Jimmy Chang, Chief Investment Officer at Rockefeller Global Family Office, used CNBC’s Closing Bell Overtime to deliver a pointed warning ahead of Big Tech earnings: the record AI CapEx cycle may already be masking an overbuild that markets have not yet learned to see. Chang argued that strong earnings are already priced in and the real risk lies in CapEx guidance tone, flagging Microsoft (NASDAQ: MSFT<\/a>) as the pivotal earnings report. “Any sign of caution could lead to a rotation of money out of some of these CapEx beneficiaries,” he said. <\/p>\n The numbers behind the concern are staggering. Microsoft told analysts on its most recent call that Q4 2026 CapEx will exceed $40 billion, with calendar-year spending pointed toward roughly $190 billion. CFO Amy Hood disclosed that roughly two-thirds of CapEx is for short-lived assets, primarily GPUs and CPUs. That mix is what worries Chang: the depreciation clock on AI silicon runs fast, and revenue conversion lags the spend. <\/p>\n Microsoft’s Q3 FY26 CapEx of $30.88 billion was up 84.39% year over year, funding an AI business now at a $37 billion annual run rate. Alphabet (NASDAQ: GOOGL<\/a>) has guided 2026 CapEx to $180 billion to $190 billion, with Q1 spend alone hitting $35.67 billion, up 107%. Amazon (NASDAQ: AMZN<\/a>) plans roughly $200 billion in 2026 CapEx, a figure that has already crushed trailing free cash flow to $1.2 billion, a 95% drop. <\/p>\n Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks \u2014 and Microsoft didn’t make the cut. Grab the names FREE today<\/a>. <\/p>\n The Moat Question <\/p>\n Chang’s sharpest point targeted competitive dynamics. With SpaceX entering hyperscaler infrastructure and Meta reportedly weighing the same, he argued that “the moat is not that strong.” That view was echoed in retail forums. A Reddit thread titled “Nikkei Investigation Finds $1.65 Trillion In Off-Balance-Sheet Commitments Across Five Major Tech Companies” received 165 upvotes as sentiment around Alphabet turned bearish. Meanwhile, sentiment toward MSFT on r\/investing swung from 82 (very bullish) to 30 (bearish) over four days, with the top thread arguing that “AI infrastructure depreciates way faster than people realise, and enterprise adoption is softening.” <\/p>\n The GPU and Memory Angle <\/p>\n NVIDIA (NASDAQ: NVDA<\/a>) NVIDIA (NASDAQ: NVDA) is the direct beneficiary. Q1 FY27 revenue reached $81.62 billion, accompanied by $119 billion in supply-related commitments. Jensen Huang called the AI buildout “the largest infrastructure expansion in human history.” Chang argued that investors should watch for “aggressive double, triple ordering” as an early warning that demand is being pulled forward. <\/p>\n Micron Technology (NASDAQ: MU<\/a>) sits at the other end of the AI supply chain. Fiscal Q3 2026 revenue surged 345.72% year over year,\u00a0with non-GAAP gross margin reaching 84.9%, according to Micron’s SEC filing. Chang’s co-panelist raised a subtler concern: chips fabricated in the United States could carry lower margins than those produced overseas, a structural headwind investors may not be pricing in given Micron’s 240.36% year-to-date gain. Shares trade at a forward P\/E of 5, appearing inexpensive on paper, but that valuation depends on the company sustaining its guided 86% gross margin. <\/p>\n What to Watch <\/p>\n Rockefeller’s CIO thinks that a more measured CapEx outlook from any single hyperscaler could trigger a rotation back into lagging stocks as expectations for free cash flow improve. Keep an eye on fiscal-year CapEx guidance, not the headline earnings beats. <\/p>\n Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks \u2014 and Microsoft didn’t make the cut. Grab the names FREE today<\/a>. <\/p>\n<\/a> Quality Stock Arts \/ Shutterstock.com <\/p>\n
<\/a> MSFT Analyst Ratings \u2014 24\/7 Wall St. <\/p>\n