AI stock sell-off: How to identify the long-term winners

Artificial intelligence (AI) stocks delivered strong returns for investors over the past three years as companies developing or selling AI products and services saw their share prices rise. In the early stages of the AI boom, many of these firms were among the first to monetize demand—particularly chip designers, whose revenue increased as customers rushed to buy chips used to train large language models.

However, the past few months have been more difficult for AI stocks and their shareholders. An AI stock sell-off unfolded as investors rotated out of several AI-focused companies in favor of stocks in other industries. The shift reflected uncertainties ranging from concerns about the broader economy to worries related to the war in Iran, which weighed on investor appetite for growth assets.

Today's AI environment

To understand where AI stocks may find long-term opportunities, the article points to the current build-out of AI infrastructure by cloud providers. Over the past several quarters, cloud companies have invested billions of dollars to expand infrastructure to meet demand, and the build-out is expected to continue. Major cloud players aim to spend nearly $700 billion this year to support this expansion.

While some investors have questioned the pace of spending, demand for the infrastructure has not slowed. The article also notes that AI use requires compute capacity, meaning capacity is needed now and likely will remain necessary for the foreseeable future.

Four elements to look for in long-term AI winners

The article outlines four characteristics that investors should prioritize when selecting AI stocks for the long term.

1. An AI growth track record

The company should have demonstrated AI strengths and generated revenue growth during the early phase of the AI boom. The example provided is Palantir Technologies (PLTR), which the article says has won government and commercial customers with AI-driven software designed to help users make better use of their data. Palantir has been operating for more than 20 years, progressively building its technology, which the article links to current returns.

2. Clear long-term prospects

The AI company should have defined goals and offer products or services expected to support growth into the future. Nvidia (NVDA) is cited as an example, with the article stating that the company aims to update its chips on an annual basis to keep its technology ahead of rivals. Because chips are needed to power real-world AI applications, the article argues Nvidia is likely to remain central as long as AI continues to be used.

3. Not a one-trick pony

The article suggests that diversification can reduce risk compared with a company highly specialized in a single area. Amazon (AMZN) is presented as an example because it operates across e-commerce and cloud computing, and through its cloud business it has become an AI powerhouse. The article states that Amazon Web Services is the biggest cloud provider globally and that Amazon is seeing high demand from both AI and non-AI customers, supporting the view that it can continue delivering growth over time.

4. A solid moat

A durable competitive advantage can help ensure a company remains a leader. Taiwan Semiconductor Manufacturing (TSM) is cited as the world’s biggest chipmaker, with the article arguing that its infrastructure and expertise make it difficult for rivals to replicate its position and win major customers. The article concludes that a moat may help distinguish AI winners from AI losers as the AI story evolves.

Impact and outlook

The article’s overall message is that while the recent sell-off has created uncertainty for AI stocks, the underlying demand for AI infrastructure and compute remains a key driver. It also emphasizes that investors may find opportunities by focusing on companies that meet the four criteria and by considering valuation, including selecting stocks that may be trading at more attractive levels following the decline.

A final thought

The article notes that the examples provided are meant to illustrate each strength, and it argues that these companies also reflect all four characteristics. It adds that investors should consider valuation and look for AI stocks that may be trading in bargain territory after the sell-off, with the goal of identifying long-term winners.