Salesforce $25B Bond Deal: Investors Demand Concessions & Buyback Funding

Salesforce is facing a tougher sell than anticipated for its planned $25 billion bond offering, according to reports from the Financial Times and other outlets. Investors are demanding significantly higher yields – the effective interest rate – than initially expected, signaling increased caution around corporate debt even for established tech giants. The company intends to use the proceeds primarily for stock buybacks, a move that is drawing scrutiny alongside the bond terms.

Debt Markets Reflect Rising Risk Aversion

The bond sale, which could be the third-largest corporate bond offering in history, comes at a time of heightened sensitivity in the debt markets. While Salesforce remains a profitable company, the combination of a large offering size and the intended use of funds for share repurchases – rather than growth initiatives – is giving investors pause. Yields on the proposed bonds have reportedly increased to around 5.25% for 10-year notes, a substantial jump from initial guidance. This increase reflects a demand for greater compensation for the risk associated with lending to Salesforce at this scale.

The situation highlights a broader trend of increased investor scrutiny of corporate debt. The Wall Street Journal noted that Salesforce’s bond sale is occurring alongside a generally busy day for corporate borrowing, but the increased yields demanded by investors suggest a shift in market sentiment. This isn’t isolated to Salesforce; companies across sectors are finding it more expensive to borrow money as economic uncertainty persists.

Buybacks and the Debt Burden

Salesforce’s decision to prioritize stock buybacks with the funds raised is a key factor influencing investor caution. Buybacks reduce the number of outstanding shares, which can boost earnings per share and potentially increase the stock price. Still, they don’t contribute to the company’s long-term growth or innovation. Some analysts and investors prefer to see companies invest in research and development, acquisitions, or other initiatives that drive future revenue. Using debt to fund buybacks is seen as particularly risky, as it increases the company’s financial leverage.

Bloomberg reported that Salesforce’s buyback plan is a significant driver of the bond offering. The company has been under pressure from activist investors to improve shareholder returns, and buybacks are a relatively quick way to achieve that goal. However, the debt-funded buyback strategy is raising concerns about the company’s long-term financial health.

AMI Labs Secures $1 Billion in Seed Funding

In a separate, but noteworthy development in the tech sector, Advanced Machine Intelligence (AMI) Labs, co-founded by former Meta chief AI scientist Yann LeCun, has raised $1.03 billion in a seed funding round. This is the largest seed round for a European company, according to Sherwood News. The funding values AMI Labs at $3.5 billion. LeCun is betting on “world models” – AI systems that understand how physical objects interact with their environment – as the next breakthrough in artificial intelligence, diverging from the current focus on large language models (LLMs). Investors in the round include Cathay Innovation, Greycroft, Hiro Capital, HV Capital, Bezos Expeditions, Nvidia, and Temasek. AMI Labs plans to establish offices in Paris, Montreal, Singapore, and New York.

The European AI Landscape

The substantial funding for AMI Labs underscores the growing ambition of the European Union to become a major player in the AI space. The EU is actively seeking to build a “Euro stack” of AI companies to reduce its reliance on American tech giants. AMI Labs, based in Paris, is poised to become a key component of this effort. The size of the seed round signals a strong vote of confidence in LeCun’s vision and the potential of world models. TechCrunch reports that the company’s approach differs fundamentally from many current AI startups, focusing on understanding the real world rather than simply processing language.

World Models vs. Large Language Models

LeCun has been a vocal critic of the limitations of LLMs, arguing that they lack a fundamental understanding of the physical world. World models, in contrast, aim to build AI systems that can reason about and interact with the environment in a more human-like way. This approach could have significant implications for applications in areas such as robotics, healthcare, and autonomous systems. AMI Labs’ first partnership will be with Nabla, a digital health startup, suggesting an initial focus on applying world models to medical applications where accuracy and reliability are paramount.

Salesforce: What’s Next

For Salesforce, the next few days will be critical. The company will need to assess investor demand and potentially adjust the terms of the bond offering – either by increasing the yield or reducing the size of the sale. A failed or significantly scaled-back bond offering could force Salesforce to reconsider its buyback plans or explore alternative financing options. The outcome will be closely watched by investors and analysts as a gauge of the current appetite for corporate debt and the market’s view of Salesforce’s financial strategy. The company’s second-quarter earnings report, expected in late May, will also be a key event, providing further insight into its financial performance and outlook. Wired highlights LeCun’s skepticism about the ability of LLMs to achieve human-level intelligence, further emphasizing the divergent paths being taken by different AI labs.