AI正凶猛“加杠杆”

Core Insights - The AI revolution is increasingly being financed through significant debt accumulation, raising concerns about potential bubble risks in the market [1][2] - In September and October alone, AI-focused tech giants issued $75 billion in investment-grade bonds, more than double the average annual issuance from 2015 to 2024 [1][2] - The financing landscape has expanded beyond traditional investment-grade bonds to include high-yield debt, private credit, and structured financial products, prompting warnings from observers about accumulating risks [1][2] Debt Market Dynamics - The $75 billion in bond issuance represents only 5% of the total $1.5 trillion in U.S. investment-grade bond issuance this year, but its growth is notable [2] - Major issuers include Meta with $30 billion and Oracle with $18 billion, while Alphabet announced new borrowing plans [2] - AI-related companies now account for 14% of the weight in investment-grade indices tracked by JP Morgan, surpassing the U.S. banking sector [2] High-Risk Financing Channels - The demand for AI data centers is pushing capital into higher-risk areas, such as the high-yield debt market, with notable issuances like TeraWulf's $3.2 billion and CoreWeave's $2 billion bonds [8][11] - The private credit market is also rapidly growing, with UBS estimating that AI-related private credit loans could nearly double in the next 12 months [11] - Morgan Stanley predicts that private credit could provide over half of the funding for global data center construction by 2028, representing an $800 billion investment opportunity [11] Structured Financial Products - Structured products like asset-backed securities (ABS) are being repurposed to support the growth of the AI industry, with a significant portion of these products backed by future cash flows from data center leases [14] - The ABS market for digital infrastructure has expanded over eight times in less than five years, with data centers supporting 64% of transactions [14] - Despite their potential, ABS products are viewed cautiously due to their role in the 2008 financial crisis, raising concerns about underlying asset quality [14]