Source: BIT Securities

Last night at 20:15, ADP released the August private sector employment data: only 38,000 jobs were added, significantly lower than the market expectation of 47,000 to 48,000, and also below the revised 46,000 from July, marking the lowest value since January this year. After the data was released, the market reacted quickly, with CME Fed rate futures retreating from a 25BP probability of a rate hike in September, and all three major indices closing higher, while the Russell 2000 index, which had dropped over 1.2% during the day, closed up 1.13% at 2953. The small-cap stocks, most sensitive to interest rates, drew a big V shape in just one day.

In the vast U.S. employment market, why can such a small number of tens of thousands repeatedly stir the U.S. stock market, U.S. bonds, and gold, even influencing the Fed's rate hike rhythm? How will the market react after today? What impact will it have on the Fed's FOMC meeting decision next week?

What is ADP's "Small Non-Farm" Employment Data?

ADP is one of the largest payroll service providers in the world, processing payroll for over 500,000 U.S. businesses and approximately 26 million private sector employees. This coverage accounts for about one-fifth of private employment in the U.S., meaning that roughly one in six American workers has their payroll processed through ADP's system. ADP typically releases its data one to two trading days before the monthly non-farm employment data, usually on Wednesday mornings in New York. It is the first systematic employment signal available to the market for the month.

In a country of 300 million, how can just a few tens of thousands affect the whole system?

In rapidly growing economies, job additions can easily reach hundreds of thousands or millions, and fluctuations of tens of thousands can indeed be ignored. However, the U.S. is a mature and saturated economy, and the employment market has long entered a steady state, with monthly net additions being a very small number. In such a market, marginal changes of tens of thousands are sufficient to determine whether employment is warming up or cooling down. Because its value has never been in absolute numbers, but in marginal direction.

Are the details of the data more alarming than the total? How significant are the employment risks in the U.S.?

The total of 38,000 actually conceals a significant internal divergence.

The entire private sector is being propped up by the service industry. The service sector saw a net increase of 48,000, with education and healthcare contributing 45,000; however, manufacturing alone decreased by 17,000, and the professional and business services sector, often seen as an economic barometer, directly reduced by 16,000. In other words, growth is highly concentrated in a few defensive industries, while most pillar industries for investment and development are actually contracting or stagnating.

The divergence in enterprise size is even more striking. Large enterprises contributed 34,000 new jobs, but medium-sized enterprises had almost zero net additions, and small enterprises only added 3,000, revealing the awkward state of U.S. employment earlier than the overall data: small and medium-sized enterprises are most sensitive to interest rates and financing costs, and their hiring willingness has cooled first.

However, the wage components provided a signal in the opposite direction. The base salary for retained employees increased by 3.0% year-on-year, but the total salary for job changers surprisingly rose by 7.3%. Job changers still earn significantly more than retained employees, indicating that companies are still willing to pay a premium to attract talent.

This "job-hopping" enthusiasm remains strong, indicating that nominal wage stickiness is still present. Hiring is cooling, but wages have not followed suit—this is precisely the combination that inflation is most unwilling to see, and also the strongest card in the hands of the hawks.

Can the dovish cooling of ADP save the market?

The "carrier pigeon" brought by the ADP data has been flying solo for most of the day but has not been able to escape the shadow of geopolitics. After the data was released, the yield on the 10-year U.S. Treasury bonds did indeed drop, but the pricing power throughout the morning remained firmly in the hands of the Strait of Hormuz—Trump threatened to retaliate "more fiercely," and Brent crude rose above $95. The inflation risk brought by supply shocks is further away than the flying bullets. As long as the oil price shock continues, the Fed has no room to retract the phrase "anti-inflation priority." This is the logic of the first half.

The turning point appeared after New York Federal Reserve Bank President Williams spoke. His statement that "there has been no unusual spillover effect from high energy prices" implies that if the energy shock does not spill over into core inflation, it is merely a one-time shock, not a trend. He added that inflation expectations "remain well-anchored," wage growth "is being restrained," and he clearly stated that he has not been convinced about "the need for a rate hike."

Thus, the same data was read twice on the same day, interpreted in the morning as "inflation risk outweighs everything," and in the afternoon as "employment is cooling while inflation spillover is limited." This led to a textbook-level intraday reversal, with the Russell 2000 index bouncing from a drop of over 1.2% during the day to close up 1.13%, with a daily fluctuation exceeding 2.3 percentage points. Small-cap stocks are the most sensitive to interest rates in the entire market, and their reversal magnitude is the true reading of the day's change in interest rate expectations.

What needs to be verified next: How will employment data influence the Fed?

Wash stated clearly last week at Jackson Hole—"We must be confident that inflation is clearly returning to target," otherwise "there is still work to do." With no discussion on the inflation front, employment should be the only variable in policy; if employment weakens, tightening should slow down.

