Goldman lays out the next phases of AI trade and stocks to play them

Published Sun, Mar 17 20247:32 AM EDTUpdated Sun, Mar 17 20248:12 PM EDT

Jesse Pound @/in/jesse-pound @jesserpound

The next steps for the artificial intelligence boom are starting to come into shape, according to Goldman Sachs, and some of the stocks involved may still be cheap relative to their potential. Ryan Hammond of Goldman's portfolio strategy research team said in a March 14 note to clients that the AI trade still has room to run and should expand beyond the biggest winners like Nvidia , even if high valuations are making some investors nervous. "In addition to NVDA, investors have been focused on a broadening of the AI trade. We expect there will likely be three broad, subsequent stages of the AI trade," Hammond wrote. As the use cases around AI come into focus, new winners will emerge. Goldman foresees a second phase that focuses on companies that build and maintain the infrastructure around AI. The industries involved in the next step could include chipmakers, cloud providers, other technology firms and utilities, according to Hammond. That should be followed by a phase three, which sees companies incorporate these new AI tools to boost their revenues, and then eventually a phase four where companies begin to see the productivity gains created by AI boost their bottom lines. There are signs that the market is already anticipating these next steps. While Nvidia has continued to push higher in the opening months of 2024, some of these more forward looking AI plays are also showing signs of life, Hammond said. "Based on performance and valuation, investors have already started to price subsequent phases of the AI trade. An equal-weighted basket of Phase 2 stocks is up 14% during the past 6 months, largely driven by valuation expansion. ... Phase 3 stocks are up 21%, primarily driven by valuation expansion. Phase 4 stocks have seen limited valuation expansion," the note said. Goldman included a list of those "phase two" stocks in the note. Some of the names are familiar to investors, including tech giants like Amazon , and major chip stocks like Broadcom . Both have easily outperformed the S & P 500 over the past year, but not nearly to the same extent that Nvidia has. GlobalFoundries is another semiconductor name, but its stock has fallen nearly 20% this year. That type of chip stock may be a better value for investors at this point, according to Goldman. "Within Phase 2, Foundry & [integrated device manufacturers] have a relatively attractive setup of strong expected EPS growth with modest valuations," Hammond wrote. There are other companies that work behind the scenes, so to speak, in developing the hardware necessary for AI, including Teradyne and Keysight Technologies . Both of those stocks have slipped in 2024, so they could prove to be well-timed bets if Goldman's phase two comes to pass. And of course, the data necessary for AI needs to be stored and protected, bringing even more companies into the fold. Some areas that could benefit include cybersecurity companies such as Palo Alto Networks , and utility companies like NextEra Energy that can help power the data centers. Those two names are also down year to date, despite being well-liked on Wall Street. For example, Palo Alto Networks has a buy or strong buy rating from about 70% of the analysts that cover it, according to LSEG. For investors looking a bit further into the future, Goldman listed software companies including Intuit and Adobe as firms that should see revenue gains from AI and whose stocks already trade with a strong correlation to Nvidia — in other words, "phase three" plays. And for those who are willing to be even more patient, the note mentioned companies that could see productivity gains from AI driving earnings growth in "phase four." Those stocks include Pinterest , Tenet Healthcare and analytics company Clarivate . — CNBC's Michael Bloom contributed reporting.