[Table_Page] 宏观经济研究报告 2026 年 2 月 12 日 证券研究报告 美国 1 月就业数据公布之后 [Tabl e_Author] 分析师： 郭磊 分析师： 陈嘉荔 SAC 执证号：S0260516070002 SFC CE.no: BNY419 SAC 执证号：S0260523120005 021-38003572 021-38003674 guolei@gf.com.cn gfchenjiali@gf.com.cn 请注意，陈嘉荔并非香港证券及期货事务监察委员会的注册持牌人，不可在香港从事受监管活动。 [Table_Summary] 报告摘要： ⚫ 根据美国劳工部 2 月 11 日公布数据1，1 月新增非农 13 万人，高于预期的 7 万人，前值 4.8 万人，且显著高于 达拉斯联储2估算的 3 万人/月就业平衡水平（即每个月使劳动力市场保持平衡的新增岗位数量）。私人部门新 增就业 17.2 万人，高于预期的 7.5 万人，前值 6.4 万人。从趋势上看，新增就业和私人部门新增就业人数 3 个 月平均值均显著回弹。就业扩散指数从 54.2%小幅回升至 55%，显示就业增长的行业覆盖面有 ...

Group 1 - The upcoming labor market report is expected to show a slowdown in job growth in December due to increased import tariffs and cautious hiring related to AI investments, with the unemployment rate projected to drop to 4.5% [1] - Economists estimate that the U.S. economy needs to create between 50,000 to 120,000 jobs monthly to keep up with the growth of the working-age population, but job creation has significantly lagged behind this requirement [2] - The sharp slowdown in job growth last year is attributed to aggressive trade and immigration policies, which reduced both demand and supply for workers, with a notable impact from the government shutdown affecting data collection [3] Group 2 - While the median forecast from economists suggests a drop in the unemployment rate to 4.5% in December, some predict a slight increase to 4.7%, which could open the door for potential rate cuts by the Federal Reserve [4] - The Federal Reserve lowered the benchmark interest rate by 25 basis points to a range of 3.50%-3.75% in December, but officials indicated a pause in further cuts to better assess economic trends, with job growth expected to remain narrow and concentrated in healthcare and social assistance sectors [5]

Group 1 - The Federal Reserve has prioritized concerns over the labor market, leading to a 25 basis point rate cut with a 9-3 voting outcome [1] - There are indications that if the labor market remains weak, policymakers may lean towards further rate cuts in the future [1] - Jerome Powell highlighted that employment growth may have been negative in recent months, suggesting a need for more accommodative monetary policy [1] Group 2 - Powell referred to a "systematic overestimation" of job growth, with a preliminary estimate indicating that employment growth was overestimated by 911,000 over the past 12 months [2] - The balance between supporting the labor market and controlling inflation will be central to policy-making as the Fed approaches 2026 [2] - There is a divergence among Fed officials regarding the direction of interest rates, with some opposing the recent cut and others suggesting further easing may be necessary [2] Group 3 - If inflation is subsiding while the labor market remains weak, the Fed is expected to adopt a more accommodative stance, especially with Powell set to step down in May [3] - Market expectations indicate that conditions for further rate cuts will be met if labor demand weakens and unemployment rises [3] - Futures markets suggest that the next rate cut may not occur until April of next year, with a possibility of two or even three cuts in 2026 [3]

