![\"\"](\"https:\/\/i0.wp.com\/karolpelc.com\/InvestorSnippets\/wp-content\/uploads\/2026\/08\/ust_yields_intraday_2026-08-08_W32.png?fit=1024%2C576&ssl=1\")
<\/figure>\n\n\n\nThe week’s defining move was the sharp repricing of near-term Fed expectations following Friday’s weaker employment report, discussed separately in the Macro update. September hike probability fell from 71.9% on 2026-07-31 to 44.0%<\/strong> on 2026-08-07, while the implied September overnight rate declined 7.1 bps to 3.741%. December pricing fell 8.8 bps to 3.914%, taking cumulative tightening priced through year-end from roughly 1.5 hikes to 1.1 hikes<\/strong>. SOFR options showed investors unwinding hawkish downside protection on Friday, while 42k Jun-2027 call spreads were bought, indicating some positioning for a softer policy path further out. The market nevertheless continues to price approximately one additional hike by year-end rather than a transition toward easing.<\/p>\n\n\n\n Inflation markets moved less dramatically. The 5Y breakeven fell 5.6 bps to 2.23%, and the 10Y declined 3.1 bps to 2.25%, while the 5Y5Y inflation forward was essentially unchanged at 2.421%. The stability of longer-run inflation compensation despite the Treasury rally suggests that much of the long-end volatility continues to reflect real yields and term premium rather than a de-anchoring of inflation expectations.<\/p>\n\n\n\n That distinction remains the central rates story. Warsh clarified during the week that he could raise rates in September if inflation strengthens, providing some of the reaction function that was missing from the July FOMC press conference, but he intends to retain his scaled-back communication framework. The initial lack of guidance had itself contributed to a higher uncertainty premium. Bloomberg’s decomposition showed the inflation component of the 30Y premium was relatively stable, while the term premium had risen sharply, reaching its highest level since 2013.<\/p>\n\n\n\n The long end is increasingly trading on fundamentals, partly independently of the expected Fed path: fiscal deficits, Treasury supply, geopolitical risk, and the marginal demand for duration. This helps explain why the 30Y remains above 5.20% despite the substantial reduction in the probability of a September hike. Positioning also remains cautious, with speculative duration shorts increasing by roughly 62k TY equivalents, asset managers reducing ultra-long futures exposure, and Friday’s Treasury rally occurring on only around 70% of average 10Y futures volume.<\/p>\n\n\n\n Treasury provided some near-term supply relief at the QRA, maintaining coupon and FRN auction sizes for at least the next several quarters and changing its language around future auction-size “increases” to the more flexible “changes,” potentially opening the door to reductions in longer-dated issuance. Treasury continues to lean heavily on bills, but the financing requirement remains substantial, with approximately $739bn of net borrowing projected for Jul-Sep. Competition for duration also remains elevated amid heavy investment-grade corporate issuance. The next direct test comes with $42bn of 10Y supply on 2026-08-12 and $25bn of 30Y bonds on 2026-08-13.<\/p>\n\n\n\n Japan remains the most relevant external risk to the Treasury supply-demand balance. BoJ September-hike pricing moved toward 60%, while JGB volatility and large unrealised losses at Japanese institutions reinforce the longer-term risk that higher domestic yields will reduce demand for U.S. duration. At the same time, the U.S.-Japan intervention framework appears designed to limit forced Treasury liquidation, including potential greater use of the Fed’s FIMA Repo Facility.<\/p>\n\n\n\n This week clarified rather than resolved the rates regime. The front end is increasingly a debate about whether and when the Fed hikes again, while the long end is a debate about term premium, fiscal credibility, supply and who ultimately absorbs duration. The reduction in near-term tightening expectations was enough to generate a Treasury rally, but not enough to eliminate the structural premium embedded further out the curve. Next week’s CPI will test the front end, while the $67bn combined 10Y\/30Y auctions will provide the more important test for long-duration demand.<\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n Credit remained constructive, but valuations are increasingly difficult to justify outside carry. IG spreads remain historically tight, with one estimate putting 10Y IG spreads below 100bps vs ~150bps over the long term, while broader fixed-income spreads were described as 95th\u201399th percentile-rich. July IG returns were the worst for the month since 2003, driven primarily by the Treasury sell-off rather than fundamental deterioration, reinforcing that, at current spreads, total returns are increasingly dominated by rates.