August CPI: A report that appears hawkish but is actually dovish and a "pre-emptive" trade in interest rate hikes.

1. August's CPI rose 0.40% month-on-month, in line with expectations, but core CPI rose 0.29% month-on-month, significantly higher than the market consensus expectation of 0.2%. Energy (gasoline +4.24%) contributed approximately 0.13 percentage points to the increase, core goods rose 0.11%, and core services rose 0.33%. The higher-than-expected increase this month came almost entirely from core services. 2. The rebound in core services month-on-month came from two sub-items: a) Telephone services increased by 5.37% month-on-month, the highest month-on-month growth rate since data became available, contributing 0.09pp to the core month-on-month increase (compared to an incremental contribution of 0.08pp in July), making it the largest single contributor; b) Hotel prices reversed from -2.75% to +2.36% (an incremental contribution of 0.08pp), driving the Shelter category month-on-month increase from 0.14% to 0.26%. The two items combined contributed an incremental increase of 0.16pp. The main drag was medical services, which decreased by 0.25% month-on-month (+0.56% in July), contributing -0.07pp. Figure 1: Distribution of Contributions to Core Inflation 3. The jump in telephone service inflation is due to two factors. First, starting in July 2025, the BLS will switch to an alternative data method for wireless services (using industry-wide package prices + monthly hedonic quality adjustments), making this category highly sensitive to overall operator price adjustments. Second, August coincides with a concentrated round of price increases by operators following the year-end promotional campaigns and equipment tariff cost pressures. Historically, such jumps are not sustainable. In the past nearly 30 years, this sub-item's month-on-month increase exceeding 1.5% has only occurred four times. Under the old method, the contribution in the following month always returned to near zero, and under the new statistical method, even a reversal is common (for example, after +2.0% in May this year, it was -3.0% in June). Therefore, the contribution of telephone services to the core CPI in September is highly likely to be zero or negative. Figure 2: Month-on-Month CPI for Telephone Services (Image 2: https://img.jinse.com.cn/7509521_image3.png) (Image 3: B0gxXfFx8hOYPcfjY9gjA3Aw3HwkqjMRpvrmQOf4.png) (Image 4: B0gxXfFx8hOYPcfjY9gjA3Aw3HwkqjMRpvrmQOf4.png) (Image 5: Sticky housing inflation continues to improve slowly.) The Ordinary Equivalent Rent (OER) decreased month-on-month from +0.26% to +0.19%, and the primary residence rent decreased from +0.26% to +0.17%. The combined contribution of rent to core inflation decreased from +0.11pp to +0.08pp. Excluding telephone and hotel services, core inflation in August was approximately +0.17% month-on-month, slightly lower than July and close to market consensus. 6. Because telephone services are included in the supercore inflation statistics, supercore inflation (core inflation excluding used cars and housing) increased by 0.31% month-on-month this month, rising to its highest level since February 2026. 7. Overall, the CPI report appeared slightly higher than expected, but if the highly volatile telephone services are excluded, the month-on-month central value is still roughly around 0.2%. A report that appears hawkish on the surface but is dovish underneath will trigger a rate hike in advance—a September rate hike is a foregone conclusion, but it's not enough to significantly increase the final rate hike date or the number of rate hikes—the 1-year Treasury yield only rose by 2 basis points, and the 10-year Treasury yield even fell from its high (a rate hike today actually means less will be needed in the future). This is more like a rate hike pre-emptive trade than a rate hike (magnitude) trade, which has instead boosted precious metals, stocks, and crypto assets. 8. The market seems quite expectant of a so-called "dovish rate hike"—a rate hike that comes as expected, a consistent statement from Kevin Warsh, and a less hawkish dot plot. If this happens, it would be the best outcome the September FOMC can deliver. However, considering the significant rise in oil prices over the past two weeks, it remains questionable whether the voting members' actual expectations will remain restrained as the market desires—after all, the sharp hawkish turn in the June dot plot stemmed from concerns about high oil prices.