Non-farm payrolls are off the charts—September rate-hike odds jump to 60%

资管舵 · 2026-09-09T08:50:16.000Z

The non-farm payroll data pushed the Fed’s September rate-hike probability up to 60%. In August, employment added 162,000 jobs, while the market expectation was only 55,000—so the gap was too big, and the market immediately repriced. I checked a historical backtest from Shenwan Hongyuan. Since 2015, whenever the rate-hike expectations before a meeting exceeded 40% in the first 10 days, it has never missed—20 for 20. Conversely, if the outcome were to fall short, the long-end term premium would likely get punished. 2018 and July this year are examples: the 10-year term premium rose by 5 and 6.2 basis points, respectively. The disagreement in the July rate decision—9 to 3—hasn’t fully resolved yet, and Woz(h) also turned hawkish by late August. The only variable now is August CPI. After running 10,000 Monte Carlo simulations, the probability that CPI comes in significantly below expectations is only 10.6%. Plus oil prices are also holding firm, so it’s hard for the high rate-hike expectations to fade.

非农数据把美联储9月加息概率顶到了60%。8月新增就业16.2万，市场预期才5.5万，差距太大，市场直接重定价。
我看了下申银万国回测，2015年以来但凡开会前10天加息预期过40%，从来没落空过，20次全落地。反过来，真要落空，长端期限溢价会被反噬，2018和今年7月就是例子，10年期限溢价分别往上走了5和6.2个基点。7月议息9比3的分歧还没消，沃什8月底也转鹰了。
现在唯一变量是8月CPI。机构跑了一万次蒙特卡洛，CPI显著低于预期的概率只有10.6%，油价又绷着，所以高加息预期很难消。
我的看法，9月加一次是沃什当前最不坏的选择。加完只要点阵图别大幅上修路径，对美股和美债短端冲击有限，加密这种风险资产也扛得住。真正要防的是加完还连续加，那才伤流动性。#美加关税战升级 $VVV 
VVVUSDT
Perp
23.51
-7.69%
 