Morgan Stanley analysts recommend buying shares in AI infrastructure companies following a market dip that occurred throughout July [1].
This guidance comes as investors weigh whether the massive spending on artificial intelligence is sustainable or if the market has reached a peak. A correction in these stocks could signal a shift in sentiment toward the tech sector's long-term growth.
Shares of companies involved in AI infrastructure fell by an average of nearly seven% in July [2]. Despite this decline, Morgan Stanley analysts said that the current weakness does not reflect a change in the underlying value of these businesses.
Stephen Byrd, an analyst at Morgan Stanley, said, "We believe a meaningful driver of weakness has been technical rather than fundamental" [1]. The firm said that the pullback provides a sizable upside opportunity for those willing to enter the market now.
As part of this recommendation, Morgan Stanley highlighted 13 AI power-infrastructure stocks [1]. These companies focus on the energy and storage bottlenecks that currently limit the expansion of AI data centers.
While Morgan Stanley views the dip as a buying opportunity, other market observers suggest a more cautious approach. Some reports indicate that buyers may need stronger stomachs for the current dip in chip ETFs, suggesting that the risk profile of these assets has increased [1].
Analysts at Morgan Stanley said that the demand for the physical infrastructure required to run AI, including power grids and cooling systems, remains robust. They said that the technical nature of the July decline creates a strategic entry point for long-term investors [1, 2].
“"We believe a meaningful driver of weakness has been technical rather than fundamental."”
This recommendation reflects a growing divide between those who view AI stocks as overvalued and those who believe the physical requirements of the technology—such as electricity and hardware—create a floor for the market. By focusing on power infrastructure rather than just software, Morgan Stanley is pivoting toward the 'picks and shovels' of the AI era, suggesting that the bottleneck for AI growth is now physical energy capacity rather than algorithmic capability.
