For two years, one number haunted every AI bull thesis on Wall Street: return on invested capital. Hyperscalers were spending at a pace that dwarfed any prior infrastructure cycle, and skeptics asked, reasonably, whether the revenue would ever justify it. This morning, JPMorgan’s strategists gave their clearest answer yet.
JPMorgan raised its year-end target for the S&P 500 to 8,000 from 7,800, citing prospects of solid corporate earnings and rising confidence that AI investments by large hyperscalers would drive faster revenue growth, Reuters reported Monday. The team led by Dubravko Lakos-Bujas has now raised its target three times in four months, having previously bumped the target to 7,800 from 7,600 in late June.
Why This Revision Is Different
The June upgrade leaned on macro optimism and a favorable geopolitical backdrop. Monday’s move rests on something harder to dismiss: actual earnings data. Of the 493 S&P 500 companies that had reported earnings for the first quarter of 2026 as of June 5, 84.2% beat analyst expectations, according to LSEG I/B/E/S data, above the long-term average of 67%. That is not a round of modest beats on sandbagged guidance. It is a genuine step-change in corporate profitability.
The specific language Lakos-Bujas used matters. “As elevated backlogs convert into recognized revenue, cloud growth should remain well supported, helping validate rising AI capex, strengthen order coverage, and further ease ROIC concerns,” JPMorgan analysts said, according to Reuters. The ROIC framing is deliberate. It signals that the firm is no longer treating AI spending as a leap of faith, but as a capital allocation cycle where the return leg is now visible.
The Backlog Evidence
The cloud metrics from Q2 make the case concrete. AWS revenue growth accelerated to 36.7% year over year in the second quarter, Amazon said, and Google Cloud revenue rose 82% year over year to $24.8 billion, Alphabet reported. Google said Google Cloud’s backlog rose to $514 billion from $460 billion the prior quarter. Those are not vanity metrics. Backlog is contracted future revenue. When it swells at that pace, the question of whether AI spending generates returns stops being theoretical.
According to JPMorgan’s analysis, the tangible benefits from escalating AI infrastructure investments became more apparent during the second quarter. Accelerating cloud revenue, expanding order backlogs, and enhanced cash-flow predictability helped alleviate market skepticism regarding AI expenditure returns.
The EPS Math
JPMorgan raised its 2026 EPS estimate to $365, implying 35% year-over-year growth and above the consensus estimate of $358, while lifting its 2027 estimate to $420, or 15% growth.
There is one asterisk worth flagging. Strategists noted that private-company stake valuations are boosting EPS by roughly $18 based on first-half 2026 marks; excluding that contribution, normalized 2026 EPS would be $347, up 28% year-over-year. Twenty-eight percent organic growth is still extraordinary by any historical standard. But investors pricing the market at 20 times forward earnings should know what they are buying.
The Broader Earnings Picture
JPMorgan is not alone in this upgrade cycle. The new target adds to a growing wave of bullish calls, with at least seven research firms raising their targets in June, Reuters reported. Goldman Sachs has independently mapped the same fundamental story: the consensus of analyst estimates is for the largest hyperscale tech companies to spend $754 billion on capital expenditures this year, an 83% increase from 2025, and $905 billion in 2027.
Goldman’s researchers also highlighted a structural shift in where the profits are landing. Semiconductor companies are the primary direct beneficiaries, and a number of tech hardware, industrials, and utilities companies are also getting a large earnings boost from the AI buildout. That broadening matters. A rally driven entirely by five or six mega-caps is fragile. One extending into industrial suppliers, power infrastructure, and mid-cap software is a different animal.
The return-on-equity data reinforces the profitability argument. S&P 500 return on equity reached a record 22% in the first quarter of 2026, surpassing the previous peak set in 2021, with rising profit margins as a key driver, Reuters reported. Critically, the S&P 500 has gained year-to-date driven primarily by stronger earnings expectations rather than valuation expansion, Reuters reported. That distinction separates a sustainable re-rating from a speculative melt-up.
The Risks JPMorgan Is Not Dismissing
The bull case has a structural vulnerability that no amount of backlog growth fully resolves. Free cash flow is expected to stay under pressure.
JPMorgan is maintaining the forward valuation multiple target at about 20 times, citing geopolitical risks and higher interest rates among key potential headwinds for the market. Holding the multiple flat while raising EPS forecasts is the right framework. It means the path to 8,000 runs through earnings delivery, not further multiple expansion. If the EPS trajectory holds, the market gets there. If it cracks, the multiple offers no cushion.
Concentration remains the clearest structural risk. The very companies that need continued hyperscaler capex to justify their earnings trajectories are the same companies whose results underpin the index. The circular nature of that dependency is not fatal, but it demands that investors distinguish between companies with genuine pricing power and sticky adoption from those simply riding the spending wave.
What Investors Should Watch Next
Three variables will determine whether JPMorgan’s 8,000 target holds or forces another revision. First, backlog conversion: backlogs at $514 billion for Google Cloud are only valuable if they convert to recognized revenue on schedule. Any slowdown in that conversion rate will show up in cloud growth deceleration by Q4 2026. Second, free cash flow recovery: except for Microsoft, JPMorgan projects negative free cash flow for most of the hyperscaler group through 2026 and 2027. The market can absorb this if it believes the depreciation cycle peaks and free cash flow rebounds by 2028. Evidence of that peaking matters more than any single quarterly number. Third, earnings breadth: while the IT sector led with gains of approximately 50% in the first quarter, non-tech sectors also delivered robust performance, posting around 20% growth on average. If that non-tech growth rate holds through the second half, the concentration risk softens considerably.
Bottom Line
JPMorgan’s move to 8,000 is the most significant Wall Street call of the week because it rests on a specific, testable claim: AI spending is generating returns, and those returns are now visible in cloud backlog, revenue acceleration, and record corporate ROE. The market is shifting from asking whether companies can spend on AI to asking whether they can turn it into profits, with upside depending on earnings delivery, not valuation hype. That framing is correct, and it is why the EPS revisions to $365 for 2026 and $420 for 2027 are the numbers that actually decide where this market goes next, not the index target itself. Watch the Q3 cloud growth figures when they arrive in October. They are the first real stress test of whether the backlog-to-revenue conversion is running on schedule.
