Dwindling cash and soaring memory costs: Tech's AI buildout has ballooning price tag

Four years into the AI boom, the world's largest tech firms are still splashing out on promises of a tech future. However, their spending is draining their cash reserves. Goldman Sachs projects AI spending among the megacaps to soar to $765 billion this year and nearly $1.2 trillion by 2027. Amazon has hiked its capital spending forecast for the year to $220 billion, the highest among the four hyperscalers.

Meanwhile, Amazon reported a negative $7.6 billion free cash flow for the trailing 12 months, and Meta disclosed a 91% drop in cash generation from a year earlier. Alphabet turned cash flow negative for the first time on record, a surprising turn for one of the most profitable companies. As the tech earnings season concludes, it's evident that AI investments are skewing balance sheets.

The memory crunch caused by the relentless demand for AI processors, which rely on memory from a few vendors, is a major factor behind the rising costs. Apple, the most budget-conscious of its peers, is facing supply constraints, and may hike prices on its devices. While memory prices are described as 'insane' by Tesla CEO Elon Musk, Amazon CEO Andy Jassy attributes his company's higher capex guidance to the 'inflated price' of memory chips.

Apple's CEO Tim Cook expects the memory crisis to continue this year, warning of a potential impact on the business if memory prices keep rising. For hyperscalers, soaring memory prices pose a significant cost hurdle. Investor reactions to the reports have been mixed, with Tesla, Alphabet, and Meta's shares dropping after reporting negative cash flow and weak forecasts.

In contrast, Microsoft's stock rose after delivering better-than-expected results and increasing capex guidance. Analysts suggest that the massive AI buildout, fueled by debt, may not ultimately yield positive returns, as skepticism grows over whether the investments will pay off.