Following last week’s discussion around more selective AI leadership, this week showed how quickly sentiment can shift even when the long-term AI story remains intact. The biggest pressure point came from reports that Meta may launch a cloud computing business by selling access to excess AI computing capacity. While this was positive for Meta, it raised a broader question for the market
If large technology companies have enough compute to resell, could hyperscaler capex eventually slow?
That concern quickly spread across semiconductors, optical networking, memory, and data center infrastructure stocks. Thursday’s selloff was not really about weak company fundamentals. Many of these companies still have strong revenue growth, healthy demand, and long-term AI tailwinds. Instead, the move appeared more technical and fund-flow driven. After a very strong first-half rally, positioning in AI infrastructure had become crowded, and the Meta headline gave investors a reason to take profits across the entire supply chain.
Optical networking stocks were hit especially hard because they had been among the strongest performers earlier this year. The pullback does not necessarily suggest demand is weakening. Faster data movement remains critical for AI data centers, but expectations had become very high, making the group more vulnerable to rotation.
The market is also becoming more focused on the return on AI spending. Earlier in the cycle, investors rewarded companies simply for increasing AI capex. Now, investors want clearer evidence that this spending can translate into sustainable revenue and earnings growth. This is a healthier but more demanding phase of the AI cycle.
Macro uncertainty also remains in the background. The June jobs report came in weaker than expected, but the Federal Reserve is still focused on inflation and price stability. That means softer labor data does not fully remove rate-hike risk, especially with valuations already elevated across growth sectors.
Looking ahead, the near-term outlook remains constructive but volatile. The AI infrastructure cycle has not ended, and large technology companies are unlikely to meaningfully reduce spending in the near term because competition for compute remains intense. However, after such a strong run, the market needs more earnings confirmation to support current valuations.
Overall, this week looks more like a technical reset than a fundamental breakdown. Long-term demand for semiconductors, memory, optical networking, and data center infrastructure remains strong, but investors are becoming more selective. In the near term, volatility and rotation may continue, but the broader AI investment theme remains intact.
If you are curious to learn how your portfolio is performing in the current scenario, review your portfolio through Quantel's
A comprehensive questionnaire that shows how your portfolio is performing.