The non-farm data to be released this Friday night will be the next harder verification. The current market consensus is an increase of 50,000 to 55,000 jobs, with the unemployment rate maintaining at 4.1%. Several major investment banks have also provided their professional estimates: Goldman Sachs and Crédit Agricole both expect 65,000, while Wells Fargo optimistically predicts 80,000. When the data is released, three scenarios may occur:

Falling within the estimated range—uncertainty. This is the most probable scenario and also the hardest to trade: it neither proves that employment is collapsing nor that it remains resilient. Market focus may shift to the August CPI to be released in mid-September, and volatility may actually converge, with the market waiting for the next catalyst. In this scenario, sector rotation may explain issues better than the overall index itself.

Stronger than the upper estimate peak—rate hikes may be locked in again. This will challenge the narrative of rapidly weakening employment, and the probability of a rate hike in September may continue to rise from 70%, with the dollar and U.S. Treasury yields possibly strengthening simultaneously. In this scenario, weaker assets may include overvalued growth stocks and small-cap stocks: the rise in discount rates impacts valuations rather than earnings themselves. Gold may come under pressure, but geopolitical risk premiums may partially offset this impact—thus, it is more likely to show a gradual decline rather than a cliff-like drop.

If the unemployment rate rises to above 4.3%—the doves will truly have ammunition. This would substantially suppress the probability of rate hikes, possibly accompanied by a weaker dollar, a rebound in U.S. Treasuries, and lower real yields pushing gold back up. However, caution is needed for second-order risks—if non-farm data is negative for two consecutive months, the trading narrative may jump from "hike or hold" directly to "are we heading into recession," at which point even rising expectations for rate cuts may not save the stock market, as what will be repriced then is earnings themselves.

How to guard against the war situation?

Last night's market reminded us that once an external shock event of this magnitude, such as geopolitical tensions, intervenes, inflation risk will once again become the dominant variable, and the directional weight of employment data will be temporarily suppressed.

If passage through the Strait of Hormuz continues to be obstructed and oil prices keep rising, inflation expectations will rise again, and long-term yields and rate hike expectations will continue to climb together. At this point, the most to be wary of may not only be a unilateral decline in the stock market but also a stagflation-style retreat that hits both stocks and bonds—this is precisely a repeat of the market from late February to late March, when the S&P saw its deepest pullback of 9.1%, and the VIX surged to 31. The yield data on 30-year long bonds will be a key threshold; once effectively broken, all assets relying on low discount rates may be repriced.

However, if there is an unexpected easing or ceasefire, this direction may be more easily missed. Last night's second half actually previewed this: once the judgment of "no spillover from energy" is confirmed, yields may fall, and U.S. Treasuries and growth stocks will welcome a repair window. However, caution is needed regarding changes in gold prices: its current price contains both "geopolitical risk premium" and "high real interest rate suppression," two opposing forces. Once the risk premium disappears, gold may fall before stocks. This year, gold has retraced 25% from its peak of $5,318 in January, and this lesson is not far off.

Data Sources:

·ADP Official Press Release: 38,000 Jobs Added in August Private Sector (Including Industry, Enterprise Size, and Wage Components)

·CNBC: Private Sector Added 38,000 Jobs in August, Below Expectations

·Fox Business: Details of ADP August Report and July Revised to 46,000

·CNBC: New York Fed President Williams Says Yield Surge Due to Strong Economic Prospects

·Bond Buyer: Williams Not Convinced That "Rate Hike" Is the Answer Yet, Advocates Watching and Waiting (Including Energy Spillover, Inflation Expectations Anchored, etc.)

·Investrade: September 2 Market Close (Three Major Indices, Russell 2000, Oil Prices, Gold, and Sector Performance)

·Barchart: Dollar Retreats from Two-and-a-Half-Week High, Gold and Silver Rise, September Rate Hike Probability 65%

·Federal Reserve H.15: Official Closing Values of Treasury Yields on September 2 (2-Year 4.39%, 10-Year 4.79%, 30-Year 5.27%)

·FXStreet: Dollar Remains Strong Initially After Data Release, September Rate Hike Probability Once Around 70%

·Blockonomi: Spot Gold Drops to Three-Week Low

·TheStreet: September 2 Market and Strait of Hormuz Incident

·Forbes: August 31 CME FedWatch Shows 66% Probability of Rate Hike in September

·CNBC: After Jackson Hole, September Decision Becomes "A Coin Flip," Rate Hike Probability Rises

·CNBC: Rate Hike Probability Declines After August 7 Non-Farm Data Misses Expectations

·TOPONE Markets: August Non-Farm Preview and Three Scenarios (Consensus +50,000 to 55,000)

·Dallas Fed: Breakeven Employment Growth Has Dropped to Near Zero

·Pew Research: Comparison of ADP and BLS Data Coverage, Methods, and Correlations

·Federal Reserve: FOMC Meeting Calendar (September 15-16 Including Dot Plot)

Disclaimer: This market data is as of the close of the U.S. stock market on September 2, 2026. This content is for informational and educational reference only and does not constitute investment advice, investment offers, or invitations to purchase or sell any financial products. The securities and market views mentioned are for introducing index compilation and passive investment mechanisms and do not represent BIT's recommendations or endorsements. Historical performance and market reactions from index adjustments do not represent future performance. Investment involves risks, and securities prices may fluctuate; investors may lose part or all of their principal and should make prudent judgments based on their own circumstances. Relevant services and products are subject to applicable laws, regulations, and regional restrictions of the jurisdiction.