Interest Rate Weekly · 2026-08-10 to 2026-08-16

One-sentence overview: Last week’s main theme was “U.S. short end pulled back on softer inflation prints while the long end was held up by supply and term premium”; Canada showed generally stable-to-slightly-higher yields, and the U.K. and Australia continued to reflect the cross-country divergence of “growth under pressure but inflation still sticky.”

I. Last Week Review

U.S. clues were the clearest: first moderate inflation prints, then supply and long-bond pressure pulled the narrative back toward “steep front, bear back” (front easing, back stiffening). On 8/12, multiple outlets reported July U.S. CPI as “inflation cooling slightly” and “energy price declines weighing on the overall reading,” and Bloomberg on 8/15 listed "US CPI Softens" as an important change in the global economic picture. On 8/13 PPI coverage showed two-sided information: on one hand “Wholesale Inflation Rolls Over” and “further supports Fed doves,” while on the other hand there were substantial upward revisions to the prior-month services PPI and core PPI. Taken together, this bundle implies U.S. price pressure is not unambiguously easing but is closer to “current prints a bit soft, components and revisions are messy.” Fed officials’ comments echoed this: Reuters on 8/11 quoted Goolsbee saying “inflation is the biggest problem,” and Richmond Fed’s Barkin on 8/13 discussed a “mysterious U.S. economy,” reflecting that growth, employment, and inflation have not formed a simple linear relationship.

Market pricing gave a more direct feedback along the curve. Over the past 5 trading days, U.S. 2-, 5-, 10-, and 30-year yields fell by 10bp, 8bp, 6bp, and 1bp respectively. The front end’s decline was noticeably larger than the long end’s, indicating the CPI/PPI combination impacted policy-sensitive tenors faster while the long end did not follow in lockstep. Looking out 20 days, U.S. 2-year is only down 1bp, while 5-, 10-, and 30-year are up 4bp, 6bp, and 12bp respectively; over 60 days, 2s–30s are mostly still higher than before. In other words, last week’s pullback looked more like a front-end correction from previously elevated rate levels rather than a trend reversal across the whole curve. U.S. 2s10s is currently +48bp, the curve remains positively sloped, and the pattern of near-term easing with a resilient long end makes the “steepening” feature more pronounced.

The long end’s reluctance to follow lower is largely explained by supply, an unavoidable variable. News on 8/15 showed the U.S. Treasury sold $742 billion of bonds in a week, 30-year Treasury auction yields hit the highest since 2001, and 10-year yields hit the highest since 2007. The same day’s “Long Bond, Big News” also highlighted that market focus has concentrated on ultra-long financing stress and a re-pricing of term premium. Combined with the Fed’s 8/13 decision to reduce Reserve Management Purchases to zero from 8/14, a diminished marginal official bid arriving at the same time as large issuance reinforces the market impression that “good data may not bring obvious downward pressure” on the long end. This aligns with curve metrics: U.S. 30-year fell only about 1bp over the past 5 days but its 1-year z-value is as high as 2.20, making it relatively the most expensive and highest-yielding segment among tenors, which indicates long-end pressure has not truly eased.

Canadian clues were comparatively muted and formed a clear divergence from the U.S. With no major new domestic macro shocks in the news set, market behavior was more that of following global rate moves but at a slower pace than the U.S. Over the past 5 days, Canadian 2-year was unchanged, and 5-, 10-, and 30-year rose 1bp, 1bp, and 2bp respectively; over 20 days, all tenors are up 9–10bp, showing Canada did not experience the same front-end rapid pullback triggered by CPI as in the U.S. Current Canadian 2-, 10-, and 30-year yields are 2.92%, 3.62%, and 4.04% respectively, 2s10s is +70bp, and the curve slope is steeper than the U.S. This shape implies Canada’s policy-sensitive short end has not rushed to price in easing, while the belly and long end are modestly higher, putting Canada closer to a “stable-to-slightly-bearish” state overall.

