AI冲击之下，新一轮“次贷危机”来了？

Core Viewpoint - The impact of AI is causing turmoil in the credit bond market, with risks reminiscent of the subprime mortgage crisis emerging, particularly affecting the software industry and leading to significant declines in leveraged loan indices and CLO assets [1][3][5] Group 1: Market Dynamics - Concerns over AI disruption are rapidly spreading to the global credit bond market, leading to asset sell-offs in leveraged loans and CLOs, raising investor awareness of systemic credit cycle risks [3][5] - The Bloomberg index indicates that the yield premium on comparable global debt has widened by nearly 4 basis points, marking the largest increase since November of the previous year [3] - Investment-grade bonds, once considered safe havens, are showing signs of pressure, with the spread of tech investment-grade bonds rapidly expanding [7][10] Group 2: Sector-Specific Impacts - The software and services sector has seen a significant decline in leveraged loans, contributing to a 1.34% drop in the Bloomberg U.S. leveraged loan index, the largest monthly decline since September 2022 [5][18] - Approximately 14% of the investment-grade index is now comprised of AI-related companies, with related debt ballooning to $1.2 trillion, surpassing the U.S. banking sector [7][9] - The high-yield bond market is under pressure, with U.S. junk bond funds experiencing continuous outflows in recent weeks [10][12] Group 3: Systemic Risk Concerns - UBS warns that the interconnectedness between private credit and leveraged loan markets poses a systemic contagion risk, particularly as borrowers frequently engage in dual financing [13][15] - The top 20 direct lending institutions dominate private credit asset management and hold significant positions in leveraged loans and high-yield bonds, indicating a potential rapid spread of default risks if the software sector is hit hard [15][17] - The refinancing risks in the software sector are escalating, with approximately $51 billion of software debt rated B- or below maturing by 2028, and another $50 billion by 2029 [19][20] Group 4: CLO and Asset Management - The underlying asset risks of leveraged loans are directly impacting structured products, with CLO managers urgently reassessing their exposure to AI-related industries [20] - Morgan Stanley estimates that between $40 billion and $150 billion of CLO assets are at risk due to their association with industries vulnerable to AI disruption [20]