Success or failure depends on non-farm payrolls?

Since July, the sharp decline in non-farm payrolls has continued to affect the nerves of the Federal Reserve and the market. Although other core employment indicators such as the unemployment rate and wages have not significantly deteriorated, there is no disagreement that the ongoing weakness in new non-farm payrolls is the "culprit" forcing the Federal Reserve to reassess employment risks and consider a "risk control rate cut" in September.

However, we believe that the pace of rate cuts by the Fed may be more convoluted than the market's expected linear rate cut path. We have continuously pointed out in previous reports that potential inflation risks in the fourth quarter will become a "roadblock" to continuous easing policies. Currently, discussions about inflation have been quite thorough.

In this report, we will focus on another key variable—employment, and whether there is a possibility of upward revision, which could inversely suppress the rate cut expectations already priced in.

The "noise" in this year's non-farm data has become increasingly prominent: The response rate to surveys has significantly declined, and layoffs by the Federal Reserve have affected the quality of data collection, leading to doubts about its accuracy. However, as the most core official indicator of the labor market, both the market and the Federal Reserve still have to rely on its fluctuations for pricing and decision-making.

On the other hand, the seasonal adjustment mechanism and model characteristics of non-farm payrolls have also amplified the short-term volatility of the data. The Department of Labor incorporates the newly released data for the month into the seasonal adjustment model each month, generating new seasonal adjustment factors and retroactively revising the non-farm employment data for the past three months. July and August are often peak periods for seasonal fluctuations in the labor market, and once included in the model, the dynamic adjustments of the seasonal adjustment factors may increase the short-term disturbances of the data from the past three months, leading to consecutive large downward revisions in non-farm payrolls.

However, historical experience shows that the initial value of August's new non-farm payrolls announced in September often undergoes seasonal upward revisions in October. Referring to data from the past 20 years (2005-2024), the probability of upward revision for August non-farm payrolls in the two revisions in October exceeds 80%, the highest among all months, while the final value announced in November also has a probability of upward revision of around 70%.

The overall upward revision magnitude should not be underestimated, as both the median and average values of the subsequent upward revisions for August non-farm payrolls rank among the highest for the year. Given that July non-farm payrolls have already shown signs of upward revision, combined with White House economic advisor Hassett's hint that "the employment report may be revised upward by nearly 70,000 jobs," the potential upward revision of August non-farm payrolls cannot be ignored in terms of its impact on rate cut expectations.

We believe the reasons for the seasonal upward revision of the August non-farm payrolls may be:

On one hand, after the significant revision of the annual benchmark in September, the overall trend component of non-farm payrolls in the CES model has clearly shifted downward (the trend component can be understood as the reference frame for seasonally adjusted non-farm data, and the downward shift makes the originally relatively "weak" data appear relatively "strong"), therefore, there is not much room for downward revision in the second revision of the August non-farm payrolls released in October;

On the other hand, the unadjusted data for August non-farm payrolls often shows a significant seasonal uplift, but the initial value did not incorporate all the survey responses when it was released (especially since the current response rate has significantly declined), and the subsequent recovered surveys may reflect a more optimistic employment tendency.

In horizontal comparison with other labor market indicators, there is also a probability of underestimating the August non-farm payrolls. The correlation of indicators such as ADP employment, PMI employment index, and small business hiring plans with the final value of non-farm payrolls is generally higher than that with the initial value, and the current downward trend of these indicators in August is controllable, not showing the severe slowdown seen in the initial value of non-farm payrolls. This comparison also suggests that there is a certain probability of upward revision for the August non-farm payrolls in the future.

Therefore, the guidance of the August non-farm payrolls on interest rate cut expectations may be disturbed, a similar situation also occurred around the first interest rate cut by the Federal Reserve last year: In June and July last year, the non-farm payroll data was continuously revised down, and the rising unemployment rate triggered market recession fears, prompting the Federal Reserve to cut rates in September, with the market pricing in a 50 basis point cut in November. However, in October, the Labor Department significantly revised up the non-farm payroll data for July and August, leading to a cooling of the recession narrative and the aggressive expectation of a 50 basis point cut.

In terms of asset reflection, U.S. Treasury yields, especially at the short end (2-year), surged about 20 basis points on that day; the dollar rose, and the prices of precious metals such as gold, silver, and copper showed a significant short-term correction; U.S. stocks experienced a slight short-term pullback but ultimately continued to rise under the retreat of the recession narrative and the support of moderate interest rate cuts. Of course, unlike last time, the current round of asset pricing is more driven by bets on interest rate cuts, without the participation of the recession narrative, which also means that the upward revision of non-farm payrolls may have a greater negative impact on risk assets.

In summary, we need to continue to observe the expectations for interest rate cuts. Although we do not rule out the possibility of the Federal Reserve continuously cutting rates from the fourth quarter to early next year (if there are clear signals of rising unemployment, etc.), the current market bets on interest rate cuts seem somewhat overly aggressive, overlooking potential volatility risks Even though there are still differences among Federal Reserve officials, market pricing shows a rare "consistent optimism."

This also means that once any reverse signals (such as rapid inflation increases and non-farm payroll revisions) lead to adjustments in expectations, market volatility will be amplified again: U.S. Treasury yields may rebound, and U.S. stock sectors that are relatively sensitive to interest rate cuts (growth stocks and cyclical stocks) will be affected, while the rise of metals with strong financial attributes (gold, silver, copper) will slow down. Recently, Powell's statements have been quite significant—emphasizing that "interest rates remain moderately restrictive" and "stock valuations are high," which may intentionally aim to cool down the overheated optimistic expectations.

Authors of this article: Wu Shuo, Lin Yan, Source: Chuan Yue Global Macro, Original Title: "Success or Failure Depends on Non-Farm Payroll?"