<\/p>\n\n\n\n The main structural theme remains the migration of the AI capex cycle from internal cash funding toward debt markets. Alphabet reportedly sought up to $25bn in U.S. IG bonds and attracted around $115bn of orders, highlighting continued depth of demand despite recent hyperscaler indigestion. Hyperscalers are estimated at roughly 5% of the IG index, potentially rising toward 10%+, while their bonds trade at around a 25% premium to IG, largely due to expected supply. Aggregate spreads have remained contained because issuance elsewhere is weak: financials, historically 40\u201350% of IG supply, are now below one-third, while issuance in industrials and transport remains below normal.<\/p>\n\n\n\n The more attractive spread opportunities are shifting toward AI infrastructure, where individual data-centre campuses may require $10\u201325bn in debt financing to cover chips, power, cooling, and transmission. Certain structures were cited at 7\u20139% yields, sometimes with investment-grade-like tenant exposure, but the premium reflects material construction, collateral, technology obsolescence, power, and documentation risk. The key underwriting distinction is increasingly structural: direct exposure to Alphabet\/Microsoft is fundamentally different from lending to an SPV dependent on a hyperscaler lease or against rapidly depreciating chip collateral.<\/p>\n\n\n\n Private credit remains more a liquidity than solvency concern. Defaults remain relatively contained, but 2021\u201322 vintages increasingly require amendments, extensions and sponsor support. Retail-oriented private-market funds reportedly received around $14bn of redemption requests in Q1 and $16bn in Q2, or roughly $30bn in 1H26, exposing the mismatch between illiquid assets and \u201csemi-liquid\u201d fund structures.<\/p>\n\n\n\n HY remains unattractive on a risk-adjusted basis. YTD returns were described as roughly equivalent to cash, while investors still assume materially greater default and liquidity risk. There have been 29 defaults totalling ~$36bn YTD, around 15% above the comparable period last year, with the HY bond default rate at approximately 2.8%, still below the historical low of 3% but drifting higher. Rising issuance adds further pressure on medium-term spreads.<\/p>\n\n\n\n Banks remain one of the cleaner relative-value areas. Financial issuance scarcity, strong capital markets activity, and positive earnings revisions support the sector, while high-quality IG floaters offer attractive carry with limited duration and lower exposure to the hyperscaler supply wave.<\/p>\n\n\n\n Consumer credit remains bifurcated. Major-bank portfolios remain healthy, with Bank of America reporting around 5% YoY growth in July spending, but lower-income stress is increasingly migrating outside banks toward BNPL, payday lending and earned-wage-access products. One cited example implied an annualised borrowing cost above 160%, suggesting that bank delinquency data may understate stress among weaker consumers.<\/p>\n\n\n\n Overall, credit is not breaking; compensation is simply thin. Fundamentals remain supportive, primary markets are wide open, and default rates remain manageable, but historically tight spreads, rising AI-related leverage and higher-for-longer base rates leave limited room for error. Bias remains toward quality, banks and floating-rate IG; remain selective in hyperscaler and infrastructure credit, and underweight HY and weaker private-credit structures unless spreads widen materially.<\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n US equities rallied strongly, led by large-cap technology. The Nasdaq surged 5.19%<\/strong>, while the S&P 500 gained 3.59%<\/strong>, the Russell 2000 3.52%<\/strong> and the Dow 2.96%<\/strong>. The equal-weight S&P 500<\/strong> rose a more modest 2.43%<\/strong>, trailing the cap-weighted index by 116 bps and confirming that mega-cap growth remained the primary driver, although strong small-cap performance pointed to some broadening. Outside the US, Germany\u2019s DAX gained 2.69%<\/strong>, and Japan\u2019s Nikkei 225 rose 1.93%<\/strong>, while the FTSE 100 advanced just 0.54%. Emerging markets lagged, with MSCI EM down 0.42% and the Hang Seng falling 0.80%. Information Technology (+7.22%)<\/strong> dominated S&P 500 sector performance, followed by Materials (+5.61%)<\/strong>, Industrials (+3.58%)<\/strong>, and\u00a0Consumer Discretionary (+2.84%), suggesting cyclical participation alongside the technology rally. Communication Services (+1.28%) also advanced, while defensives generally lagged. Real Estate (\u22120.12%)<\/strong>, Utilities (\u22121.68%)<\/strong> and particularly Energy (\u22123.30%)<\/strong> were the weakest sectors.