Global multi-asset implications of US inflation breadth and AI adoption Report Interpretation

Overview

This global multi-asset strategy report assesses inflation breadth after Warsh's Jackson Hole remarks and the economic and market effects of AI adoption. HSBC sees recent inflation data improving at the margin and believes AI-linked productivity, semiconductor demand and US profitability support its US equity overweight.

Core views

HSBC revisits concerns that US inflation remains too broad. Using nearly 200 core-PCE subcomponents, it finds that about 54% of the basket is currently inflating at 3% or more. That is below the post-pandemic peak of roughly 77%, though above the 32% level seen in the two decades before the pandemic. HSBC notes that this breadth is high relative to the 2010s but only modestly above the range seen in the 2000s; the shares of components above 2%, 5% and 10% also resemble the pre-global-financial-crisis period more than the post-pandemic extreme. The report argues that the more recent signal is constructive: while year-on-year measures show limited change, a three-month annualised version of the PCE heatmap indicates broad disinflation since February. With a September rate hike around 70% priced, HSBC says upcoming US releases are pivotal. Its labour and inflation dashboards already show a significant loss of momentum in the third quarter, and further weak data over the following weeks would be needed to convince markets that the Federal Reserve will not hike. On AI, HSBC believes the productivity effect is now becoming visible in company fundamentals. Across the US and Europe, AI adopters have delivered stronger revenue growth, stronger earnings growth and faster-rising margins than non-adopting peers, helping explain the recent strength of earnings despite higher yields. The US shows stronger improvement in fundamentals than Europe when comparing the reported charts on a like-for-like scale, although banks are an exception. HSBC continues to favour European banks and views them as one of the better ways to participate in AI adoption. The institution's adoption measures count companies that provide tangible examples of AI use on earnings calls, and these measures continue to rise. It cautions that this does not fully capture intensity, such as token consumption or agentic adoption. However, RampAI's measure of US business AI spending accelerated again through the summer. Together with renewed increases in DRAM and NAND spot prices after a period of semiconductor underperformance, HSBC believes the conditions are forming for renewed semiconductor outperformance. HSBC also expects hyperscalers to begin performing well again and says these trends underpin its US equity overweight. It accepts that RoEs should rise across major markets with AI adoption, but rejects the consensus expectation that US RoEs will fade relative to other regions from 2027. Its reasoning is that US companies show stronger AI-adoption evidence outside financials, the US remains the centre of data-centre construction, the US index has greater semiconductor exposure than Europe, and a near 15-year structural trend has been for US profitability to improve relative to the rest of the world.

Analysis framework

HSBC first measures inflation breadth across nearly 200 PCE components and compares current readings with the 2000s, 2010s and post-pandemic peak. It then shifts to three-month annualised inflation to assess recent momentum, compares the fundamentals of AI adopters with non-adopters in the US and Europe, tracks AI spending and memory-chip pricing, and links those findings to equity-sector performance and relative RoE expectations.

Methodology notes

• Other

  PCE inflation heatwave

  HSBC sorts nearly 200 PCE subcomponents by their inflation rates to show how widely price pressures are spread across the consumption basket, including on a three-month annualised basis.

• Corporate Fundamentals and FinanceEarnings Quality Analysis

  Comparison of AI adopters and non-adopters

  The report compares revenue growth, earnings growth and margin trends between companies that disclose tangible AI adoption and peers that do not, using the comparison as evidence of AI-related productivity gains.

• Quantitative, Factor, and Portfolio TheoryStyle factor analysis

  Systematic sentiment and positioning indicators

  HSBC combines normalized market and positioning inputs, including momentum and moving-average signals across several horizons, to monitor contrarian buy and sell sentiment conditions.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• US equities

  HSBC maintains an overweight, supported by AI adoption, data-centre investment, semiconductor exposure and a favourable relative RoE view.

  Strengths

  Stronger evidence of improving AI-adopter fundamentals, leadership in data-centre buildout and meaningful semiconductor index weights.

  Weaknesses

  Consensus expects US RoEs to wane relative to other major markets from 2027.

  Comparison

  HSBC sees stronger recent fundamental improvement in the US than in Europe outside financials.

  Risks

  A lack of further weakening in US labour and inflation data could sustain expectations of Fed tightening.

• European banks

  Highlighted as one of HSBC's preferred ways to participate in AI adoption.

  Strengths

  The report identifies banks as an exception to the broader pattern of stronger US fundamental improvement.

  Comparison

  HSBC continues to prefer European banks relative to the broader banking comparison presented.

• Semiconductors

  HSBC expects conditions to support renewed outperformance after a summer of underperformance.

  Strengths

  Renewed increases in DRAM and NAND spot prices alongside accelerating AI demand.

  Weaknesses

  The sector had underperformed over the summer.

  Comparison

  US equity indices have greater semiconductor weights than Europe, according to HSBC.

• US hyperscalers

  HSBC expects them to begin performing well again as AI demand and investment remain elevated.

  Strengths

  They are linked to the US data-centre buildout and accelerating AI adoption.

Key data

• PCE components with inflation at or above 3%Around 54%Below the roughly 77% post-pandemic high but above the 32% pre-pandemic two-decade level cited in the report.
• September rate hike priced by the marketc.70%HSBC says upcoming US labour and inflation data are crucial for whether that pricing persists.
• Timing of broad disinflation signalSince FebruaryBased on HSBC's three-month annualised PCE inflation heatwave.
• S&P 500 Q2 2026 EPS growth51.2%Reported year-on-year growth in the report's S&P 500 earnings table.
• US RoE relative outlookConsensus expects weakening from 2027HSBC disagrees and expects the longer-running US profitability advantage to reassert itself.

Impact & implications

HSBC's interpretation is that easing inflation momentum can reduce the need for further Fed tightening, while observable AI productivity and rising memory-chip prices support renewed leadership by semiconductors and hyperscalers. These factors underpin its US equity overweight; European banks are highlighted as an additional AI-adoption exposure.

What to watch

• Upcoming US labour and inflation releases, which HSBC says must show further loss of momentum to reduce expectations of a Federal Reserve hike.
• Whether AI spending continues to accelerate and adoption broadens beyond disclosed examples on earnings calls.
• DRAM and NAND spot-price trends as evidence of sustained AI-related memory demand.
• Whether US RoEs hold up relative to other major equity regions from 2027, contrary to consensus expectations.