Bull-Bear Divergence on US Q2 Earnings: Goldman Sees 22% Growth While Morgan Stanley Warns of AI Monetization Risks
Taylor Wilson
Ahead of Q2 earnings season, Goldman Sachs projects S&P 500 profits up 22% year-on-year while Morgan Stanley warns that AI capex returns are dulling — the core dispute is whether AI investment can generate real profit beyond hardware.
What is the consensus number for Q2 earnings?
FactSet-tracked Wall Street consensus has climbed from roughly 15% at the start of the year to an intra-year high of 21.3%, then eased slightly to 19.9%.
Goldman's team is more bullish: S&P 500 Q2 earnings growth of 22%. Bloomberg Intelligence aggregate data suggests the gain could approach 23%.
This means → mainstream houses agree on a 20%–23% growth band. The debate is not *whether* earnings grow, but *the quality* of that growth.
Who is carrying this growth?
Goldman estimates AI-infrastructure stocks will account for nearly 60% of S&P 500 Q2 EPS growth. Micron and Nvidia alone may contribute over 40%.
In plain terms = the whole S&P 500 looks like it is growing, but close to six-tenths of that comes from a handful of AI-hardware sellers — strip them out and the number shrinks sharply.
Energy is another variable: Goldman notes rising oil prices are lifting oil-producer profits, but growth expectations diverge widely across sectors.
What are the bulls betting on?
Bank of America argues Q2 will mark a milestone — AI dividends spreading from "hardware monopoly" to the full value chain, reaching power, legacy server assembly, industrials and financials.
UBS points out that the 493 S&P 500 constituents outside the tech "Magnificent Seven" are expected to post their first collective earnings acceleration in nearly two years.
This means → the bull case is not just "earnings keep rising" but "the base is widening." If confirmed, the bull market no longer depends on a few tech giants alone.
What are the bears worried about?
Morgan Stanley warns: if downstream software, healthcare and retail firms cannot prove in their filings that AI is generating actual profit, the current tech-stock premium becomes unsustainable — risk of a sharp correction once "buy the rumour" turns to "sell the news."
In plain terms = upstream is selling shovels furiously (chips, servers), but if the downstream gold-diggers find no gold, even the shovels lose their premium.
JPMorgan flags the "double-edged sword" of rising energy prices: higher oil lifts energy-giant profits but effectively extracts margin from the real economy — consumer goods, airlines and logistics are expected to see profit-margin compression.
What is the single question this earnings season must answer?
Both sides converge on one verification point: can AI capex generate visible earnings returns beyond the hyperscale cloud companies?
JPMorgan adds a transmission chain: rising crude intensifies inflation pressure → may force the Fed to hold rates higher for longer → weighing on Q3 equity valuations.
This reflects a pricing logic that has moved past "how much do they earn" to "whose pocket does the AI money actually land in" — the answer will emerge report by report over the coming weeks.
Content is for reference only, not financial advice.