Interest Rate Weekly · 2026-08-03 to 2026-08-09

One-sentence summary: The main theme over the past week was U.S. rates continuing to tick up at high levels with the curve maintaining a positive slope, combined with US-Japan FX and intervention rumors, disputes over U.S. jobs and inflation data, and China’s July CPI rebound — together pulling global rate trading back onto the track of “supply pressure + policy uncertainty + cross-market linkage.”

I. Review of Last Week

The U.S. produced a high density of information this week, but the impact on the yield curve was not complicated: yields across maturities continued to rise, though the slope of the rise was slightly slower than earlier. Curve snapshots show US 2Y at 4.25%, up 2bp over 5 days, up 9bp over 20 days, up 30bp over 60 days; US 5Y at 4.40%, up 2bp over 5 days, up 13bp over 20 days, up 33bp over 60 days; US 10Y at 4.69%, up 1bp over 5 days, up 15bp over 20 days, up 27bp over 60 days; US 30Y at 5.22%, up 1bp over 5 days, up 17bp over 20 days, up 24bp over 60 days. More importantly is the level itself: from the 1-year z-value, 2Y is 1.96, 5Y is 2.12, 10Y is 2.25, 30Y is 2.45, meaning the entire U.S. curve remains at relatively high percentiles over the past year, with the long end’s “expensive rate” status particularly pronounced.

On the news front, U.S. signals mainly concentrated in three groups. The first group is Fed independence and policy jockeying. On 8/6 there was “Trump Calls Fed Chief”, and on 8/7 and 8/9 there were consecutive items “Trump reportedly renews bid to oust Fed Gov. Lisa Cook” and “Trump Revives Effort to Fire Fed Governor”; the same day also saw “The Fed Status Quo vs. Kevin Warsh”. These reports at least indicate the market is trading not only macro data but also the policy framework and personnel uncertainty. For rates markets, such news may not provide a single-day directional signal, but will raise term premium and policy uncertainty premium, especially when long-end yields are already at high levels.

The second group concerns data itself and data credibility. On 8/7 there was “Will the negative jobs report hold off a September rate hike?”, indicating a weak jobs report sparked debate over a September rate hike; on the same day there was “U.S. NEWS Change in Inflation Data Questioned”, showing the inflation data’s definitions or quality were being questioned by the market. Taken together, the U.S. this week was not simply “weaker growth, lower rates,” but rather “some growth data weakened, yet policy and inflation signals remain unclear.” This helps explain why yields did not fall materially due to the negative jobs signal, but merely slowed their pace of ascent.

The third group involves cross-market linkage from supply and FX intervention. On 8/9 there was “US Government Sold $638 Billion of Just T-Bills this Week, 10-Year Treasury Yield Dips after Big Kahuna Yen Intervention”; from 8/3 to 8/5 there were consecutive items “Why the US-Japan Joint Intervention to Prop Up the Yen? Fear of Treasury Yields Blowing Out if Japan Becomes a Forced Seller”, “Can the U.S. Treasury Save the Yen?”, “Why Bessent Leans on Fed to Help Japan”, “Yen intervention Trump's Treasury czar forges new era of US activism with increasing involvement in the financial sector”. These stories point to a core logic: if Japanese asset allocation is forced to adjust, U.S. Treasuries could face additional selling pressure; conversely, stabilizing the yen is also seen as part of stabilizing the U.S. long end. Looking at the outcomes, the 10-year yield “dips” after intervention, but on a weekly and monthly basis U.S. Treasuries have not abandoned the upward trend, implying intervention affects pace rather than rewriting the medium-term rate center.

Canada was much quieter compared with the U.S., but the curve showed its own characteristics. CA 2Y at 2.92%, up 7bp over 5 days, up 7bp over 20 days, down 1bp over 60 days; CA 5Y at 3.22%, up 3bp over 5 days, up 4bp over 20 days, up 3bp over 60 days; CA 10Y at 3.61%, up 3bp over 5 days, up 5bp over 20 days, up 7bp over 60 days; CA 30Y at 4.02%, up 3bp over 5 days, up 7bp over 20 days, up 12bp over 60 days. Canadian yields also rose over the past week, but the magnitude was clearly weaker than the U.S., especially over the 60-day horizon: U.S. 2Y to 30Y mostly rose 24–33bp, while Canada was roughly between -1 and +12bp. In curve shape, Canada 2s10s is +69bp, noticeably steeper than the U.S. +44bp, reflecting a lower short-end rate level in Canada, more front-end compression, with the long end relatively following global pricing. This “lower short-end, steeper curve” structure did not converge this week.

