Inflation is still accelerating: producer prices 6.3%, consumer prices 3.7%

Interest Rate Weekly · 2026-10-11 to 2026-10-18

U.S. September CPI rose to 3.7%, rekindling Fed rate-hike expectations; U.S. yields, after retreating within the week, remain at one-year highs; the Canadian curve, however, has diverged from the U.S.

1. Review of Last Week

U.S. thread. The main story this week was the September inflation data. Multiple media outlets reported CPI year-on-year rose to 3.7%, above expectations, with rising gasoline and diesel prices as the main drivers. Marketplace’s headline summed it up as "blame energy prices, but mostly so," and other commentary highlighted that core CPI and core services CPI also increased, suggesting inflation may be spreading into the broader economy. Producer-side pressure was larger: there were reports that PPI year-on-year was 6.3%, the highest since 2022. The Fed’s Beige Book at the same time described moderate economic growth and rising price pressures.

On policy, Fed Chair Warsh (Kevin Warsh) emphasized price stability in remarks on 10/15 to 10/16, committing to tamp down inflation while giving no signal about the next rate hike and suggesting AI could bring shocks to the economy. Outside the Philadelphia Fed, Fed official Collins was reported by Reuters as still considering another rate hike before the end of 2026. Commentaries ran titles saying this inflation surprise disrupted the Fed’s prior wording of "patience." On the market side, early-week coverage included stories like "investors tough it out against U.S. Treasury selloff" and "why U.S. stocks and Treasury yields rose together," and U.S. equities dropped on the day CPI was released. CNBC reported on 10/15 that U.S. Treasury yields fell sharply after the hot CPI; another headline noted that long-end yields faced a dual disturbance from high inflation and the AI boom.

Canada thread. Available Canadian news this week was limited. BMO’s view is that Canadian bond yields and lending rates will rise with U.S. inflation; another headline warned the Canadian dollar faces headwinds in Q4 due to weak domestic growth and the Fed rate-hike bets. Taken together, these items point to externally imported pressure combined with weak domestic demand.

Global thread. In the U.K., GDP grew 0.1% amid heightened rate-hike concerns; the IMF warned that the U.K. faces inflation risks ahead of the government’s first budget and expects households to face more pronounced price increases, while U.K. GDP growth exceeded expectations. In Europe, there were reports that the possibility of ECB hikes is again weighing on the European real estate recovery. The IMF’s forecast for the U.S. is higher inflation in 2026 and a decline in 2027; the U.S. Treasury rebutted the IMF’s debt warnings and forecasting methods. On the energy front, there were reports about U.S. sanctions or blockades related to Iran, but this week’s materials did not elaborate on specific price transmission, so no inference is made here.

Curve data align broadly with the above narratives, with some directional differences. U.S. 2-year, 5-year, 10-year, 30-year yields were 4.74%, 4.89%, 5.09%, 5.43%, with intraweek changes of -1bp, -10bp, -13bp, -17bp; the longer the tenor, the larger the retreat, and the curve flattened during the week. Over 20 days, U.S. tenors still rose 7 to 15bp; over 60 days they rose 28 to 48bp, with one-year z-scores in the range 1.85 to 2.06, at the high end of the past year. Canadian 2-year, 5-year, 10-year, 30-year yields were 3.20%, 3.53%, 3.83%, 4.17%, down 3 to 13bp on the week and down 12 to 15bp over 20 days, but still up 21 to 35bp over 60 days, with z-scores 1.54 to 1.77, slightly lower than the U.S. On credit, U.S. high-yield OAS was 3.23%, widening 53bp over 20 days; investment-grade OAS was 0.84%, widening 6bp over 20 days. These two series have only about 3.1 years of history, so they are described only and no conclusion is drawn.

2. Views on Future Rates

Method boundary: the system’s directional judgments rely solely on quantitative signals, and directions are assigned only for tenors that passed historical tests; other tenors are treated as random walk, with direction unpredictable and therefore not judged — only static factors such as carry are discussed. Positioning horizon is approximately three weeks.

U.S. short- to mid-end (2-year, 5-year, 10-year). All three tenors are marked as proven, with directions biased toward higher yields. Historical hit rates are 2-year 56.2% (base rate 50.5%), 5-year 54.0% (base rate 50.6%), 10-year 52.7% (base rate 50.4%), with samples each exceeding ten thousand observations. The edge shrinks with tenor: the 2-year is about 5.7 percentage points above the base rate, the 5-year about 3.4 points, the 10-year about 2.3 points. The upward bias aligns with this week’s macro narrative: inflation surprised on the upside, Fed officials have not ruled out another hike, and 60-day momentum is upward. Hit rates are only a few percentage points above base, indicating a tendency rather than certainty; intraweek yield retreats already occurred, showing paths will have reversals.

