Morgan Stanley Global Macro Forum: Bank earnings, cooling inflation, and AI efficiency together point to a more moderate rate path

Overview

This edition of Morgan Stanley Global Macro Forum centers on three signals: Moonshot’s release of a low-cost open-source large model, strong U.S. Q2 bank earnings, and June inflation showing disinflation progress. The report combines these signals into one macro judgment: the U.S. economy and financial conditions remain resilient, the AI investment cycle continues to broaden, and the inflation path may be more moderate than the market fears.

Core views

The core views are as follows. First, large and mid-sized U.S. banks broadly beat expectations in 2Q26, with solid performance in capital markets, investment banking, and loan growth, indicating continued support from the economy and credit environment. Second, June CPI/PCE-related data released disinflation signals, and Morgan Stanley maintains its non-consensus view that the Fed stays on hold this year and cuts 50bp in 2027. Third, market pricing still implies a relatively high probability of rate hikes; the report argues that this path is too hawkish and that if inflation continues to cool, the rates market should reprice. Fourth, improved AI model efficiency may lower the unit cost of using AI, but could also, through the Jevons paradox, increase demand for compute power and AI infrastructure.

Analysis framework

The report uses a cross-asset macro framework, combining signals from bank earnings on lending, capital markets, and credit quality with inflation, oil prices, tariffs, software prices, and Fed expectations, and further mapping them to U.S. yield curve trades. Bank earnings are used to observe the real economy and risk appetite, inflation data to assess the policy reaction function, and market-implied rate paths to identify pricing distortions.

Methodology notes

• Macro policy pathFed reaction function and annualized inflation path

  Use monthly, three-month, and six-month annualized changes in core PCE to judge whether the Fed needs to continue hiking.

  The report forecasts June headline PCE and core PCE at -0.08% and 0.17% month over month, respectively, and argues that inflation in the second half of the year will be closer to a 2% annualized pace, thereby supporting the Fed holding rates steady.

• Bank earnings signalsEPS beats, capital markets revenue, and loan growth

  Use bank earnings to observe consumers, commercial credit, capital markets, and credit quality.

  15 of 16 U.S. banks beat consensus EPS expectations, trading and investment banking revenue was strong, commercial loan growth broadened, and problem assets declined, supporting the view of economic resilience.

• Rates strategyYield curve steepening

  When the market gradually removes near-term hike risk, curve steepening trades become more favorable.

  The report argues that the current market-implied path of nearly 50bp of hikes through March 2027 is too hawkish, and if inflation continues to cool, there is room for the front-end rate risk premium to decline.

• Technology and macro demandJevons paradox

  Improved model efficiency lowers the cost of AI use, but may instead lift total demand and compute consumption.

  Moonshot’s Kimi K3 is described as potentially delivering frontier-level performance at 40%-70% lower cost; if true, this would raise the likelihood of broader AI adoption and increased demand for AI infrastructure.

Asset mapping & comparison

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

• Large and mid-sized U.S. banks

  Benefit from earnings beats, strong capital markets revenue, and broadening loan growth.

  Strengths

  EPS broadly beat expectations, trading and investment banking businesses grew, consumer credit and commercial lending performed better than expected, and credit trends were benign.

  Weaknesses

  Valuation, earnings revisions, and credit quality still depend on the macro cycle and the interest-rate environment.

  Comparison

  Within capital markets, the year-over-year growth rate of equities trading was significantly higher than FICC, and investment banking also showed a broader recovery.

  Risks

  If the economy slows, employment deteriorates, or credit losses rise, the resilience of bank earnings could weaken.

• U.S. Treasuries and the yield curve

  Less tightening pressure supports yield curve steepening.

  Strengths

  Cooling inflation and a labor market that is not overheating reduce near-term hike risk, and market repricing could lower the front-end rate risk premium.

  Weaknesses

  The trade is sensitive to the view that inflation remains moderate and that the Fed will not hike again.

  Comparison

  The report argues that the market-implied path is more hawkish than Morgan Stanley economists’ baseline.

  Risks

  If inflation reaccelerates or the Fed delivers a more hawkish signal, the steepening trade could come under pressure.

• AI infrastructure and compute demand

  Improved AI model efficiency may lower usage costs and expand demand.

  Strengths

  If low-cost models are validated, they could drive broader AI adoption and increase demand for compute and infrastructure.

  Weaknesses

  Moonshot-related performance and cost advantages still require validation, and adoption is also affected by geopolitics, regulation, and strategic considerations.

  Comparison

  The report views innovation in Chinese models as a signal of accelerating AI innovation outside the U.S.

  Risks

  Regulatory restrictions, geopolitical scrutiny, insufficient model performance validation, or the upward price impact of AI demand.

Key data

• U.S. bank EPS beats15 of 16 banks beat consensus, with the median beat around 10%From the 2Q26 bank earnings section.
• Equities trading wallet growth+71% y/yThe 2Q26 equities trading wallet hit a record and exceeded FICC in nominal terms for the first time.
• FICC trading wallet growth+12% y/yTogether with equities trading, this drove capital markets revenue.
• Investment banking wallet growth+46% y/yIncluding ECM +87%, DCM +41% and a record high, and M&A +27%.
• June PCE forecastHeadline PCE -0.08% m/m, core PCE 0.17% m/mThe report uses this forecast to support its disinflation and Fed-on-hold view.
• Annualized core PCE changeJune three-month annualized 3.00%, six-month annualized 3.82%Below May’s 3.52% and 4.14%.
• Fed policy viewUnchanged in 2026, 50bp of cuts in 2027Morgan Stanley’s relatively non-consensus view.
• Market-implied policy pathNearly 50bp of hikes implied through March 2027The report argues the market assigns more than a 50% probability to a scenario its economists view as a tail risk.

Impact & implications

For asset allocation, the implications are that U.S. bank earnings and loan growth support the fundamental resilience of risk assets, while cooling inflation reduces the need for further rate hikes, and the rates market may be overpricing a hawkish path. If this view plays out, U.S. yield curve steepening trades would have a stronger rationale, while banks and capital-markets-related assets would benefit from growth in trading, investment banking, and lending; however, price pressures from oil, tariffs, software prices, and AI demand could still disrupt the disinflation path.

Risks

• Conflict in the Middle East could push WTI oil prices close to or above $80 per barrel, potentially undermining the base-case disinflation scenario.
• Section 232 and 301 tariff reviews are still ongoing, and tariff pass-through may continue to affect price levels.
• Persistently rising software prices could become an upside risk to core inflation.
• AI-related demand could offset the disinflationary effects from tariffs and other goods prices.
• If subsequent inflation data reignites the Fed’s hawkish tail risk, the yield curve steepening trade could fail.
• If bank credit quality shifts from declining problem assets to migration into losses, the current constructive view on banks would be challenged.

What to watch

• Whether the three-month and six-month annualized trends in subsequent core CPI and core PCE continue to move toward 2%.
• Whether the Fed’s communication on keeping rates unchanged for the remainder of 2026 remains stable.
• The pass-through from WTI oil prices and the Middle East situation to energy prices.
• The outcome of tariff reviews and their impact on goods price levels.
• Subsequent trends in U.S. bank loan growth, credit quality, problem assets, and capital markets revenue.
• The actual performance of low-cost AI models such as Moonshot Kimi K3, adoption constraints, and their pull on compute demand.