MIT Technology Review warns that hyperscalers need to nearly triple their productivity by 2030 to break even on their trillion-dollar infrastructure investment. According to estimates developed by Jessica Wachter, a finance professor at the Wharton School, hyperscalers, including Alphabet, Microsoft, Amazon, Meta, and Oracle, will spend nearly $1.1 trillion on data centers by 2027. This raises questions not only about the future of these companies but also about the economy as a whole. Wachter, who previously served as chief economist at the Securities and Exchange Commission (SEC), noted that hyperscalers' productivity must increase by 2.7 times to reach the break-even point by 2030. This calculation takes into account the cost of capital, an expected return of 15%, and asset depreciation. Without such growth, Wachter and her co-author arrive at a stark conclusion: "The current build-out will become the largest capital misallocation in history."
Alphabet announced a free cash flow deficit of $5.9 billion in the last quarter, marking the first occurrence since the company went public in 2004. Investors are increasingly concerned about the risks of AI spending as the concentration of funds grows. According to Morgan Stanley, hyperscalers are set to finance more than half of their planned $2.9 trillion on data centers by 2028. These funds are increasingly coming from external capital rather than internal reserves. In Louisiana, Meta transferred an 80% stake in its Hyperion data center to financial firm Blue Owl Capital, demonstrating how challenging hyperscaler financing has become.
Stijn Van Nieuwerburgh from Columbia Business School warns that such debt is increasingly penetrating through pension funds and private credit instruments. He emphasizes that many people do not realize how deeply this exposure has spread to their own retirement savings and insurance. Crypto strategist Arthur Hayes has also proposed a similar scenario, arguing that a potential credit crash in the AI sector could force the Federal Reserve to print money, which would drive the price of Bitcoin (BTC) to $1 million. Gary Gensler, former SEC chairman and MIT Sloan professor, expects a downturn, but the timeline for this process remains uncertain.
Related articles