<\/p>\n\n\n\n <\/p>\n\n\n\n The rebound reversed a meaningful portion of July\u2019s technology unwind and brought US equities back toward record highs. Semiconductors were at the centre of the recovery<\/strong>, with the Philadelphia Semiconductor Index gaining 9.3% after falling 20.6% in July, its worst month since 2008. Nvidia rose 11.6%, Marvell 16.6%, Qualcomm 13.7% and Broadcom 9.9%<\/strong>. The speed of the rebound reflected both improving fundamentals and cleaner positioning following July\u2019s deleveraging, but the recovery did not fully reverse the correction, and investors remained more selective about rebuilding exposure. Semiconductor ETFs attracted unusually strong inflows during the sell-off, suggesting investors largely treated the correction as a positioning event rather than a break in the structural AI investment cycle.<\/p>\n\n\n\n Q2 earnings provided unusually strong fundamental support. Around 90% of S&P 500 companies had reported, with blended EPS growth tracking above 30% YoY and eight of eleven sectors delivering double-digit growth. Importantly, the earnings expansion broadened beyond technology: AI and technology\u2019s contribution to S&P 500 earnings growth fell from roughly 90% a year earlier to 57%, as Energy, Consumer Discretionary, Materials and other sectors accelerated. Guidance was similarly constructive, with companies including Caterpillar, Amgen and DuPont raising full-year expectations. Yet exceptionally strong results were increasingly embedded in prices: GS noted that companies beating EPS estimates by more than one standard deviation outperformed the S&P 500 by only 33 bps the following day, versus a historical median of 95 bps, while comparable technology beats actually underperformed by 99 bps.<\/p>\n\n\n\n The most important earnings development was growing evidence that AI CapEx is translating into revenue rather than simply higher investment costs. AWS revenue accelerated 37% to $42.2bn, its fifth consecutive quarter of accelerating growth, while Microsoft Azure grew 43%. Across Google Cloud, Azure and AWS, cloud revenues grew approximately 43% on a $364bn trailing revenue base, alongside improving margins. BCA estimates hyperscaler marginal return on incremental invested capital at close to 30%, despite substantial infrastructure still under construction and therefore not yet generating revenue. That helped address one of the market\u2019s principal concerns following the July correction: whether hundreds of billions of dollars of AI infrastructure spending could generate sufficient returns to justify the capital committed.<\/p>\n\n\n\n The market nevertheless imposed a much higher bar on individual AI companies. Amazon surged around 15% following its results as accelerating AWS growth provided clear evidence of monetisation, while Palantir rallied sharply after another major beat and guidance increase. Conversely, AMD declined despite strong revenue and data-centre growth because guidance failed to clear elevated expectations and investors focused on spending and margins. Memory and storage names displayed the same asymmetry: strong pricing and AI demand remained fundamentally supportive, but results from SanDisk and Western Digital were met with caution as investors questioned the durability of pricing, future capacity additions, and already-extreme expectations. The defining feature of the week was therefore the widening gap between business performance and stock performance: simply being exposed to AI was no longer sufficient; companies increasingly needed to demonstrate accelerating monetisation, operating leverage and credible returns on incremental capital.<\/p>\n\n\n\n At the same time, the AI opportunity continued to broaden beyond GPUs. Hyperscalers lifted aggregate 2026 CapEx guidance by around 5% to approximately $748bn, almost double the prior year, supporting demand across networking, memory, semiconductor equipment, data-centre construction, cooling and electrical infrastructure. Earnings and investor discussions increasingly highlighted bottlenecks in high-speed connectivity, physical data-centre construction and power alongside compute itself. This broadening helps explain why Industrials and Materials participated in the week’s rally and suggests the next phase of the AI trade may be increasingly determined by identifying infrastructure bottlenecks rather than simply owning the largest semiconductor companies.<\/p>\n\n\n\n Earnings strength was also visible outside technology. Consumer and travel results remained generally constructive, with Uber, Expedia and Booking pointing to resilient demand, while industrial and defence companies including Caterpillar and General Dynamics reported strong results and\/or guidance. European earnings also continued to improve, particularly across banks, semiconductors, transport and aerospace. The broader message from the season was therefore more constructive than the index concentration implies: earnings growth and revisions were broadening beyond the largest US technology companies even though those companies continued to dominate headline equity returns.