The Fed's main inflation measure will be released Wednesday. Here's what to expect

August PCE inflation is expected to rise 0.3% month over month in both headline and core measures, leaving annual inflation at 3.7% and 3.3%, respectively—well above the Fed's 2% target. Consumer spending is forecast to accelerate to 0.8% in August from 0.2% in July, while Bank of America data showed card and debit spending up 6.9% year over year, reinforcing demand resilience. Fed officials cited tariffs, the Iran war and AI-related goods demand as contributors to persistent price pressures, and markets are pricing a strong probability of an October rate hike followed by another increase in December or January.

Analysis

The key market variable is not the headline inflation print but whether nominal consumption remains resilient enough to keep real policy rates from becoming restrictive. That combination extends the “higher-for-longer” regime: front-end yields and real yields should stay supported, pressuring long-duration software, unprofitable growth and rate-sensitive housing, while favoring cash-generative value. A downward historical revision would likely create a one- to three-day relief rally, but it does not alter forward policy unless sequential core services and wage-sensitive categories decelerate materially.

BAC is a qualified beneficiary: elevated short rates support asset yields, but the upside is capped if deposit betas reaccelerate and commercial-real-estate losses rise as refinancing costs reset. GS has less deposit-franchise protection and greater sensitivity to a risk-asset drawdown, though sustained volatility can offset this through trading revenue; the more attractive expression is likely long large-bank net-interest-income exposure versus capital-markets beta. For ALV, resilient vehicle demand is constructive, but tariffs and input-cost pressure create a lagged margin risk that OEM pricing may not fully absorb over the next two quarters.

The underappreciated second-order effect is that AI infrastructure demand can keep goods and power-related investment firmer than conventional rate-cycle models imply, delaying the expected disinflation in capital goods and services tied to buildout. That is supportive of select power, electrical-equipment and semiconductor-capex suppliers, but a policy mistake risk rises if the Fed reacts to backward-looking nominal spending inflated by energy prices. A meaningful downside catalyst for yields would require weaker real consumption, not merely a revised prior inflation level; conversely, another upside surprise could reprice a terminal-rate premium quickly and compress equity multiples over the next one to three months.

Key Decisions for Investors

• Maintain a 1-3 month long BAC / short GS pair, sized market-neutral: BAC should retain relative earnings support from balance-sheet income while GS is more exposed to multiple compression if financial conditions tighten. Exit if BAC deposit costs accelerate materially or if GS trading and underwriting estimates are revised upward enough to close the earnings-growth gap.
• Use a tactical long 2-year Treasury yield expression via short SHY or long 2-year Treasury puts into the next inflation and labor releases; the asymmetry favors higher front-end yields if sequential core inflation remains firm. Cover on a clear downside surprise in monthly core inflation coupled with weakening real spending, rather than on backward-data revisions alone.
• Reduce exposure to high-duration, cash-flow-negative technology over the next 1-3 months; hedge broad duration risk with a relative long XLF / short ARKK position. The trade fails if incoming data force a rapid policy-pivot repricing and long-end yields fall decisively despite firm consumption.
• Keep ALV on watch rather than initiate: require evidence that OEM production schedules remain intact and that tariff/input-cost pass-through preserves margin guidance. A cut to 2026 margin outlook or renewed European auto-volume weakness would support a short; absent those datapoints, the macro signal is insufficient.