What's the take on today's CPI? Honestly, it's the same old story. The overall CPI (headline) at 3.6% is already priced in... The key is still the decimal point of core CPI (core). The core is expected to round to 0.2%, but the real question is whether it's a "cool" 0.2% like 0.18% or 0.20%, or a "hot" 0.2% like 0.23% or 0.24%. --------------- Current predictions suggest this month will see a hot overall CPI but a cooling core CPI. High oil prices in September are expected to push the overall CPI to a monthly rate of 0.6% and an annual rate of 3.6%. However, institutions predict the core CPI will drop from August's 0.29% to around 0.2%. The main reason? The one-time spike in August's mobile phone bills won't happen again. Adding to that, September's weaker-than-expected non-farm payrolls data means a rate hike in October is basically off the table, with the market pricing in just an 18.5% chance. The real debate is over a December rate hike, currently at a 74% probability. Tonight's suspense lies in whether the core monthly rate rounds to 0.2% or 0.3%. If it's 0.2%, is it a "cool" 0.2% or a "hot" 0.2%? --------------- Looking at recent speeches from several Fed voting members, an October hike is highly unlikely, but rate hike expectations for the future aren't over yet. Core members like Jefferson, Williams, and Bowman have said there's no rush and that they can keep monitoring the data. Waller wants to hike but doesn't think it needs to be consecutive. Logan and Hammack remain hawkish and want to keep tightening. Williams also mentioned that one more hike this year would be appropriate. --------------- In short, the main focus of tonight's CPI is on the second decimal place of the core monthly rate and whether inflation has spread from gasoline to airfare, rent, and services. Market expectations are centered around 0.22%. So: Below 0.20% = dovish 0.20%–0.22% = in line with expectations 0.23%–0.24% = neutral to slightly hawkish 0.25%+ (which would round to 0.3% in the news) = hawkish --------------- Right now, $83k is near the edge of the range... There's buffer room both up and down. So as long as the data isn't too extreme, the range should hold.