Group 1 - Morgan Stanley's latest report revises down the market's perception of U.S. employment resilience, indicating a downward adjustment of 910,000 jobs for March 2025, with 880,000 in the private sector [1] - The average monthly job growth from March 2024 to March 2025 has been revised down from 147,000 to 71,000, reflecting a significant decrease in employment growth expectations [1] - The report suggests that the Federal Reserve may have anticipated this adjustment, implying that it will not affect the decision to lower interest rates by 25 basis points in September [1] Group 2 - The benchmark revision only reassesses employment stock prior to March 2025, meaning subsequent monthly changes are not mechanically reduced, which may temporarily alleviate concerns about a sudden cooling labor market [3] - There are two critical warnings for policymakers: the negative revisions for 2023, 2024, and 2025 indicate a potential systematic overestimation in the survey framework, and revisions tend to be pro-cyclical, suggesting that if unemployment continues to rise, further significant downward adjustments are likely [3][4] - The report identifies that the birth-death model has consistently overestimated net job additions, leading to an overcount of 256,000 jobs in 2024, and while adjustments were made for 2025, the gap between actual openings and closures remains narrow, indicating potential for continued negative errors [4] Group 3 - The report argues that the lack of empirical support for the impact of illegal immigration on job numbers suggests that the recent decline in immigration may not be the primary reason for repeated negative revisions [4] - Investors should monitor three key developments: the new seasonal adjustment factors to be released in January, updates to the birth-death parameters that will affect job additions post-April 2025, and the downward adjustment in employment numbers that will also impact hours worked, potentially revising Q1 non-farm productivity growth from 1.2% to approximately 1.7% [4]

Core Viewpoint - The upcoming annual benchmark revision of non-farm payroll (NFP) data by the U.S. Bureau of Labor Statistics (BLS) is expected to reveal a significant downward adjustment of 550,000 to 800,000 jobs, which may impact market confidence in the U.S. labor market and prompt the Federal Reserve to consider a substantial interest rate cut of 50 basis points, similar to actions taken in September of the previous year [2][5]. Group 1: Reasons for Data Overestimation - The primary reasons for the overestimation of employment data include the flawed "birth-death model" used by BLS, which inaccurately estimates job creation from new businesses without relying on actual business registration or tax data [3][4]. - A significant reduction in illegal immigration has also contributed to the overestimation, as the labor supply has been affected, leading to a decrease in the actual need for new jobs [8]. - Historical trends indicate that original employment data is often revised downward during economic slowdowns, a pattern observed since 1979 [10]. Group 2: Additional Indicators of Data Issues - Seasonal adjustment models may misinterpret genuine trends as seasonal fluctuations, leading to delayed corrections in employment data [9]. - ADP data has raised questions about BLS's overestimation of job growth in the healthcare sector, suggesting that the actual growth is not as robust as reported by BLS [11]. - Household surveys may be overestimating population and employment growth, with current models potentially overstating U.S. population growth by about 1 million, which could lead to an overestimation of employment growth by approximately 50,000 jobs per month [12].

Core Viewpoint - The upcoming annual benchmark revision of non-farm payroll (NFP) data by the U.S. Bureau of Labor Statistics (BLS) is expected to reveal a significant downward adjustment of 550,000 to 800,000 jobs, which could impact market confidence in the U.S. labor market and potentially lead the Federal Reserve to consider a 50 basis point rate cut in September [1][2]. Group 1: Reasons for Data Revision - The primary reasons for the anticipated downward revision include the distortion of the birth-death model, which overestimates job creation by new businesses, and a significant reduction in illegal immigration, leading to a systematic overestimation of the labor force [1][2]. - Estimates suggest that these biases may result in an overstatement of actual employment by 40,000 to 70,000 jobs per month, accumulating to a total overstatement of 550,000 to 800,000 jobs annually [1][2]. Group 2: Implications for Federal Reserve Policy - If the upcoming revision mirrors last year's adjustment, where the BLS revised down by 800,000 jobs, the Federal Reserve may face a decision on whether to implement a substantial 50 basis point rate cut, reinforcing that last year's cut was based on genuine economic slowdown rather than political compromise [1][2]. Group 3: Additional Signals of Data Inflation - Goldman Sachs identified at least five additional reasons indicating the employment data may be inflated, beyond the birth-death model [4]. - A notable factor is the recent decline in illegal immigration, which has significantly impacted labor supply and job creation estimates [5]. - Seasonal adjustment models may misinterpret genuine trends as seasonal fluctuations, leading to delayed corrections in employment data [6][7]. - Historical patterns show that original employment data is often revised downward during economic slowdowns, a trend observed since 1979 [8]. - ADP data challenges BLS's overestimation of job growth in the healthcare sector, suggesting that the actual growth is not as robust as reported [9]. - Household surveys may overestimate population and employment growth, with current models potentially overstating annual population growth by 1 million, leading to an overestimation of job growth by approximately 50,000 jobs per month [10].

