Nvidia, Microsoft AI Trade Still Has Cycle Support
Investors may be growing more skeptical about the sustainability of the AI spending frenzy, but TS Lombard says the trade is not ready to go against the boom yet because the U.S. business cycle may still have 12 to 18 months left.
That matters because AI stocks have become one of the market’s biggest macro expressions of the cycle, with capital spending, cloud demand and semiconductor orders tied to the same investment wave. TS Lombard’s view suggests the near-term risk is less about an immediate bust than about how long growth, corporate spending and easy earnings comparisons can keep supporting the trade.
Nvidia and Microsoft underscore the scale of that setup. Nvidia shares were last at $218.29 after a sharp run-up this summer, while Microsoft closed at $495.63, still far above its 200-day moving average of $429.65. Both have been buoyed by the AI buildout, but the market has also become more selective: Nvidia’s recent pullback from above $227 comes as technical momentum cools, with its RSI at 52.6 and price still close to the upper half of its recent range.
The credit market is sending a similar message, though not yet a full alarm. The ICE BofA U.S. high-yield spread sits at 2.70 percentage points, well below stress levels, while the 10-year Treasury yield is around 4.949% and the 2-year at 4.568%, keeping financial conditions tight enough to matter but not tight enough to signal a recession is already underway. TS Lombard’s call rests on that gap between rising AI risk and an economy that still has room to run.
That helps explain why betting aggressively against the group has been a crowded and dangerous trade. Microsoft’s latest filing said AI infrastructure spending lifted cost of revenue, but it also highlighted continued gross-margin growth and strong Microsoft Cloud performance. Nvidia’s latest filing showed $36 billion of commitments tied to capacity and supply-chain demand, reinforcing that the industry is still in the middle of a capital-intensive expansion rather than a completed cycle.
For investors, the message is that AI names may be more vulnerable to valuation resets, margin pressure and positioning unwinds, but not necessarily to a synchronized collapse yet. The key risk now is timing: if the business cycle lasts another year or more, the short thesis can stay early even as volatility rises.
The next catalyst is whether growth data, rates and hyperscaler capex guidance confirm that the AI buildout still has runway — or start to show the kind of slowdown that would finally make the short case work.
| Entity | Gains | Losses |
|---|---|---|
| Nvidia longs | ▲Cycle still supports AI demand | ▼Pullback risk, valuation pressure |
| Microsoft shareholders | ▲AI capex still drives cloud growth | ▼Margin drag from infrastructure spending |
| AI shorts | ▲Volatility may rise | ▼Can be too early if cycle extends |
| Credit bulls | ▲No recession-style stress yet | ▼Higher yields keep conditions tight |