth number released last week will be revised in November, then revised again in December for its "final" figure, says Mike Englund of Action Economics.
Ah, but the final is not final, as it is subject to an annual revision next July and each July afterward for four more years, Mr. Englund says. And then? A potential revision every five years after that, indefinitely.
You won't care about 2010 third-quarter GDP in 2025, in all likelihood, but you'll care if market reaction to the initial number turns out to be unwarranted later on.
So far in 2010, the preliminary numbers have been bumped up by 0.5 percentage points and down by 0.7 in the first and second quarters, respectively. Last year saw a downward revision of nearly two full percentage points for the third quarter. And the terrible fourth quarter of 2008? Originally thought to be a drop of 3.8 per cent, it was eventually revised to negative 6.8 per cent.
Fund flows
The recent comeback of U.S. stocks has been of little use to the shell-shocked retail investor, who has been participating in an epic run of doubt about U.S. equities.
The evidence is in mutual fund flow data from the Investment Company Institute, a trade group for the fund industry. Funds focusing on U.S. equities have seen negative flows for 25 consecutive weeks, through Oct. 20. The last positive week for U.S. equity funds was in late April, and the money pulled out, on balance, has totalled nearly $83-billion (U.S.).
The big winner? You know the answer: Bond funds, which haven't had a week of negative flows since December, 2008. In those 96 weeks, they've taken in more than $661-billion.
This could be yet another example of the long-standing tradition of investors chasing performance and timing it all wrong. Or it could have greater meaning, says Charles Hugh Smith, a columnist at AOL's Daily Finance site. He says American investors have finally lost faith in the stock market.
"Institutional money managers and individual investors alike have been voting with their feet and abandoning the market for both equity funds and mortgage-backed securities."
Given the market's September and October performance - and the sharply lower outflow numbers of the last two weeks - we'll see whether the remainder of this year's numbers support his theory.
Special to The Globe and Mail