AI狂潮下的债务杠杆——下一场金融风暴导火索？

Group 1 - The core trend driving the demand for debt in AI infrastructure is the significant increase in capital expenditures by tech giants, exemplified by Amazon's capital spending surging 75% year-over-year, nearing the company's operating cash flow levels [1] - Traditional equity financing is becoming insufficient to meet the capital needs, leading companies to turn to the bond market and private credit for support, as evidenced by Amazon's recent $15 billion bond issuance attracting $80 billion in subscriptions [1][2] - The strong demand for AI-related debt instruments is being fueled by institutional investors such as pension funds and insurance companies, which are seeking higher yields in a low-interest-rate environment [3] Group 2 - The total bond issuance by the five major U.S. cloud computing companies has reached an astonishing $121 billion this year, significantly higher than the average of $28 billion over the past five years, indicating a substantial impact on the market [2] - It is projected that AI infrastructure projects will require approximately $800 billion in private credit funding between 2025 and 2028, representing one-third of the expected total infrastructure investment in the sector during that period [3] - The financing structure poses a systemic risk due to its broad distribution across the financial system, as debts are held by various institutional investors, which could lead to widespread repercussions if the AI sector underperforms or experiences defaults [4][5]