Global clues concentrated in the U.K. and Australia, with the shared theme “growth under pressure but inflation constraints remain.” In the U.K., Bloomberg on 8/13 noted an unexpected improvement in the economy and on 8/15 listed “UK Economy Expands” alongside “US CPI Softens”; but the Guardian on 8/16 warned a new round of cost-of-living pressure could push inflation higher due to rising energy bills, and This is Money on 8/12 noted Iran-related conflict risk could weigh on U.K. growth. The U.K. combination is a typical coexistence of growth and inflationary disturbances: near-term data are not poor, but energy and geopolitical factors raise inflation tail risk again. In Australia, the RBA on 8/11 held rates unchanged but multiple reports emphasized inflation remains stubborn and hikes remain possible if necessary; 8/13 coverage that “high rates are breaking the Australian economy” reflects that policy bias remains tight while the growth cost is rising. Cross-country, the U.S. looks like “current inflation easing + long-bond supply shock,” the U.K. is “growth OK + inflation re-acceleration risk,” and Australia is “high rates squeezing growth + central bank reluctant to loosen.” This helps explain why global rate markets lacked a uniform direction last week.

On credit, there was no clear deterioration last week. U.S. HY OAS 20-day is unchanged at 2.71%, IG OAS 20-day widened only 1bp to 0.79%. Because this credit history sample is only 3.1 years, one can only make static descriptions, but at least by these two measures the long-end rate pressure has not yet translated into a synchronous widening of credit risk.

II. Views on Future Interest Rates

First, clarify methodological boundaries. Judgments for the next 3 weeks must adhere to validated signals: direction signals for U.S. 2-, 5-, and 10-year have statistical backing and can provide directional tilt; U.S. 30-year and Canadian tenors are, within the given framework, close to random walk, so their directions are unpredictable and not judged — only carry and static curve structure when yields do not move can be discussed. Carry is not a direction forecast; it only represents expected returns if yields remain unchanged; if yields rise, price losses can easily overwhelm carry.

Start with the U.S. short end. U.S. 2-year is labeled [Proven] but direction is mixed, with a hit rate of 56.2% versus a base rate of 50.4% — an improvement but not supportive of a strong one-sided view. Given last week’s softer CPI, harder PPI revisions, and Fed officials’ continued vigilance on inflation, a 3-week view of the U.S. 2-year is better described as “oscillatory and tradeable” rather than a trend bet. On static returns, US_2Y carry is 4.51, with correlation 0.326 and long-win rate 68%, making it the most robust carry segment among the given assets. That means if yields sideways, the 2-year has clear coupon and roll-down cushioning; however if markets refocus on inflation revisions, Fed commentary, or fiscal supply spillovers and the 2-year ticks up, carry will not fully offset price volatility.

U.S. 5- and 10-year direction signals are clearer: both are [Proven] and dir=up. The 5-year hit rate is 53.9% versus a base rate of 50.8%, and the 10-year hit rate is 52.6% versus a base rate of 50.6% — not huge edges but sufficient to support a baseline that medium-term rates still lean toward higher. This conclusion is consistent with the news: on the one hand, softer current inflation readings lowered the front end; on the other hand, PPI revisions, Fed vigilance on inflation, large Treasury supply, and the Fed’s RMP going to zero all provide upward stickiness for the belly and long segments. Data-wise, the U.S. 5- and 10-year 1-year z-values are 1.68 and 1.80 respectively, still at high percentiles over the past year, and markets demand higher yields to absorb supply and uncertainty. For a 3-week holding period, the U.S. 5–10 year segment looks like “carry exists, but direction still leans toward higher yields.” US_5Y carry is 4.603, US_10Y carry is 5.25; the latter offers higher static return but will suffer larger duration losses if yields continue to move up per the proven signals.

