
The Zhitong Finance App learned that on the evening of November 19, Beijing time, the latest statistics released by the US Department of Labor showed that for the week ending November 14, the number of initial jobless claims in the US after the seasonal adjustment was 203,000, higher than the 198,000 expected in the Econoday survey, and an increase of 1,000 from 202,000 after the previous week's revision. The initial value of the previous week was 200,000.
The four-week average of smooth single-frequency motion increased by 2,000 to 197,750, and is still below 200,000. What changed even more was the renewal data after a week's delay: as of November 7, the number of renewals jumped from 1.693,000 after the revision to 1.763,000, an increase of 70,000 in a single week; the four-week average increased 10,250 to 1,717,000, and the insured unemployment rate remained 1.1%. Compared with the same period last year, initial and renewal requests were still about 8.6% and 9.5% lower, respectively.
Businesses are generally willing to keep their current employees, but those who lose their jobs may take longer to find their next job. Non-farm payrolls increased by 165,000 in October, and the unemployment rate fell from 4.2% to 4.1%, but the median duration of unemployment rose from 11.5 weeks to 11.9 weeks, in line with the signal conveyed by this rise in renewals. Citibank economist Veronica Clark also pointed out earlier that the reduction in the size of the labor force itself may reduce the number of applicants. Therefore, the initial request left room for the Federal Reserve to continue fighting inflation, and the continued increase meant that this report alone was not enough to support the Fed in choosing an aggressive interest rate hike of 50 basis points in December.
The latest initial request and renewal data undoubtedly leaves room for the Federal Reserve to continue fighting inflation. Whether it needs to raise interest rates by 50 basis points at a time depends even more on whether price pressure will continue to spread. The US core CPI has been rising 0.3% month-on-month for three consecutive months. AI investment is driving up the prices of some technology products. Federal Reserve Governor Waller is particularly concerned that price increases in the service sector will continue. Therefore, he proposed raising interest rates by 50 basis points once in December, or adding 25 basis points first, and clearly releasing a signal to raise interest rates in January next year. Before the initial data was released, interest rate futures were priced at about 90% of the 25 basis point rate hike in December, and the probability of directly raising interest rates by 50 basis points was only about 8%. The market is still more inclined to tighten gradually.
At a time when core inflation is still sticking, energy gains are cooling down. Why is Waller rushing to put 50 basis points on the table?
The month-on-month increase in US energy prices fell to 0.7% in October from 5.2% in September, and the overall CPI increase also slowed from 0.6% to 0.4%, but the core CPI rose 0.3% for the third month in a row. The PCE price index, which the Federal Reserve is more concerned about, rose 3.6% and 3.1% year-on-year respectively in September as a whole and core.
Waller said on November 18 that he is particularly concerned about the spread of price increases to the service industry, which is less affected by oil prices, and ultimately changing corporate pricing and consumer expectations. He proposed two paths: directly raise interest rates by 50 basis points in December, or add 25 basis points first, while clearly conveying a signal that interest rates will continue to be raised in January next year. According to the minutes of the October meeting released on the same day, almost all participants thought that further interest rate hikes might be appropriate. Two of them believed that if subsequent inflation data were unfavorable, a range of more than 25 basis points could be considered in December
AI investment distinguishes this round of inflation from a cycle simply driven by overheating employment. Data centers can quickly purchase additional servers and storage, but it takes time to expand chip production capacity. Higher prices paid by companies due to tight supply will also spread to consumer electronics products with shared components.
According to a Federal Reserve estimate released by Federal Reserve Supervisory Vice Chairman Bauman, in the 12 months up to September, AI-related price pressure accounted for about a quarter of the core PCE inflation exceeding the 2% target. Meanwhile, the average hourly wage in the US increased by only 3.1% year-on-year in October. This explains why wage pressure is moderate, and the Federal Reserve's FOMC will continue to discuss interest rate hikes — short-term demand for equipment has been fulfilled, and the effects of AI to increase the productivity of the whole society and reduce unit costs will take longer.
The December rate hike is already a mainstream deal, and the Wall Street disagreement will fall in January next year
Prior to the release of the initial request data, interest rate futures on November 19 had already given a probability of 25 basis points to 4.00% — 4.25% of about 90.7%. The probability of a cumulative increase of at least 50 basis points by the end of January next year is 73.8%; according to LSEG data, the probability that the market will raise interest rates by 50 basis points once in December that day is only about 8%. The most obvious expected warming is the subsequent path.
Market data compiled by the International Monetary Fund shows that from November 6 to 13, the market's pricing of the cumulative interest rate hike up to June next year increased by 39 basis points to 81 basis points in one week. This round of revaluation occurred before the report was requested at the beginning of this week. The main catalysts were higher than expected inflation and hawkish signals from Federal Reserve officials.
I watched Goldman Sachs, Morgan Stanley, and UBS twice; Bank of America, Deutsche Bank, and J.P. Morgan Chase three times. Goldman Sachs expects to add 25 basis points each in December this year and March next year, with interest rates peaking at 4.25% — 4.50%; Morgan Stanley and UBS have the same benchmark path. Seth Carpenter, chief global economist at Morgan Stanley, pointed out the premise of this judgment: inflation will peak in the first quarter of next year, and it will only be conditionally suspended in January. Deutsche Bank and Bank of America included the January interest rate hike in the benchmark forecast after the October CPI was released; J.P. Morgan Chase also announced continuous interest rate hikes for December, January, and March, with 25 basis points each three times, corresponding to the end point of 4.50% to 4.75%.
As far as technology stock trading logic is concerned, the 25 basis points difference between the two paths is particularly direct for two-year US bonds and floating rate financing costs. Take the $10 billion unhedged floating rate debt that is completely repriced with the benchmark interest rate. An additional 25 basis points would mean about $25 million more in interest each year. This is why for data center operators like CoreWeave that rely on quarterly debt expansion, this directly reduces project cash returns. The next key verification is the October PCE announced on November 25: if core price pressure continues to spread, the forecast of continuous interest rate hikes will be more supported; if core inflation cools down and the four-week average continues to rise, Goldman Sachs's judgment of “December first, January observation” will be more persuasive.