Core Viewpoint - The upcoming annual benchmark revision of non-farm payroll (NFP) data by the U.S. Bureau of Labor Statistics (BLS) is expected to reveal a significant downward adjustment of employment figures, potentially by 550,000 to 800,000 jobs, which could impact market confidence and lead to a 50 basis point rate cut by the Federal Reserve [1] Group 1: Reasons for Data Revision - The primary reasons for the anticipated downward revision include the distortion of the birth-death model, which overestimates job creation from new businesses, and a significant reduction in illegal immigration, leading to an overestimation of the labor force [1][2] - Estimates suggest that these biases may result in an overstatement of actual employment by 40,000 to 70,000 jobs per month, accumulating to 550,000 to 800,000 jobs annually [1][2] Group 2: Employment Data Analysis - Goldman Sachs indicates that the BLS's birth-death model is a major source of employment data distortion, as it relies on estimations rather than actual business registration or tax data, making it prone to systematic overestimation [2] - The Quarterly Census of Employment and Wages (QCEW) and Business Dynamics Statistics (BDM) are considered more reliable benchmarks for employment data, as they are based on actual unemployment insurance records [2][3] Group 3: Employment Trends and Signals - From early 2024, established companies are reportedly adding only 25,000 jobs per month, while the BLS estimates new companies contribute over 100,000 jobs monthly, a discrepancy highlighted by BDM data [3] - To maintain a balanced labor market, a reasonable level of non-farm employment should be around 170,000 jobs per month, with 100,000 from natural growth and 70,000 from model overestimations [3] Group 4: Additional Indicators of Data Issues - Goldman Sachs identifies five additional signals indicating the employment data may be inflated, including a decrease in illegal immigration, seasonal adjustment model inaccuracies, historical patterns of data revisions during economic slowdowns, discrepancies in healthcare employment growth compared to ADP data, and potential overestimations in household surveys regarding population and employment growth [4][5][6][7][8][9]

Group 1 - The core viewpoint indicates that the significant downward revision of June non-farm data reflects substantial disruptions to the U.S. economy due to tariffs, suggesting that the resilience of the U.S. economy should not be overestimated, and the direction of interest rate cuts remains highly certain [2][4][17] - In July, non-farm employment increased by 73,000, which is below the expected 110,000, and the previous value was revised down from 147,000 to 14,000, indicating pressure on the U.S. job market [6][11][22] - The unemployment rate in July rose to 4.2%, up from 4.1% in the previous month, while the average hourly wage increased by 3.9% year-on-year, exceeding the expected 3.8% [1][6][31] Group 2 - In July, the financial activities, healthcare, and retail sectors added 15,000, 79,000, and 16,000 jobs respectively, showing a stable demand in the service sector [3][22] - The manufacturing sector has seen negative job growth for three consecutive months, indicating insufficient production willingness among enterprises [3][22] - The labor force participation rate decreased to 62.2% in July, down from 62.3% in the previous month, with a notable decline in employment willingness among younger demographics [26][31] Group 3 - The downward revision of June non-farm data was primarily due to significant adjustments in government, leisure, and hotel employment, which collectively accounted for a 90,000 downward revision, representing nearly 70% of the total revision [12][17] - The cumulative downward revision for May and June non-farm data reached 258,000, while the July employment figure of 73,000 is a significant drop compared to the average monthly increase of over 100,000 in the first quarter [4][17] - The market anticipates that the Federal Reserve will cut interest rates three times in 2025, with an 80% probability for the first cut in September [4][21][37] Group 4 - The average hourly wage growth has shown an upward trend, with a month-on-month increase of 0.3% in July, higher than the previous 0.2% [37][39] - The service sector's job growth in July rebounded to 96,000, compared to a previous value of 16,000, indicating a relatively stable demand in the service industry [22][31] - The overall economic environment remains challenging, with second-quarter GDP growth at 3.0%, driven by a "import rush" effect, while core GDP growth has declined [18][22]