Nvidia customers face receivables and credit risk
Nvidia’s blistering AI growth is masking a quieter risk: its customers are taking on more receivables to keep the spending cycle alive, and that leaves the semiconductor boom more exposed to a single weak link than the market may appreciate.
That matters because the AI buildout is no longer just a story about demand for chips; it is becoming a question of balance-sheet strain across the supply chain. When orders are financed through stretching payment terms and rising customer credit, revenue can stay strong for a while even as the quality of that revenue deteriorates. If one large buyer slows purchases, delays payment or tightens capital spending, the effect can cascade quickly through suppliers, foundries and equipment makers.
The latest price action shows investors are still willing to pay up for the theme, but are also becoming less forgiving of any wobble. Nvidia’s shares closed at $218.29 on Sept. 11, below their 50-day moving average of $212.36? actually still above it, with RSI readings around 52.6 and a flattening MACD, suggesting momentum has cooled from earlier highs near $235.20 in May. Taiwan Semiconductor Manufacturing Co., Nvidia’s most important manufacturing counterpart, has also pulled back from recent peaks, while ASML has slipped below its own 50-day trend after a strong run, a sign the market is reassessing the durability of the spending surge even as AI demand remains intact.
The broader supply-chain data still point to extraordinary activity. TSMC reported August revenue of NT$514.8 billion, underscoring how much AI-related demand is flowing into advanced nodes. Micron has rewarded employees in Taiwan with larger bonuses, another sign that capacity is tight and competition for talent remains intense. ASML continues to win orders for its high-end lithography tools from Samsung, TSMC and Intel. In other words, the industrial side of AI is still expanding.
But that is exactly why receivables matter. In a normal cycle, a strong order book is enough. In this one, Nvidia’s customers are often hyperscalers, cloud providers and data-center operators making multibillion-dollar commitments up front. That can support a long runway of sales, yet it also means Nvidia’s earnings are increasingly tied to the health of a handful of giant buyers rather than a broad customer base. If those customers become more selective on capital spending, the effect would hit Nvidia first through slower revenue growth and then ripple outward to TSMC, memory suppliers and equipment makers.
For investors, the bull case is still straightforward: AI compute demand remains real, supply remains constrained, and Nvidia’s earnings power is still unmatched. The bear case is subtler and more dangerous. A business that looks nearly invulnerable while growth is being financed through a tight circle of customers and suppliers can re-rate quickly once credit conditions, inventory levels or spending priorities shift. The risk is not that AI demand disappears; it is that the financing structure behind it becomes less stable than the market has priced in.
That leaves Nvidia as both the clearest winner of the AI boom and the stock most exposed to a change in its plumbing. As long as the customer base keeps spending, the story stays intact. If receivables keep building and one major buyer blinks, the market could discover that the foundation of the AI trade is narrower than it looks.
| Entity | Gains | Losses |
|---|---|---|
| Nvidia | ▲AI revenue growth | ▼Credit and receivables risk |
| Hyperscale buyers | ▲Immediate AI capacity | ▼Cash flow flexibility |
| TSMC | ▲Advanced-chip demand | ▼Any capex slowdown |
| ASML / Micron | ▲Equipment and memory orders | ▼Cycle reversal risk |