On the evening of August 21st (Wednesday) Beijing time, the US Bureau of Labor Statistics (BLS) will release the preliminary report of the Quarterly Census of Employment and Wages (QCEW) for the first quarter, which includes revisions to non farm employment data for the one-year period from April 2023 to March 2024.
Tonight's correction may mean that the US Bureau of Labor Statistics has severely overestimated the number of new non farm payroll jobs in the United States, and the degree of exaggeration may even be the highest in 15 years.
Earlier this month, after the far less than expected July non farm payroll data triggered a widespread panic on Wall Street, the US stock market experienced a large-scale sell-off. Therefore, investors will closely monitor the revised data released tonight to determine whether the market's initial reaction was correct.
According to a report by Barron's Weekly, if the data released tonight shows a significant downward revision in the number of new non farm payrolls, it may reignite the debate on the necessity for the Federal Reserve to accelerate or intensify interest rate cuts, and even affect Federal Reserve Chairman Powell's speech at the Jackson Hole Global Central Bank Annual Meeting on Friday.
Jonathan Millar, Senior US Economist at Barclays Bank, stated in a comment email to the Daily Economic News that although he also believes that tonight's data is likely to be a significant downward revision, the final (actual) magnitude of the revision will be greater than initially estimated.
Even with a massive reduction of one million, employment growth remains at a healthy level
The QCEW data comes from unemployment insurance tax records in various states in the United States, covering almost all job positions, so the accuracy of this data is much higher than the monthly non farm payroll data, but the disadvantage is that the timeliness will be much worse.
The revised data has not been released yet, but the expectations of major Wall Street investment banks are not optimistic. JPMorgan expects BLS to cut 360000 new non farm jobs added in the year ending in March this year; Economists at Wells Fargo expect to cut 600000 jobs; Goldman Sachs even predicts that in the worst case scenario, the downward revision could reach one million job positions.
Goldman Sachs
According to foreign media reports, if the data is revised downwards by more than 501000 people, it will be the largest downward adjustment in 15 years, indicating that the cooling of the US labor market will take longer than originally thought, and may even last longer.
According to data previously released by BLS, the United States added a total of 2.9 million non farm jobs from April 2023 to March 2024, with an average monthly increase of 242000 jobs. However, even if Goldman Sachs lowered its maximum forecast by 1 million, 1.9 million new non-agricultural jobs had been created in the 12 months up to March this year, that is, the average number of jobs added each month was about 158000, but this is still a level of employment growth that economists regard as healthy, only slowing down compared with the peak period of employment after the COVID-19 epidemic.
Investment bank: In recent years, there has been a significant "downward deviation" in QCEW
Affected by the significant downward revision of QCEW expectations, the US dollar continued to weaken on Tuesday. As of press time, the US dollar index has fallen to a new low since January 2nd this year, while other non US currencies rose across the board during the three week trading session.
Bloomberg
The global market is closely monitoring the Jackson Hole Global Central Bank Annual Meeting to be held on Thursday Eastern Time, as well as Powell's guidance on the Fed's subsequent interest rate cuts at the meeting. QCEW data may affect the tone of Federal Reserve Chairman Powell's speech.
If the data released on Wednesday really shows a situation where "US non farm payroll data is revised down by one million", Powell's wording may trigger a different trading rhythm compared to the current market celebration of a "soft landing of the US economy".
According to the "Federal Reserve Watch" tool of Zhishang Institute, as of press time, traders currently expect the Federal Reserve to cut interest rates by a total of 100 basis points within the year: 25 basis points each in September and November, and 50 basis points in December.
Zhishang Institute
It should be noted that QCEW has two obvious limitations. Firstly, QCEW largely excludes illegal immigrants, whose employment is included in non farm employment data. This alone could lead to a downward revision of data ranging from 300000 to 500000; Secondly, the QCEW data to be released tonight is also the initial value, and the final figures in recent years are on average 100000 people higher than the initial value.
Jonathan Millar, Senior US Economist at Barclays Bank, stated in a comment email to the Daily Economic News that there is still a significant "downward bias" in the QCEW released by BLS in recent years, so there may be some deviation in the authenticity of the data released tonight.
Millar believes that the preliminary data released by BLS on Wednesday is likely to be a significant downward correction. The latest QCEW estimate as of December 2023 suggests that, on the surface, the cumulative growth of non farm employment from business surveys from March 2023 to December 2023 is 720000 (80000/month) higher than the corresponding QCEW estimate.
"Although the possibility of BLS data downward revision on Wednesday is high, we strongly suspect that the (actual) downward revision will be much smaller than the preliminary estimate. As we emphasized in previous studies, since the COVID-19 epidemic, the QCEW estimate has been significantly downward biased. This became very evident last year, when the early downward revision on new non farm employment was largely offset by the subsequent upward revision on QCEW. If the recent pattern remains unchanged, BLS is likely to significantly increase its initial value when it releases the final QCEW values for the 12-month period ending in March 2024 on December 5th of this year. The 2023 correction value, which will be announced on September 4th, will provide some preliminary clues about the possible magnitude of the correction Millar added.
