Workflow
AI spending war
icon
Search documents
AI-Spending War and AI-Debt Pile-Up Could Squeeze Share Buybacks
Wolfstreet· 2025-11-24 23:22
Core Insights - Major tech companies have spent a total of $1.1 trillion on share buybacks over the past five years, significantly impacting stock prices [1][2] - The trend of share buybacks may be at risk as companies shift focus towards capital expenditures, particularly in AI infrastructure [10][11] Share Buyback Spending - Apple led the share buyback spending with $437 billion, followed by Alphabet ($281 billion), Meta ($151 billion), Microsoft ($107 billion), and Nvidia ($87 billion) [3] - Nvidia has recently ramped up its buyback program, spending $43 billion over the past four quarters [3] Funding Sources for Buybacks - Some share buybacks were financed through debt issuance, with Apple holding $112 billion, Microsoft $120 billion, Meta $50 billion, and Alphabet $30 billion in debt [4] - In the last three months, five companies issued $88 billion in new investment-grade bonds to fund various expenditures [5] Shift in Capital Expenditures - Companies are increasingly investing in AI-related capital expenditures, with four major firms spending $114 billion in Q3 alone and projected to exceed $400 billion for the year [4] - This shift in focus may lead to reduced share buybacks as companies prioritize AI investments over returning cash to shareholders [11][12] Off-Balance-Sheet Debt Concerns - Meta has utilized a strategy to keep a $27 billion AI bond sale off its balance sheet to protect its credit rating [7] - Rating agencies are expressing concerns about the opacity of off-balance-sheet debt and its implications for financial transparency [8] Future of Share Buybacks - Companies may eventually reduce or halt share buybacks as they allocate more resources to AI spending, which could impact stock prices negatively [12][13] - Amazon has already ceased share buybacks in favor of capital expenditures, indicating a potential trend among other companies [10]
Meta Losing AI Spending War To Google, Microsoft, Warns 'Big Short' Investor Steve Eisman - Meta Platforms (NASDAQ:META)
Benzinga· 2025-11-06 08:12
Core Viewpoint - Steve Eisman warns that Meta Platforms Inc. is losing the AI competition against Alphabet Inc. and Microsoft Corp., stating that Meta cannot sustain the industry's high capital expenditures [1][2]. Group 1: Market Reaction and Earnings - Despite all three companies (Meta, Google, and Microsoft) beating revenue and earnings estimates, Meta's stock fell while Google's stock rose significantly [2]. - Eisman attributes the market's reaction to the scrutiny of who can afford the massive AI capital expenditures, rather than common press explanations [2]. Group 2: Capital Expenditure Comparison - Meta's capital expenditure is in the "low 70 billion" range, while Microsoft is at $80 billion and Google exceeds $90 billion [3]. - The key difference is that Google and Microsoft have substantial cloud businesses generating revenue from their AI investments, whereas Meta lacks a cloud business and its AI spending is solely for new product development [4]. Group 3: Financial Health and Cash Reserves - Meta's cash and cash equivalents decreased by 43% from $77.8 billion at the end of 2024 to $44.4 billion in 2025, contrasting with the increasing cash reserves of Google and Microsoft during the same period [4]. - Eisman concludes that Google and Microsoft are better positioned to handle the financial burden of AI investments compared to Meta [5]. Group 4: AI Boom Outlook - The significant capital expenditures from all three companies indicate that the AI boom is ongoing, countering any notions that the AI narrative is nearing its end [5]. Group 5: Stock Performance - Meta's stock is up 6.13% year-to-date and 11.17% over the year, but it maintains a weaker price trend compared to Microsoft and Google [6]. - Microsoft has increased by 21.16% year-to-date and 20.70% over the year, while Google has surged 49.37% year-to-date and 59.68% over the year, showing stronger price trends across all time frames [7].