Artificial intelligence investment momentum is expected to remain a vital pillar of corporate earnings through at least 2027, according to Goldman Sachs, which argues that an outright decline in AI spending is unlikely before 2028 or even later. The firm's chief economist, Jan Hatzius, projects that earnings per share for S&P 500 constituents could climb 36% by 2026, fueled by private equity investment gains and the outsized dividends stemming from AI-related capital expenditure, including margin expansion across semiconductor and memory companies.

Goldman Sachs does anticipate that this exceptional pace of growth will eventually cool. The firm's US equity strategists believe that as private equity investment gains fade and AI spending approaches its peak, S&P 500 EPS growth will moderate to 11% in 2027. However, Hatzius emphasized that a significantly sharper slowdown would only materialize when AI investment begins to decline in absolute terms, a scenario the bank views as unlikely before 2028 or beyond.

The bank contends that talk of an earnings bubble is exaggerated, even though current profitability does reflect a state of "earnings excess" relative to a more normal cycle, driven primarily by AI-related boosts. Goldman estimates that AI infrastructure alone will contribute approximately 15 percentage points to S&P 500 EPS growth in 2026, with contributions tapering to 11 points in 2027 and 8 points in 2028. Hyperscaler cloud providers are expected to add another 4, 3, and 3 percentage points in those respective years.

The real test arrives in 2028, and Goldman's stance that AI spending will not see an absolute decline is far from isolated on Wall Street. Over the past few months, major investment banks have aggressively revised their S&P 500 earnings forecasts upward, with the debate shifting from whether growth will occur to how long AI can sustain index-level earnings expansion.

Morgan Stanley stands out with the most aggressive upward revision. Chief Cross-Asset Strategist Andrew Sheets raised his 2026 S&P 500 earnings growth forecast from 17% to 23% in his medium-term outlook, citing how the AI capital expenditure wave is reshaping the "inelasticity of demand" across the US economy. He projects capital spending by large cloud computing firms to reach $805 billion in 2026, nearly doubling previous year-ago estimates. The bank also flagged the flip side of growth, noting that combined capital expenditures for Microsoft, Alphabet, Amazon, and Meta are expected to surge another 57% from 2026 to 2027. Morgan Stanley has already trimmed its 2027 free cash flow forecasts for these four companies, suggesting that AI-related borrowing will need to accelerate before cash flow can catch up with capital outlays.

JPMorgan's narrative centers on transitioning from "committed spending" to "realized revenue." The bank has raised earnings expectations twice this year, lifting its 2026 S&P 500 EPS estimate from $350 to $365 and its 2027 figure from $390 to $420, with the year-end price target climbing to 8,000 points. Strategists there point out that as backlogged orders convert into recognized revenue, cloud business growth gains strong support, helping to justify the expanding AI capital expenditure outlay and alleviate market concerns about returns on invested capital. Notably, JPMorgan has not raised its valuation assumptions, keeping the forward P/E ratio near 20 times, which indicates that its bullish case leans almost entirely on earnings delivery rather than multiple expansion.

Barclays and HSBC have struck a more exuberant tone. Barclays US Equity Strategy Head Venu Krishna lifted his year-end target to 7,950 points, describing a "stellar" earnings season led by the tech sector that reinforces the case for durable, AI-supported earnings growth. He also raised his 2026 EPS forecast sharply from $337 to $365, while candidly acknowledging lingering doubts about AI spending durability, inflation stickiness, and a more hawkish rate outlook, which is why his valuation assumptions remain restrained. HSBC Equity Strategy Head Nicole Inui made a dramatic single-step move, raising her target by 450 points to 8,100, projecting S&P 500 earnings growth approaching 40% in the first half of 2026 and around 25% in the second half.

UBS emphasizes the "prepaid" nature of demand, noting that hyperscalers have already placed pre-purchase compute orders worth up to $2 trillion. The bank expects roughly half of 2026 index-level earnings growth to originate from the tech sector. Across Wall Street, there is now little disagreement that AI serves as the core engine for S&P 500 EPS growth. The genuine points of divergence center on three key issues: the speed at which capital expenditure converts into revenue, the financing approach for closing the free cash flow gap, and who picks up the growth baton once spending peaks around 2028.