Increasingly jittery investors need to be on the lookout for false signals about the economy, whether from one-month anomalies or readings that may later be revised.
Markets worldwide are particularly nervous as Europe's debt crisis threatens to infect the global financial system and the U.S. economy sputters. Money managers – and their clients – are hunting for any indications of whether their investments are getting safer or more risky, and what to buy and sell in response.
Fluctuating data aren't helping.
Investors got whacked this summer by conflicting signals. U.S. stocks climbed on July 7 after a report from ADP Employer Services showed that employers hired more than twice as many people in June as economists had expected. The market dropped the next day, when statistics from the Department of Labour showed hiring fell short of even the most pessimistic forecast.
"At any turning point, when there's unusual uncertainty around the outlook – which there is now and there has been really for three or four years – you're going to see more significant market volatility around significant numbers," said Benjamin Reitzes, senior economist at BMO Nesbitt Burns. "Markets will take any information they can get, and while it is revised, it does generally give a decent picture of the direction that the economy is headed in. But any one month can give you a false signal."
The U.S. Labour Department's employment report for September, published Friday, showed an increase of 103,000 jobs outside the farm sector, compared with the 60,000 expected on average by economists.
But the department also revised figures for the previous two months. July was initially estimated to have seen 117,000 new jobs; that was later cut to 85,000, and then raised to 127,000 in the Friday report. The August numbers climbed from a preliminary estimate of zero to 57,000.
"When things are better, they do tend to trend upwards, you do tend to get positive revisions, and when the economy is going downhill, you tend to get negative revisions," Mr. Reitzes said. "For the past while, the economy has been – not stagnating – but bumping along at a low level. You've gone back and forth, oscillating between positives and negatives."
While the revisions may seem large, they need to be put in the context of total non-farm jobs of more than 131 million.
Economists largely concluded that the figures didn't show a U.S. economy bouncing along in perfect health, but they did ease concern of a stall. The jobs growth is insufficient to cut the unemployment rate from 9.1 per cent, economists said.
"There is no hint in September's Employment Report that another recession is starting," Paul Ashworth, of London-based Capital Economics Ltd., said in a report. "Nevertheless, the 103,000 increase in non-farm payrolls last month is still consistent with what would normally be considered very weak economic growth."
What did investors and speculators make of all this? Futures contracts on the S&P 500 index, a benchmark for the U.S. stock market, rose after the labour numbers were released at 8:30 a.m. ET, reversing earlier declines. An hour later when the stock exchange opened, the index was almost unchanged from the previous day's close. It ended the day 0.8-per-cent lower. The U.S. dollar weakened slightly following the report, and Treasury bonds declined.
The last three times the Department of Labour released its monthly Employment Situation (Sept. 2, Aug. 5 and July 8), the S&P 500 fell on the day, by 2.5 per cent, 0.06 per cent, and 0.7 per cent, respectively. Of course, the jobs release was never the only thing going on. This summer, investors have also grappled with the risk of Greece defaulting on its debt and the U.S. losing its top credit rating, among other concerns.
It is normal for statisticians to revise their findings from broad economic surveys, including employment, gross domestic product and manufacturing. The Labour Department says it sends the jobs survey to a sample of about 140,000 businesses and government agencies around U.S. The volume, magnitude and complexity of what needs to be collected, processed and calculated means preliminary results get updated as more data come in.