📰 This article attributes the weakening of U.S. long-end Treasury yields to a somewhat counterintuitive adversary: AI giants. On September 24, the yield on the U.S. 30-year Treasury note rose to 5.48%, a two-decade high; the 10-year yield also broke above 5.2%, the highest since the financial crisis.
🔥 As of August this year, the combined debt issuance of five ultra-large cloud providers is close to $230 billion—more than double the full-year figure last year—and a large portion is concentrated in 20-, 30-, and even 40-year maturities. Meta issued one 40-year bond totaling $30 billion. Alphabet has also issued a century-maturity sterling bond, and Amazon raised $54 billion in a single financing.
To be honest, these companies don’t seem to care much about interest rates, because they believe the future returns from AI will be high enough to cover today’s borrowing costs. The problem is that long-duration capital is limited. Pension funds, insurance companies, and bond funds buying more corporate tech debt naturally means they may reduce their allocations to U.S. Treasuries. Nomura estimates that long-term funding borrowed by tech giants is about $200 billion—roughly one quarter of the Treasury Department’s total annual issuance of intermediate- and long-term Treasuries.
💡 Even more troublesome: even if AI succeeds, it may not directly rescue the U.S. fiscal situation. If AI improves efficiency, corporate profits may rise, but the total wage bill could fall. Meanwhile, individual income taxes and social security taxes make up more than three-quarters of federal revenue. Data disclosed by Microsoft shows that current tax expense dropped from $14.1 billion in the same period last year to $2.5 billion; over the first 11 months of the fiscal year, the company’s income tax revenue decreased by one quarter year over year.
🤔 On one side, revenue falls while spending grows. On the other, total U.S. Treasury debt has crossed $40 trillion, and annual interest payments are nearing $1 trillion. So the question is: will AI first help the U.S. economy repay its debts, or will it first force the government to impose heavier taxes on AI profits?
#AI #美债 #美国财政 #Technology bonds
🔥 As of August this year, the combined debt issuance of five ultra-large cloud providers is close to $230 billion—more than double the full-year figure last year—and a large portion is concentrated in 20-, 30-, and even 40-year maturities. Meta issued one 40-year bond totaling $30 billion. Alphabet has also issued a century-maturity sterling bond, and Amazon raised $54 billion in a single financing.
To be honest, these companies don’t seem to care much about interest rates, because they believe the future returns from AI will be high enough to cover today’s borrowing costs. The problem is that long-duration capital is limited. Pension funds, insurance companies, and bond funds buying more corporate tech debt naturally means they may reduce their allocations to U.S. Treasuries. Nomura estimates that long-term funding borrowed by tech giants is about $200 billion—roughly one quarter of the Treasury Department’s total annual issuance of intermediate- and long-term Treasuries.
💡 Even more troublesome: even if AI succeeds, it may not directly rescue the U.S. fiscal situation. If AI improves efficiency, corporate profits may rise, but the total wage bill could fall. Meanwhile, individual income taxes and social security taxes make up more than three-quarters of federal revenue. Data disclosed by Microsoft shows that current tax expense dropped from $14.1 billion in the same period last year to $2.5 billion; over the first 11 months of the fiscal year, the company’s income tax revenue decreased by one quarter year over year.
🤔 On one side, revenue falls while spending grows. On the other, total U.S. Treasury debt has crossed $40 trillion, and annual interest payments are nearing $1 trillion. So the question is: will AI first help the U.S. economy repay its debts, or will it first force the government to impose heavier taxes on AI profits?
#AI #美债 #美国财政 #Technology bonds
