Goldman Sachs Sees Long-Term AI Trade Reaching Far Beyond Nvidia

The artificial-intelligence boom has investors transfixed on finding the next hot stock to explode out of the market. But Goldman Sachs Group Inc. has a different idea for playing the AI trade: forget today, think long-term.

(Bloomberg) — The artificial-intelligence boom has investors transfixed on finding the next hot stock to explode out of the market. But Goldman Sachs Group Inc. has a different idea for playing the AI trade: forget today, think long-term.

The bank’s analysts, including Ryan Hammond and David Kostin, devised a basket of firms with the largest potential long-term EPS boost via AI’s ability to create a more productive workforce, lower labor costs, or both. Using 2024 consensus earnings estimates as a baseline, the team found that the median stock in their index (ticker GSTHLTAI) could see its earnings per share some 72% higher via AI-related productivity increases. 

The analysis adds to an onslaught of recent efforts from academia and Wall Street alike to understand AI’s gargantuan market impact. Productivity is but one AI-linked factor investors weigh in evaluating stocks, but in a market obsessed with ChatGPT and seemingly nothing else, the Goldman analysis lends one more tool for picking winners from losers.

“It might be a useful tool for equity researchers and investment managers who are wondering how to discipline the hyperbole coming through media channels and convert the breathless onslaught of technology news into more actionable investment decisions in combination with their own assumptions,” said Gregor Schubert, an economist at the University of California Los Angeles and co-author of a recent study “Generative AI and Firm Values.”

Any such projections come with plenty of caveats. For one, it’s unclear when each company might see these EPS gains and when the market might react. Goldman analysts expect a “meaningful macro impact” from AI sometime between 2025 and 2030, but the stock surges of individual companies could happen before AI alters any given firm’s operations, given that valuations tend to be forward-looking.

At the moment, many of the companies in the index are not exactly Wall Street darlings. Occidental Petroleum Corp. made the cut, despite shares struggling to break even this year. Then there’s Walgreens Boots Alliance, Inc., one of the S&P 500’s worst performers year-to-date. Overall, the Goldman index, which is equal-weighted, has trailed the broader S&P 500 benchmark this year, reflecting in part a stock-picking strategy focused on what AI will do well into the future. 

“There is still a lot of uncertainty about AI and the adoption timeline,” Hammond, a Goldman strategist, said in a phone interview. “That is part of the reason why I think as of now, it has not outperformed.”

The underperformance thus far stands in sharp contrast to this year’s AI all-stars — big tech stocks that have driven the Nasdaq 100 up 37% year-to-date. Notably, most of these so-called FAANGs — save Inc. — are absent from the long-term beneficiaries basket. Hammond said that those stocks could still outperform in the years to come, but the gains likely will not be driven by AI-linked increases in productivity.

The degree to which AI will revolutionize workforces depends on the task at hand. For example, a recent study found that customer service workers when given access to generative AI tools became 14% more productive than those who did not have such access. Other researchers have found that firms in finance, professional services and tech are primed to become more productive via models like ChatGPT. While seemingly fodder for job replacement woes, such findings could also imply that clever chatbots can do the grunt work, freeing humans to focus on other tasks.

In crafting the index, the analysts began by assessing how easily AI could complete a variety of workplace tasks. Taking into account the company’s implied labor costs, the team then calculated productivity increases based on two scenarios: a revenue boost with stable margins or stable revenues with a margin boost. By averaging the two together, they calculated an average EPS boost for each company. 

Based on this productivity analysis, the strategists found that the median company in the Russell 1000 will see an EPS increase of some 19%. But some firms stand to gain more than others, thus securing slots in the index of long-term beneficiaries. For example, Goldman’s team estimates that Guidewire Software Inc. could see its EPS swell some 388% due to productivity enhancements.

Such outsized estimates, though, are just that — estimates. What AI will do for any given firm is very much up for debate, which is exactly why it’s the hottest one on Wall Street.

“On paper, the potential is interesting, particularly on productivity enhancements,” said Nikhil Devnani, a Sanford C. Bernstein analyst. But, “it’s still very much in experimentation mode.”

More stories like this are available on

©2023 Bloomberg L.P.