AI fears add a 200-basis-point premium to SpaceX’s debt
When SpaceX issued its first corporate bonds, it paid 6.65% for the most expensive piece. The most expensive slice of the $40 billion of private credit borrowing it’s finalized in recent days will cost as much as 8.75%.
Wall Street is gauging fear through numerical measures such as the VIX and credit-default spreads. When SpaceX introduced its first corporate bonds four months ago, it was charged 6.65% for the most expensive portion. Currently, the most expensive slice of the $40 billion in private credit borrowing finalized by the company could reach 8.75%, according to a source familiar with the matter.
This widening gap can be attributed to at least 25 basis points due to the Fed's September rate hike and expectations of further hikes, along with the remaining 175 basis points stemming from pure fear. This 200 basis point premium translates into millions of dollars in extra borrowing costs that cannot be attributed solely to the Fed or a more cautious lending environment.
The blame lies squarely on growing doubts about the return on investment (ROI) of AI and the question of borrowers' ability to repay the borrowed funds. Despite Elon Musk's assertion that the deal showcased SpaceX's strength, emphasizing its "investment grade" status and revenue from Starlink and launch services, the market continues to price it as if a BBB-rated company were two notches lower.
Additionally, the supply-demand imbalance in the market, where AI requires substantial funding while investors seek to catch their breath, exacerbates the situation. For now, AI anxiety predominantly affects credit investors, who are typically more cautious and pessimistic compared to stockholders. However, if this fear spreads to stockholders, the consequences could be swift and severe.
Financial experts warn that the credit market is struggling to find effective ways to offload AI-related risks, fearing that the rising tide of AI may eventually sink some boats.
Written by urgent.news from Semafor's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.