Last week, there was a flurry of earnings releases from "Big Tech" companies. Artificial intelligence is big business, and competition for many of them is stiff these days. The resulting spending spree, which has been coined "hyperscaling," is resulting in billions of dollars to buy semiconductor chips, build data centers, and develop the software to run AI.
In a recent edition of the investment newspaper, the Financial Times, an article questioned whether the "Big Tech borrowing spree raises fears over AI risks in US corporate bonds." It cited estimates of the scale of borrowing – Morgan Stanley estimates hyperscalers will raise around $400 billion in corporate bonds in 2026 to scale AI. And JPMorgan recently estimated that AI and data center firms account for 14.5% of its $10 trillion investment-grade bond index. That's nearly $1.5 trillion in existing debt.
Image source: Getty Images
This is a lot of spending, and if the returns on these massive investments don't turn out as planned, it could spell disaster for both these firms and the investors who bought their stocks.
Below is an overview of the debt levels of the key hyperscalers. This list of leading Big Tech firms includes Nvidia (NVDA), whose GPU and chips are seen at the forefront of providing the computing power to run AI. Oracle (ORCL) is also a primary supplier and one of the largest to lead the charge in building data centers and cloud computing capabilities. Alphabet (GOOG) (GOOGL), Apple (AAPL), Microsoft (MSFT), Meta (META), and Amazon (AMZN) offer a mix of cloud computing, as well as the ability to bring AI directly to consumers and businesses.
| Company | Ticker | Debt-to-Equity (D/E) % | Debt-to-Capital (D/C) % |
|---|---|---|---|
| Oracle | ORCL | 519.6 | 83.9 |
| Apple | AAPL | 152.4 | 60.4 |
| Amazon | AMZN | 43.4 | 30.3 |
| Microsoft | MSFT | 32.7 | 24.6 |
| Meta | META | 38.6 | 27.9 |
| Alphabet | GOOGL | 11.4 | 10.3 |
| Nvidia | NVDA | 9.1 | 8.3 |
| average | 115.31 | 35.10 |
Looking at the below chart, despite the many billions of dollars of spending going into scaling AI, most of these big tech bellwethers still have very solid balance sheets. Only Apple and Oracle owe more debt than they have cash on the balance sheet.
| Company | Ticker | Total Cash & ST Investments | Long-Term Debt |
|---|---|---|---|
| Alphabet | GOOGL | $98.5B | $21.6B |
| Nvidia | NVDA | $60.6B | $7.5B |
| Microsoft | MSFT | $89.5B | $35.4B |
| Amazon | AMZN | $94.2B | $57.9B |
| Meta | META | $81.6B | $58.7B |
| Apple | AAPL | $54.7B | $78.3B |
| Oracle | ORCL | $19.8B | $100.0B |
| source: KoyFin |
The debt-to-equity (D/E) and debt-to-capital (D/C) ratios are two primary measures of a firm's solvency and of whether debt levels are excessively high. Apple and Oracle are again the two standouts that warrant further attention. Apple's debt level is arguably risky, but it has steady demand through its subscriptions and service revenues. Its iPhones are also very popular and not seen as too economically sensitive, as consumers consider their mobile phones a necessity.
NASDAQ: AAPL
Key Data Points
On the other hand, Oracle's debt levels are much higher. They could arguably be too high, putting the firm (and shareholders) at risk if the returns it expects from AI do not materialize as planned.
NYSE: ORCL
Key Data Points
One last channel check is each firm's debt ratings. Each and every firm has an investment grade rating by S&P and Moody's – the two largest and most well-known credit rating agencies. And all, save Oracle, have A ratings or higher. Only Oracle is in the B range with a credit rating of BBB by S&P and Baa2 from Moody's. Again, still investment grade, but something to keep an eye on. At this point, it shouldn't be too much of a surprise that Oracle's debt is on a negative watch, meaning it could be downgraded at some point.
The Foolish Bottom Line
The above analysis sets a good baseline from which to judge the additional spending and debt raised for these big tech firms to continue to scale their AI ambitions. I am only worried about Oracle's debt levels, but it still could end up being a big winner as consumers and businesses increasingly turn to AI for research and building more efficient operations.