Other global clues were mainly China and Europe. For China, on 8/9 the statistics bureau released “CPI year-on-year rose 0.5% in July”, while on 8/5 the market had originally assessed “July CPI and PPI year-on-year growth will slightly fall back.” The actual release at least indicates the market’s prior expectation for a Chinese inflation pullback was not fully realized. Looking at a single CPI item cannot directly be extrapolated to China’s interest rate path, but for global bond markets it means deflation worries on the demand and commodity chain have not deepened further, slightly easing the pricing pressure for “nominal growth being too low” at the long end globally. In Europe, on 8/7 the ECB published consolidated banking data through end-March 2026; the information was structural and had limited direct impact on global sovereign rate trading during the week, but disclosures on the stability of the European financial system still form part of the background for global risk appetite.

On credit, the data are only suitable for description and not for extended inference. US HY OAS is 2.71%, 20-day widened 1bp; US IG OAS is 0.78%, 20-day widened 2bp. The changes are very small, indicating this round of U.S. long-end rise, at least so far, has not been accompanied by a notable re-evaluation of credit risk, but rather has been dominated by risk-free rate and term premium.

II. Views on Future Rates

First clarify methodological boundaries. Directional judgments for the next three weeks can only be built on momentum signals already labeled as [Proven]; any series labeled as “≈ random walk” will not have directional predictions and will only be discussed in terms of static carry and curve position. The meaning of carry must also be separated: it is merely the expected return “if yields do not move,” a tailwind, not a direction prediction; if yields continue to rise, price losses can completely offset carry.

For the U.S. short- to mid-end, we can currently give a bias toward upward direction. Directional forecasts for US 2Y, 5Y, and 10Y are all labeled [Proven] with dir=up. The 2Y has a hit rate of 56.2%, which offers statistical gain relative to a base rate of 50.4%; the 5Y hit rate is 53.9%, above a 50.8% base; the 10Y hit rate is 52.6%, above a 50.6% base. Combined with the spot curve’s steady lift over the past 60 days, this means the probability that U.S. front- and mid-section yields continue to rise over the next three weeks remains biased to the upside. The emphasis here is not that weekly moves are only 1–2bp, but that both 20-day and 60-day windows are steadily moving up, indicating the trend has not been broken.

From a trading implication standpoint, US 2Y and 5Y are the segments where direction and static yield resonate most. US 2Y annualized carry is 4.63, corr=0.326, long-win rate 68%; US 5Y annualized carry is 4.65, corr=0.178, long-win rate 59%; US 10Y annualized carry is 5.27, corr=0.112, long-win rate 56%. If looking at carry alone, the 10Y is higher; if incorporating directional risk, 2Y and 5Y have stronger defensive characteristics because short-end carry is not low and duration is shorter, making them less price-sensitive to further yield increases. Conversely, even if the 10Y has higher static carry, as long as direction remains “yields biased higher,” its price volatility can more easily swallow coupon and roll-down income.

U.S. long end requires more restraint. US 30Y currently yields 5.22%, 1-year z-value 2.45, at a very high level, with annualized carry reaching 6.015, the highest in the table; but the direction forecast explicitly notes “≈ random walk.” Although dir is marked up, the hit rate of 52.7% versus a base of 51.3% does not form an effective advantage, so no directional judgment can be made on that basis. The only conclusion is: U.S. 30-year’s static yield is very high and carry is attractive if yields remain unchanged; but the direction over the next three weeks is unpredictable and not judged, and with large duration exposure in the long end, if supply pressure, policy uncertainty, or overseas official asset reallocation continue to push yields up, price losses could entirely offset carry. High carry at the long end is not by itself a buy signal.

