Cloud backlogs jump 150% to $1.7 trillion as AI capex converts

Hyperscaler cloud order backlogs jumped more than 150% year over year to roughly $1.7 trillion, more than double the pace of capital spending, as AI infrastructure investment converts into contracted revenue.

"Cloud backlogs and net new annual recurring revenue are growing well ahead of capital expenditure, meaning future revenue should cover current infrastructure outlays," Mark Schilsky, an analyst at JPMorgan, said.

Amazon chief executive Andy Jassy raised his long-term view of AWS, saying the cloud unit could eventually become a $1 trillion annual revenue business, double his prior estimate, and that demand through 2028 looks "stunning." Microsoft, Alphabet and Meta executives echoed the theme, describing AI commercialization as still in its early expansion phase.

The S&P 500 information technology sector now trades at about 20 times forward earnings, near the 1st percentile of its historical range and below the 10-year average of 23 times. JPMorgan estimates megacap technology stocks excluding semiconductors sit more than two standard deviations below their post-2018 average multiple, implying 30 percent to 56 percent upside if valuations normalize.

Backlogs Outpace Capex as Demand Shifts to Inference

The gap between order growth and spending marks a turning point in the AI investment debate. For more than a year, investors questioned whether the trillions of dollars poured into AI infrastructure would ever generate returns. The backlog data suggests the opposite: hyperscalers are locking in revenue faster than they are building capacity.

JPMorgan's Schilsky said the divergence — 150 percent backlog growth versus roughly 80 percent capex growth — shows the potential revenue return on AI infrastructure is beating market expectations. Amazon's Jassy said enterprise adoption of AI inference services remains in its early stages, with corporate customers only beginning to scale deployments. The second-quarter earnings season reinforced the theme: consensus estimates for S&P 500 earnings growth have been raised from 22.4 percent at the start of reporting to roughly 45 percent today, one of the steepest revision paths on record outside post-recession recoveries.

Valuations Compress Even as Earnings Estimates Rise

The fundamental improvement has not been matched by multiple expansion. After July's selloff, the S&P 500 IT sector trades at roughly 20 times forward earnings, near one-year lows and below its 10-year average of 23 times. Megacap technology names excluding semiconductors sit more than two standard deviations below their historical average multiple.

Deutsche Bank data shows institutional investors hold only a modest overweight in large-cap technology despite sharply higher earnings growth and forecasts — well below the positioning seen in past strong earnings cycles. Money has concentrated in semiconductor names this year, leaving broader megacap tech under-owned.

JPMorgan argues that if the market narrative shifts from "is capex excessive" to "returns are materializing," the next leg of tech gains may come from rotation within the sector rather than further semiconductor strength. The MAGS index has rebounded about 10 percent from its recent low, reclaiming its 200-day moving average, which has flattened — a pattern that historically precedes stronger breakouts after longer consolidation.

For investors, the setup pairs compressed valuations with improving fundamentals and underweight positioning. Citadel's Scott Rubner said July delivered the technical reset markets needed, with retail selling, lower leverage and reduced concentration clearing excess built up during the early-summer rally. Buyback blackout windows are expiring — roughly 45 percent of the S&P 500 by weight is eligible to repurchase shares today, rising to 75 percent by next week and 85 percent by mid-August — restoring a major source of equity demand just as positioning rebuilds. If JPMorgan's thesis holds, the re-rating of megacap technology could draw significant capital inflows.

This article is for informational purposes only and does not constitute investment advice.