ADP payrolls data is expected to show that 175,000 jobs were added in April, but traders are skeptical after it reported a huge gain in March jobs while the government payrolls number fell below 100,000, missing forecasts by a landslide.
The ADP report is scheduled for 8:15 a.m. ET Wednesday, and the government's April jobs data is expected Friday morning.
ADP's data is treated as a kind of preview for the monthly government report, which is typically released two days later. While ADP hasn't been a perfect guide, the market does watch it as a barometer and does take it into account when anticipating the government's monthly report from the Bureau of Labor Statistics.
So when ADP reported March jobs rose a whopping 263,000, well above expectations, it encouraged some economists to up their forecasts for that month's employment report.
But the BLS reported just 98,000 total March nonfarm payrolls, and 89,000 of those were private sector payrolls, stunningly below ADP and also the 185,000 consensus number that economists were forecasting. So the question is do markets trust ADP this time.
"It missed by [165,000] last month, so it's credibility as a predictive measure has been brought into question. More importantly it occurs just hours before the FOMC announcement, and now with Trump saying we could use a good shutdown in September, the political tensions continue to heat up in Washington," said Ian Lyngen, head of U.S. rate strategy at BMO. "I would suspect ADP would be a sideshow to the political show."
Economists expect the government to report 185,000 nonfarm payrolls Friday, according to Thomson Reuters.
ADP (as of March) versus government data (as of February)
Source: JP Morgan Chase
Moody's Analytics Chief Economist Mark Zandi said some of the discrepancy between the numbers has to do with the way they are derived. Moody's crunches the data for ADP.
He said the difference in the two numbers had to do with the fact that there were winter storms in March.
"There's one conceptual difference between BLS and ADP. With BLS, you have to have been paid during the survey week. ADP is simple. You have to be on the payroll. ... The bar for BLS is higher," explained Zandi.
He said ADP data counts as employed those workers who are on the payroll but can't make it to work due to weather. The government data, however, will not count a worker who does not show up to work for a week due to weather, if they are not paid.
"For BLS, you have to be on the payroll and you have to have been paid during the survey week," said Zandi.
"You saw a big decline in the BLS number and in the number of industries that were sensitive to weather. Construction is the best example."
Many economists included forecasts for those winter storms in their estimates for nonfarm payrolls, and they still were wrong.
Separately, ADP said in a statement that it bases its report on real-time payroll data, versus the government's survey data. It also said besides the weather factors, the gains in its last report may have reflected small business hiring that may not have been picked up in the government report.
Jim Caron, fixed income portfolio strategist at Morgan Stanley Investment Management, said the market will still watch ADP.
"It's hard to predict that number because it's all over the place, but people are going to look at it. It was a little misleading last time. People are going to look at that number because it's an early view of what might come out Friday. It's important," he said.
Caron said it's conceivable that both the March government report and the ADP number are revised, and that could put them closer together.
Lyngen said both reports have their flaws. "BLS is not a perfect gauge of how many jobs are being created, but it's the one with the longest history and it's the one the markets are watching," he said.
Watch: Pro expects big swing back in jobs