U.S. 30-year. The signal is labeled ≈ random walk, although the historical hit rate 52.8% is slightly above the base 51.1%, the gap is insufficient to support a directional judgment. The 30-year yield direction is unpredictable and not judged. This tenor has the highest carry on the leaderboard (annualized 5.94%), but its reliability is very low (corr 0.077, long-win rate 53%, marked as marginal), so it can only be treated as a static tailwind.

Canadian tenors. 2-year, 5-year, 10-year, 30-year are all ≈ random walk, with historical hit rates comparable to or slightly below the base (10-year 49.8% vs 52.1%, 30-year 51.5% vs 52.4%). Canadian yield directions are unpredictable and not judged. BMO believes Canadian yields will follow U.S. higher; that is an institutional view and differs from this system’s test results, and therefore does not form this report’s directional judgment. On the static side, Canadian 30-year carry annualized 4.94% and 10-year 4.43% are both marked as proven, but correlations are only 0.109 and 0.139 with win rates 54% and 56%, indicating carry’s explanatory power for actual returns is limited.

Carry ranking and usage boundary. Annualized carry plus roll rank as: U.S. 30-year 5.94%, U.S. 10-year 5.49%, U.S. 2-year 5.33%, U.S. 5-year 5.14%, Canada 30-year 4.94%, Canada 10-year 4.43%. Among these, the U.S. 2-year has the highest reliability (corr 0.326, long-win rate 68%), followed by the U.S. 5-year (corr 0.178, win rate 59%). Carry is the expected return if yields do not move, a tailwind, and does not constitute a directional forecast. Converted to a 3-week horizon, an annualized carry just above 5% equates to roughly 0.3%. Using a crude duration of 7 years, a 10-year yield rise of about 5bp would produce enough price loss to erase three weeks’ carry; the 2-year’s duration is short, and an approximately 15bp rise would offset it. This system’s directional judgments for U.S. 2-year, 5-year, 10-year are biased toward higher yields, and in this environment price losses are likely to outweigh carry, so a high carry ranking alone should not be taken as a buy signal.

Cross-country divergence. The U.S.-Canada yield spread has clearly widened over the past 20 days: U.S. yields rose 7 to 15bp while Canada fell 12 to 15bp, widening the U.S.-Canada spreads by about 22bp, 23bp, 27bp, 26bp for the 2-year, 5-year, 10-year, 30-year respectively, with current levels about 154bp, 136bp, 126bp, 126bp. The divergence matches the news: hot U.S. inflation and higher rate-hike expectations, while Canada is described as having weak domestic growth. This system has directional signals for U.S. short- and mid-ends but not for Canadian tenors; whether spreads will widen further or converge next is unpredictable, so only the current state is described.

Curve shape. U.S. 2s10s is +35bp, Canada is +63bp, so the Canadian curve is steeper. This week U.S. 2s10s narrowed by about 12bp and Canada by about 8bp; over 20 days, the U.S. curve steepened about 8bp and Canada steepened about 3bp. Curve shape depends on relative moves across tenors, and because this system has no directional signals for the 30-year and for Canadian tenors, it makes no judgment on steepening or flattening — the figures above merely describe current positions.

3. Risk Warnings

First, uncertainty in inflation and policy. The Fed Chair gave no signal about the next hike and officials’ statements are not uniform; if subsequent data again surprise on the upside, yields could move up more steeply; if data cool, the intraweek retreat may continue. Momentum signals have limited edge and single-instance misjudgment probability is not small.

Second, the trade-off between carry and price. U.S. yields are at one-year highs with z-scores around 2, and that position alone does not guarantee a retreat. In a rising-yield environment, price losses can overwhelm carry, especially for long-duration instruments.

Third, credit and equity linkage. U.S. high-yield OAS widened 53bp over the past 20 days, while reports noted cracks in the AI boom and U.S. stocks fell. Credit data have only about 3.1 years of history and limited statistical significance; if risk appetite weakens further, the correlation between rates and credit could change.

Fourth, external and geopolitical factors. Energy prices are a main driver of this round of inflation; developments related to Iran, the IMF’s warnings on U.S. debt, and rate-hike concerns in the U.K. and Europe could transmit to North American rates via global channels, but this week’s materials are insufficient to quantify those channels. Canada lacks domestic data, so judgments rely mainly on external information and the information set is thin.

This report is for research purposes only, for reference only, and does not constitute investment advice.