U.S. 30-year cannot be given a directional call. The signal clearly indicates ≈ random walk; although dir=up, a hit rate of 52.7% is only slightly above a base rate of 51.3% and lacks usable predictive power, so the long bond’s direction is unpredictable and no directional call is made. On the static side: US_30Y carry is highest at 6.08, but its reliability is only 0.077 and the long-win rate is 53%, making it marginally effective. Combined with last week’s 30-year auction yields reaching highs since 2001 and supply pressure concentrated on the ultra-long end, the 30-year’s coupon and roll look attractive only as a tailwind, not as a buy signal. If yields climb further, long-duration price drops will quickly erase carry — a risk especially relevant in a 3-week holding horizon.

For Canada, directions are essentially non-predictable. CA 2Y, 5Y, 10Y, and 30Y are all close to random walk, so directions over the next 3 weeks are unpredictable and no directional calls are made. Two points about Canada can be discussed. First, the curve is steeper than the U.S., with 2s10s at +70bp, indicating markets demand higher term compensation and the short end has not quickly priced in easing. Second, static carry is not bad: CA_10Y carry is 4.4 and CA_30Y is 5.012, with long-win rates in the sample of 56% and 54% respectively, which can be interpreted as “if yields sideway, Canada’s belly and long end offer some hold-return buffer.” But this is still not a directional judgment; under a random-walk framework, the safest conclusion is that Canadian bonds are better viewed from curve carry and coupon perspectives rather than overclaiming trend success probability.

Cross-country divergence is likely to persist in the short term. The U.S. core contradiction is “front end supported by softer current inflation while the long end is constrained by supply and term premium”; Canada lacks an equally strong domestic catalyst and behaves more like a mild follower; the U.K. faces renewed inflation risk from energy, and Australia still operates in the framework of “growth under pressure but inflation stubborn.” On the curve, U.S. steepening looks event-driven while Canadian steepening is more static. If over the next 3 weeks the U.S. continues to show “soft data but weak long-bond absorption,” maintaining or further steepening 2s10s would not be surprising.

From the holding experience perspective, the easiest misread is “high carry = good buy.” The data actually warn the opposite: U.S. 30-year carry is highest but direction is unpredictable, and last week the ultra-long end was the most exposed to auction and supply stress; U.S. 10-year carry is second highest but the direction signal is skewed toward higher yields, i.e., price bearish. For medium-term positions, carry is better used as a cushion in a sideways market, not a substitute for directional judgment. The only segments that provide directional tilt remain the proven U.S. 5- and 10-year tenors, and their tilt is toward higher yields; the U.S. 2-year is closer to oscillation.

III. Risk Warnings

The largest short-term risk remains U.S. long-end supply. If subsequent Treasury issuance, auction tails, or take-up conditions remain weak, term premium on the 30-year could rise further and transmit to the 10-year, offsetting the downward pressure on the belly from softer inflation prints. Last week’s 30- and 10-year auction yields already reached highs since 2001 and 2007 respectively, and such signals often affect market sentiment for more than a week.

The second risk is “surface softness in inflation with messy components.” Last week CPI slowed due to energy drag, but PPI prior-month revisions were strong and Fed officials did not send clear easing signals. If markets refocus on services, core components, or the revised price resilience, front-end declines may be reversed and U.S. 2-year volatility could increase.

The third risk stems from global energy and geopolitical variables. U.K. reporting has already linked rising energy bills to re-accelerating inflation, and Iran-related conflict has been mentioned as a drag on U.K. growth. If oil and gas prices become dominant again, the developed-market mix of “weaker growth but persistent inflation” could be reinforced, and global long-end rates may not benefit.

The fourth risk is cross-country policy divergence. The RBA is on hold but retains the option to hike, the U.K. is toggling between better-than-expected growth and rising inflation risks, and the U.S. sits with simultaneous inflation cooling and long-bond supply pressure. Continued moves in cross-country spreads could, via FX and relative-value flows, affect U.S./Canada curve behavior and render single-market narratives ineffective.

This report is for research purposes and is not investment advice.