Can the S&P 500 rise further despite high valuations? HSBC says yes: here's why

HSBC's Willem Sels said US equities may have further upside despite elevated valuations, arguing that AI-driven earnings and productivity gains are not yet fully priced in. He noted that investor skepticism remains high over the durability of corporate earnings growth, particularly for technology and semiconductor companies, leaving room for positive repricing if growth proves sustainable.

Analysis

The investable question is not whether AI raises aggregate productivity, but whether earnings revisions can outrun the valuation premium already embedded in the narrow AI complex. Broad market upside requires the benefit to migrate from infrastructure vendors into software, financials, industrial automation, and consumer-facing enterprises; otherwise, index gains remain vulnerable to a de-rating in a small group of high-weight names. The most attractive second-order beneficiaries are firms where labor expense is a large, measurable share of costs and implementation can be verified through margin guidance rather than promotional AI announcements.

Over the next 1-3 months, the key catalyst is third-quarter guidance: evidence of lower SG&A, customer-service, coding, or sales-support costs can support upward EPS revisions and broaden participation beyond semiconductors. Over 6-18 months, the risk shifts to AI capex digestion: if cloud spending continues to rise faster than enterprise monetization, hyperscaler free-cash-flow expectations and their suppliers' order books become exposed. HSBC is not a clean expression of the thesis; any benefit is indirect through wealth-management risk appetite and capital-markets activity, while its valuation remains more sensitive to Hong Kong/China credit conditions and rate normalization.

Contrarian view: skepticism alone is not bullish if it reflects a rational demand for proof of return on capital. The consensus may be underestimating the lag between deployment and margin realization, particularly for regulated industries with costly data integration and compliance requirements. A durable broadening thesis is falsified if AI-adopting companies fail to raise 2027 margin targets during the next two earnings cycles, or if hyperscalers signal capex moderation without corresponding acceleration in enterprise AI revenue.

Key Decisions for Investors

• Do not add HSBC solely on the AI narrative; treat HSBC as a watch item for improved wealth inflows and investment-banking fees, with China/Hong Kong credit indicators as the primary risk variable over 6-12 months.
• Favor a 6-12 month quality-broadening basket: long MSFT and ORCL versus short a basket of lower-quality AI infrastructure proxies after checking current valuation and short interest. The thesis requires enterprise AI revenue and operating-margin upside; exit if either company cuts cloud/AI growth guidance.
• Use an AI-adoption versus infrastructure pair over the next two earnings seasons: long ETF IGV or selected profitable software beneficiaries versus short SOXX only if semiconductor order visibility weakens while software guidance begins to show measurable margin lift. This is an alert, not an immediate recommendation, because relative valuations and current earnings revisions are required.
• For broad equity exposure, prefer staged S&P 500 participation rather than chasing index strength: add only following earnings-revision breadth expansion beyond mega-cap technology. A failure of revision breadth to improve by year-end would argue for reducing beta, as multiple expansion alone is unlikely to sustain the advance.