AI’s Investment Boom Meets the Discipline of Credit Markets

SpaceX shares trading below their IPO price has attracted attention, but the more instructive signal may be found in its debt. The company recently completed a US$25 billion, five-tranche bond offering. Its 2056 bond has subsequently weakened, with its yield moving towards 7.5%. Meanwhile, Goldman Sachs’ hyperscaler credit basket has widened to its broadest level since its launch in February. Together, these suggest that 𝗶𝗻𝘃𝗲𝘀𝘁𝗼𝗿𝘀 𝗮𝗿𝗲 𝗯𝗲𝗴𝗶𝗻𝗻𝗶𝗻𝗴 𝘁𝗼 𝗿𝗲𝗮𝘀𝘀𝗲𝘀𝘀 𝘁𝗵𝗲 𝗰𝗼𝘀𝘁 𝗼𝗳 𝗳𝗶𝗻𝗮𝗻𝗰𝗶𝗻𝗴 𝘁𝗵𝗲 𝘄𝗶𝗱𝗲𝗿 𝗔𝗜 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 𝗯𝘂𝗶𝗹𝗱-𝗼𝘂𝘁.

This is exactly why we have remained selective on AI/tech lending (as highlighted in an earlier post). Three considerations stand out:

𝗙𝗶𝗿𝘀𝘁, 𝗰𝗮𝗽𝗶𝘁𝗮𝗹 𝗶𝗻𝘁𝗲𝗻𝘀𝗶𝘁𝘆. Data centres, power and compute capacity require substantial upfront investment. As financing costs rise, so does the hurdle rate, placing greater scrutiny on expected returns.

𝗦𝗲𝗰𝗼𝗻𝗱, 𝗳𝗿𝗲𝗲 𝗰𝗮𝘀𝗵 𝗳𝗹𝗼𝘄. Higher interest expense reduces cash available for reinvestment and balance-sheet flexibility, weakening the cushion available to absorb execution delays, cost overruns or slower revenue growth.

𝗧𝗵𝗶𝗿𝗱, 𝗰𝗼𝗹𝗹𝗮𝘁𝗲𝗿𝗮𝗹 𝗾𝘂𝗮𝗹𝗶𝘁𝘆. A significant portion of technology-related capital expenditure is invested in hardware that depreciates rapidly, alongside infrastructure whose value may depend on future utilisation and revenue assumptions. Where performance falls short, recovery values may be less certain than in businesses supported by tangible assets, contracted cash flows or established asset values.

Our approach to private credit remains anchored in 𝗰𝗼𝗹𝗹𝗮𝘁𝗲𝗿𝗮𝗹 𝗰𝗹𝗮𝗿𝗶𝘁𝘆, 𝗱𝗼𝘄𝗻𝘀𝗶𝗱𝗲 𝗽𝗿𝗼𝘁𝗲𝗰𝘁𝗶𝗼𝗻 and 𝘀𝘁𝗮𝗯𝗹𝗲 𝗳𝗿𝗲𝗲 𝗰𝗮𝘀𝗵 𝗳𝗹𝗼𝘄, not simply on growth narratives or assuming that capital will remain readily available at favourable terms. Businesses without an adequate margin of safety are often the first to come under pressure when financing conditions tighten. And lenders with limited collateral protection may bear a disproportionate share of that risk.

hashtag#PrivateCredit hashtag#CreditInvesting hashtag#ArtificialIntelligence hashtag#TechLending 

𝘛𝘩𝘦 𝘪𝘯𝘧𝘰𝘳𝘮𝘢𝘵𝘪𝘰𝘯 𝘱𝘳𝘰𝘷𝘪𝘥𝘦𝘥 𝘪𝘯 𝘵𝘩𝘪𝘴 𝘱𝘰𝘴𝘵 𝘪𝘴 𝘧𝘰𝘳 𝘪𝘯𝘧𝘰𝘳𝘮𝘢𝘵𝘪𝘰𝘯𝘢𝘭 𝘱𝘶𝘳𝘱𝘰𝘴𝘦𝘴 𝘰𝘯𝘭𝘺 𝘢𝘯𝘥 𝘴𝘩𝘰𝘶𝘭𝘥 𝘯𝘰𝘵 𝘣𝘦 𝘤𝘰𝘯𝘴𝘵𝘳𝘶𝘦𝘥 𝘢𝘴 𝘧𝘪𝘯𝘢𝘯𝘤𝘪𝘢𝘭, 𝘪𝘯𝘷𝘦𝘴𝘵𝘮𝘦𝘯𝘵, 𝘰𝘳 𝘱𝘳𝘰𝘧𝘦𝘴𝘴𝘪𝘰𝘯𝘢𝘭 𝘢𝘥𝘷𝘪𝘤𝘦. 𝘈𝘭𝘸𝘢𝘺𝘴 𝘤𝘰𝘯𝘥𝘶𝘤𝘵 𝘺𝘰𝘶𝘳 𝘰𝘸𝘯 𝘳𝘦𝘴𝘦𝘢𝘳𝘤𝘩 𝘢𝘯𝘥 𝘤𝘰𝘯𝘴𝘶𝘭𝘵 𝘸𝘪𝘵𝘩 𝘢 𝘲𝘶𝘢𝘭𝘪𝘧𝘪𝘦𝘥 𝘧𝘪𝘯𝘢𝘯𝘤𝘪𝘢𝘭 𝘢𝘥𝘷𝘪𝘴𝘰𝘳 𝘣𝘦𝘧𝘰𝘳𝘦 𝘮𝘢𝘬𝘪𝘯𝘨 𝘢𝘯𝘺 𝘪𝘯𝘷𝘦𝘴𝘵𝘮𝘦𝘯𝘵 𝘥𝘦𝘤𝘪𝘴𝘪𝘰𝘯𝘴. 𝘗𝘢𝘴𝘵 𝘱𝘦𝘳𝘧𝘰𝘳𝘮𝘢𝘯𝘤𝘦 𝘪𝘴 𝘯𝘰𝘵 𝘪𝘯𝘥𝘪𝘤𝘢𝘵𝘪𝘷𝘦 𝘰𝘧 𝘧𝘶𝘵𝘶𝘳𝘦 𝘳𝘦𝘴𝘶𝘭𝘵𝘴.