AUDUSD TECHS: Support Remains Intact For Now

• Treasuries see-sawed off early Wednesday lows to steady/mixed in late trade, focus on Thursday's heavy data (weekly claims, CPI, regional Fed and Tsy TICS) and Riksbank, Norges, Bank of England and ECB policy announcements. Fed speakers are absent tomorrow.
• We won't get aggregate CPI levels for October tomorrow, but as part of the November report's unprecedented format, the BLS will include data for some but not all subcomponents for October.
• Currently, TYH6 trades -1 at 112-15.5, yielding 4.1509% (+.0059). Curves mildly steeper, 2s10s +.995 at 66.588, still well off yesterday's 69.086% high.
• The short-term technical condition in Treasuries is bearish and near-term resistance points remain. A continuation lower would refocus attention on key support at 111-29, the Dec 10 low. Clearance of this level would confirm a resumption of the bear leg and open 111-19, a Fibonacci projection.
• On the upside, a clear breach of 112-23, the Dec 12 high would instead strengthen a short-term bull cycle.
• Gov Waller indicates in Q&A earlier that he's forecasting GDP growth of 1.6% this year and 2.5% in 2026 (vs FOMC medians 1.7% / 2.3%), and that supply-side improvements mean that stronger growth will not translate into stronger inflation (an argument advanced by the Trump administration as well as some of Waller's FOMC colleagues such as Gov Miran).

The October data could include several of the major categories that we include in our Core CPI consensus tables: used cars and trucks and new vehicles; airfares; lodging away from home; and various communications and medical prices (for headline CPI, gasoline data could be published).

• At least one analyst (Nomura) speculates that the BLS could estimate rents and OER as well as rent data are collected through the Housing Survey and not the Commodities and Services Price survey.
• Below is MNI's collation of sell-side consensus on individual categories. It expresses the average M/M inflation rate over October / November, either from specific month-by-month estimates, or by halving the 2-month change.

We won't get aggregate CPI levels for October, but as part of the November report's unprecedented format, the BLS will include data for some but not all subcomponents for October.

• The vast majority of price data collection are collected by BLS employees in person or by phone, which obviously wasn't possible in October due to the federal government shutdown. However over 20% of the basket is taken from a variety of sources that don't entirely require personal collection, relying instead on online prices and private data providers.
• The BLS hasn't announced which components this will include for October. The BLS previously noted that it "expects the number of publishable indexes to be small".
• We don't know how the BLS will present such data either, as while it could simply publish the usual tables with % changes for both October and November (they advised "news release and database update will not include 1-month percent changes for November 2025 where the October 2025 data are missing"), this could clash with most items which would only have the November data (potentially with just the indices published and the 2-month changes noted).
• This could make it difficult to quickly interpret the data; the one sure thing is that we will get a % Y/Y reading for the main aggregates (headline and core CPI) because a comparison to November 2024 indices will be possible, though we won't get % M/M readings as October aggregates won't be compiled.