<\/p>\n\n\n\n Europe also continued to present an increasingly credible case for diversification. European earnings are experiencing their strongest growth in several years, while valuations remain substantially below those of comparable US companies, and equity inflows are tracking among the strongest of the past decade. The region’s heavier exposure to Financials, Industrials, Defence, Utilities and other cash-generative businesses has also provided relative protection against volatility in expensive US AI exposures. The structural disadvantage remains Europe’s limited representation among high-growth technology companies, but in an environment where investors are increasingly questioning the funding requirements and returns on US hyperscaler CapEx, Europe’s combination of stronger cash generation, buybacks and lower valuations has become more attractive.<\/p>\n\n\n\n Positioning remained an important part of the market setup. July’s technology correction materially reduced leverage and crowding in several AI trades, but the rapid Week 32 rebound also reduced some of that positioning cushion. Global equity flows remained positive, although capital increasingly diversified beyond the US, while European equities continued to attract significant foreign inflows. Beneath the index, high dispersion and low stock correlation persisted, creating a favourable environment for stock selection but a considerably less forgiving one for companies missing elevated expectations. The combination of cleaner positioning, strong earnings and continued AI investment supports the market, but after the rapid rebound the risk\/reward increasingly depends on earnings delivery rather than another broad valuation re-rating.<\/p>\n\n\n\n Weekend developments did not affect Week 32 returns but complicated the setup for the following week. Renewed incidents around the Strait of Hormuz and Gulf energy infrastructure threatened to reverse part of the week’s decline in geopolitical risk, with potential implications for airlines, logistics and other fuel-sensitive industries while supporting selected Energy and Defence exposures. Reports that the Pentagon was seeking faster production of surveillance, interceptor and missile-tracking equipment reinforced an already strong demand backdrop for defence contractors. In the technology sector, reports that Apple was evaluating CXMT memory for China-market products highlighted both the severity of global memory constraints and the regulatory and competitive implications of sourcing Chinese semiconductors. These developments leave geopolitics, energy and semiconductor supply chains as important cross-currents entering Week 33.<\/p>\n\n\n\n Bottom line: Week 32 was a powerful rebound from July’s technology unwind, supported by cleaner positioning, exceptionally strong earnings and, crucially, growing evidence that the enormous AI investment cycle is beginning to generate measurable returns. The rally broadened into small caps, cyclicals and the physical AI infrastructure stack, but technology remained overwhelmingly dominant. The key change is that the market is becoming substantially more discriminating: AI exposure itself is no longer enough. With valuations elevated and expectations rebuilt quickly, the next phase should increasingly reward companies able to convert AI investment into sustained revenue growth, margins and free cash flow, while punishing even fundamentally strong companies that fail to clear an increasingly demanding bar.<\/p>\n","protected":false},"excerpt":{"rendered":" Rates Treasuries rallied across the curve, partially reversing the previous week’s post-FOMC selloff. The 2Y fell 9.6 bps to 4.20%, 5Y \u22129.7 bps to 4.354%, 10Y \u22129.0 bps to 4.65%, and 30Y \u22127.3 bps to 5.20%. The curve bull-steepened modestly, with 2s10s widening by 0.6 bps to 45.0 bps, 5s30s by 2.4 bps to 84.9 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"nf_dc_page":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-1894","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"blocksy_meta":{"styles_descriptor":{"styles":{"desktop":"","tablet":"","mobile":""},"google_fonts":[],"version":7}},"yoast_head":"\n
\n\n\n\nCredit<\/h2>\n\n\n\n
![\"\"](\"https:\/\/i0.wp.com\/karolpelc.com\/InvestorSnippets\/wp-content\/uploads\/2026\/08\/credit_yield_vs_spread_2026-08-08_W32.png?fit=1024%2C576&ssl=1\")
<\/figure>\n\n\n\n
\n\n\n\nEquities<\/h2>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/i0.wp.com\/karolpelc.com\/InvestorSnippets\/wp-content\/uploads\/2026\/08\/global_index_weekly_2026-08-08_W32.png?fit=1024%2C576&ssl=1\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/i0.wp.com\/karolpelc.com\/InvestorSnippets\/wp-content\/uploads\/2026\/08\/sp500_sector_weekly_2026-08-08_W32.png?fit=1024%2C576&ssl=1\")
<\/figure>\n\n\n\n