Canada is simpler: the entire curve’s direction signals are all labeled ≈ random walk, therefore no directional judgment is made for the next three weeks. Canada 2Y is currently 2.92%, 5Y is 3.22%, 10Y is 3.61%, 30Y is 4.02%, absolute yields are materially lower than the U.S., and corresponding carry is generally lower than U.S. Treasuries, but not negligible. The carry ordering in the data shows CA 30Y annualized carry is 4.98, corr=0.111, long-win rate 54%; CA 10Y is 4.39, corr=0.141, long-win rate 56%. This indicates that under the assumption of unchanged yields, Canada’s mid- to long-end has some static yield support, but since direction is unpredictable, it should be understood as income to hold, not as an extrapolation that rates will fall.

Cross-country divergence remains the most worth-tracking point for the next three weeks. The most notable differences between the U.S. and Canada are twofold. The first is level differences. U.S. 2Y and CA 2Y differ by 133bp, U.S. 10Y and CA 10Y differ by 108bp, U.S. 30Y and CA 30Y differ by 120bp. This shows the entire U.S. curve is more in a “higher-rate” state. The second is trend differences. Over the past 60 days, U.S. maturities mostly rose 24–33bp, whereas Canada was only -1 to +12bp, indicating the recent global long-end rise has been driven more by U.S.-domestic factors, including supply, policy jockeying, and term premium, rather than a synchronous North American macro reheat. If this divergence continues, then on a cross-country relative value basis, the U.S.’s high-rate status will remain a core feature.

In curve shape, both the U.S. and Canada maintain positive slopes but with different natures. US 2s10s is +44bp, indicating front and mid both rose together and the curve has not re-inverted; CA 2s10s is +69bp, the curve is steeper, reflecting a lower Canadian front end. For the next three weeks, this usually means the U.S. policy path remains the main axis for rate trading, while Canada more often plays a following but not fully synchronized role. If the U.S. short end continues to rise, 2s10s may not necessarily flatten significantly, because the 10Y itself also has a proven upward momentum; but for the 30Y, since direction is unpredictable, changes in 10s30s are not suitable for clear judgment at present.

Global background will continue to influence the U.S. long end more than Canada. This week’s intensive news around yen intervention and potential U.S.-Japan joint action has put the question of “whether overseas official institutions will adjust U.S. Treasury holdings” back on the table. Even if single-week data show the 10-year yield fell after intervention, the 60-day upward trend has not been broken. In other words, FX intervention can ease the pace but does not necessarily absorb U.S. supply and policy uncertainty. For medium-term positions, this point matters more than single-day volatility.

China’s July CPI year-on-year 0.5% release, placed in the global rates framework, looks more like a slight correction to the narrative of “global disinflation dragging down long-end nominal rates” rather than a new dominant variable. European banking data plays a similar role: it provides background information but does not rewrite the week’s North American rate regime led by the U.S.

Overall, the next three weeks can be summarized into three judgments. U.S. 2Y, 5Y, 10Y yield directions remain biased upward — this is the only main judgment supported by [Proven] signals; U.S. 30Y direction is unpredictable, no judgment, acknowledging its carry is high but also most vulnerable to upside risk; Canada’s entire curve direction is unpredictable and is better viewed from static carry and relative value angles rather than expressed as a one-sided directional view.

III. Risk Warnings

The biggest risk is mistaking carry for direction. Currently US 30Y, 10Y and even CA 30Y all provide non-trivial annualized carry, but if yields continue to rise, especially at the long end, price drawdowns will quickly eat up coupon and roll returns. High carry can only improve the holding experience, it cannot replace a directional judgment.

The second risk is treating the “single-day shock” of news events as a trend reversal. Whether it is a weak U.S. jobs report, contested inflation data, political messages around Fed personnel and independence, or U.S.-Japan yen intervention, these events can bring intraday and single-week volatility, but from currently validated signals the U.S. front- and mid-end upward momentum remains, and one-off news should not be used lightly to infer a trend turning point.

The third risk is uncertainty in cross-market transmission. If yen stability relies on sustained intervention, U.S. long-end may temporarily receive support; but if overseas funding and allocation constraints intensify, term premium on U.S. Treasuries could rise again. These forces currently influence the long end more, and the long end is precisely the segment with unpredictable direction.

The fourth risk is that credit spreads remain calm. HY and IG OAS have only slightly widened in the past 20 days, indicating the market has not yet used credit risk to confirm the macro consequences of rate rises. If credit spreads later catch up, the linkage between rates and risk assets could change, and the current static understanding of curves and carry would need recalibration.

This report is for research reference only and is not investment advice.