Capitalizing on AI: The $196 Billion Investment Surge and Its Financial Implications

AI

Hi everyone!

As the dust settles after the earnings reports of the biggest AI players, I wanted to take a deeper look at some key trends emerging from their capital expenditures (CapEx) and investment strategies. Companies like Microsoft, Google, Meta, and Amazon are making significant investments in AI infrastructure, and it’s worth examining the financial implications behind these moves.

AI Capital Expenditures: Growing at an Unprecedented Pace

Over the past 12 months, four of the biggest names in AI—Microsoft (MSFT), Google (GOOG), Meta (META), and Amazon (AMZN)—have collectively invested $196 billion in capital expenditures to fuel their AI ambitions. This surge in CapEx growth has been accelerating, with a 10% compounded quarterly growth rate since Q1 2023. Microsoft is leading the charge with a 14% quarterly growth rate, while Meta has the slowest at 4.4%.

One thing is clear: AI infrastructure is critical for these companies’ long-term growth. In fact, this heavy investment in capital appears to be a strategic hedge against the risk of under-investing. As Sundar Pichai, CEO of Alphabet, pointed out during Alphabet’s Q2 2024 earnings call:

“The risk of under-investing is dramatically greater than the risk of over-investing. Even if we end up over-investing, these are assets with long lifespans and widespread utility. The downside of not investing, however, is much greater.”

While this rationale might seem conservative, it reflects a broader trend where the leading tech giants are front-loading their AI investments to secure a competitive edge in what they see as a long-term infrastructure race.

CapEx Growth vs. Sales Growth: A Sign of AI’s Growing Burden

Interestingly, while these companies are investing heavily in AI infrastructure, not all are seeing the same return on investment. A closer look at the relationship between CapEx growth and sales growth reveals some important differences. Across the group, CapEx is growing faster than sales, a sign that these companies are prioritizing long-term capacity over short-term profitability. On average, CapEx is growing 10% per quarter, while sales growth across these companies is just 4%.

Meta Platforms stands out as the only exception. Meta is growing its sales at a 6% quarterly rate, which outpaces its 4% quarterly CapEx growth. This reflects a more balanced approach, where the company’s CapEx is more aligned with its revenue generation.

Altogether, they are growing their CapEx at a 10% rate, compared to a 4% growth in total sales, which somewhat verifies that most of them are thinking in the same way.

Nevertheless, all the comments made by the management teams of these companies during their latest earnings calls indicate that this trend won’t stop anytime soon, and may even accelerate.

The Ricardian Effect of AI Capital

A lesser-known but fascinating economic principle comes into play here—the Ricardian effect, originally proposed by economist David Ricardo. Ricardo argued that if a society increases or decreases its savings rate while wages/rents remain unchanged, it directly affects the goods and services provided within that economy.

If we use the same example with companies, it works the same way. If all the companies decide to increase or decrease their free cash flow rate while their sales remain the same, it would have an immediate effect on the factors of production (or capital goods).

In the case of these AI companies, we can apply this concept to understand the growing demand for capital goods (such as GPUs) and its broader market implications. These companies are decreasing their free cash flow (FCF) margins due to that capital into AI-related infrastructure. As a result, we’re seeing a significant increase in the CapEx-to-sales ratio, which has increased 4.7% to 16.5% since Q1 2023.