A $1 Trillion Depreciation Bomb Is Ticking Beneath the AI Market

Key takeaways

• AI earnings growth is largely an accounting illusion; the real cash‑flow strain could erode profits once depreciation and debt costs kick in.
• U.S. corporate profits $4.7 trillion (12 months through mid‑year)
• 17% increase in profits
• 42 AI firms drove ~66% of S&P 500 gains

What happened

U.S. corporate profits climbed to a record $4.7 trillion in the 12 months through mid‑year, up 17%. Only 42 AI‑related companies have driven about two‑thirds of the S&P 500’s gains since early 2024. Analysts argue that the surge is largely an accounting artifact: Nvidia and other chip makers book full profits when they sell GPUs, while buyers depreciate the chips over roughly six years, creating a $1 trillion depreciation “mirage” that won’t hit earnings until data centers are online. Tech…

Why it matters

The apparent earnings boom masks a deepening cash‑flow gap: hyperscalers are spending almost all operating cash on capex, issuing $225 billion of unsecured bonds and holding $1.1 trillion in off‑balance‑sheet lease commitments. Morgan Stanley projects negative free cash flow of about $550 billion next year, and the…

Short-term and longer-term

Short-term. Over the next few weeks, watch for any signs of tightening credit or a slowdown in data‑center deployment, as these could trigger a sharp drop in reported earnings.

Longer-term. For the coming months, the key risk is that the $1 trillion depreciation wall and negative free cash flow could erode the high earnings multiples that have driven the AI rally.