The article below explores how stock markets struggle to predict the real impact of artificial intelligence on businesses and industries. It highlights investor uncertainty, historical examples of technological disruption, and why markets often misjudge major innovations.
In a literal sense, stock markets are fortune tellers; their role is to predict which companies will prosper in the future and which will not. This is a straightforward extrapolation problem when there is little change.
Stock Markets and the Uncertain Impact of Artificial Intelligence
It becomes more difficult when change occurs. This is undoubtedly true during periods of drastic change, like the current global fog of war. However, this also applies to more significant but slower-moving disruptions, such as those caused by artificial intelligence.
AI is causing confusion everywhere. As shown in chart 1, Goldman Sachs has developed a share-price index of companies most vulnerable to disruption. This has decreased by more than 20% in the last year. Even while several stock markets are close to record highs, the bank’s mirror index of “long-term AI beneficiaries,” whose earnings stand to gain the most from increased AI-fueled productivity, has dropped by almost 5%.
🤖 AI Impact on Stock Markets
- Market Confusion: Investors struggle to identify AI winners and losers
- AI Risk Index: Goldman Sachs index of vulnerable firms fell over 20%
- AI Beneficiaries: Expected winners also declined nearly 5%
- Investor Sentiment: Markets are adjusting expectations frequently
- Economic Question: AI productivity gains are still uncertain
- Key Issue: Predicting AI disruption remains difficult
Investor Confusion Over AI Winners and Losers
Investors frequently are not even able to determine whether a company is an AI winner or failure. The share price of the language-learning company Duolingo doubled between May 2024 and May 2025. It has decreased by 80% since then. AI seems to have destroyed Google not too long ago. Its parent firm, Alphabet, has seen an 85% increase in share price in the last year.
For their part, bond traders believe it is all for nothing. Real interest rates ought to increase in an era of greater AI-driven economic expansion. However, yields on 30-year Treasury bonds are essentially unchanged from the beginning of the year at 4.9%.
Bond Markets Show Little Reaction to AI
Long-term yields decreased when Isaiah Andrews and Maryam Farboodi of the Massachusetts Institute of Technology examined bond-market movements surrounding significant AI model releases.
AI is therefore both everything and nothing, depending on where you look: a rounding mistake for the economy and an existential menace for businesses. Market observers are irritated by this. In terms of history, it is also typical. Because stock markets may be worse at pricing technical advances if they are poor at pricing conflict.
Lessons From Past Technological Disruptions
For every Blockbuster, which the stock market began devaluing two years prior to the video-rental darling’s 2004 revenue high, there is a BlackBerry or a Kodak, where investors were unaware of problems until the company was in dire straits. Markets frequently expect a technological disruption that never occurs, much like economists and recessions.
It was the ChatGPT of the 1870s, when Thomas Edison’s electric lighting revolution got underway. With the support of John Pierpont Morgan and other financiers, the smart money of the time poured into the new technology and out of gas businesses, which up until that point had produced the majority of artificial illumination. However, gas did not lose its significance. Investors realized that for years to come, electric light would be more expensive than gas.
How Old Industries Adapt to New Technology
Gas companies discovered bigger markets to target even as the cost of electric light decreased over the ensuing decades. Just 2% of its clients in London had gas cookers in 1892, according to the Gas Light and Coke Company. More than two-thirds did by 1911.
To support the anecdote, The Economist looked at American and European stocks between 2005 and 2026 and discovered about 80 cases in which a whole industry—from media to telecoms to luxury goods—entered a sectoral bear market, with share prices falling by at least 20 percentage points over the course of three months in comparison to the larger index.
📉 Sector Bear Markets and Tech Disruption
- Study Period: American and European stocks (2005–2026)
- Total Cases: About 80 sector bear markets
- Definition: Sector falls 20% vs wider market in 3 months
- Cause: Investors fear structural industry change
- Examples: Media, telecoms, luxury goods
- Key Insight: Markets often misjudge technological disruption
Sector Bear Markets and Structural Change
(We did not include the energy sector, which fluctuates with the price of fossil fuels.) When this occurs, we presume that the market is concerned that a structural shift will negatively impact that sector. This can be a new institutional structure or technology (like globalization).
The share price declines half the time after such a sectoral bad market. In those instances, investors had accurately anticipated a long-term shift in the conditions of an industry. Investors valued this early in the 2010s, when competition from Chinese rivals irreversibly destroyed a large portion of the European solar business. Additionally, they predicted that as the internet grew, telecom companies would suffer (see chart 2).
When Markets Get It Wrong
Within a few years, the share price climbs above the market as a whole in the other half of bear markets, indicating that the initial bets have soured. American and European tobacco companies are prime examples. Investor concerns about the impact of e-cigarette and vaping advancements caused their share prices to plummet several times in comparison to the market. Tobacco stocks repeatedly emerged from the ashtray.
Our findings imply that the stock market finds it difficult to reflect structural shifts in the economy. This aligns with Song Ma’s research at Yale University. According to Mr. Ma, analysts frequently overestimate future profitability even when a company’s technological foundation is becoming outdated (as determined by how advanced its patents are). This supports the share prices of outmoded companies.
Why AI Creates Unique Market Uncertainty
In light of Mr. Ma’s findings, it would not be shocking if investors today overestimated the threat posed by artificial intelligence to some businesses while underestimating the harm to others. After all, compared to earlier technology changes, there is much more uncertainty around the AI transition.
There are two main causes. The first relates to the technology itself. In some areas, including coding, AI capabilities have advanced quickly. However, progress varies depending on the task. Compared to a few months ago, performance on open-ended writing and idea generating is not appreciably better.
The Economics of Superintelligent AI
The economics of a superintelligent AI is the second area of uncertainty. The beneficiaries of a “artificial global intelligence” are unknown. Businesses’ profit margins may decrease if AI lowers entry barriers. Prominent AI labs claim both large computer power expenses and quick income growth.
Numerous scholarly studies indicate that as investors become enthusiastic about the future, new technologies lead to market bubbles. However, some, like Boyan Jovanovic of New York University and Jeremy Greenwood of the University of Pennsylvania, contend that the stock market may actually decline because investors anticipate that newly unlisted companies will absorb the earnings. Many supporters of OpenAI and Anthropic, two rival AI labs, are undoubtedly hoping for it.
Corporate Reinvention in the Face of Technology
Today’s losers could become tomorrow’s winners in a world where fundamental perspectives on AI shift rapidly. A lot will depend on how well businesses use AI to enhance their customer offerings. A recurring topic in corporate history is how to innovate your way out of perceived technical dangers.
Alexander Graham Bell offered to sell Western Union, the top telegraph company at the time, the telephone patent, but Western Union foolishly turned him down. However, as phone networks expanded, Western Union established a money transfer market with no direct technology rivals.
Examples of Business Transformation
American Express began as a freight forwarder before becoming well-known for credit cards. The multinational electronics behemoth Samsung offered dried fish. Some of the software companies that are currently facing significant challenges may also decide to reinvent themselves.
This inability to precisely predict the future may irritate proponents of efficient markets. However, markets merely represent the cumulative knowledge of today’s investors. No one will be any wiser as long as two Silicon Valley technologists have three different replies when asked about how AI will affect the world.
Frequently Asked Questions
1. Why are investors unsure about the implications of AI?
Investors suffer because it is unclear how AI will affect the economy in the future. While some businesses might profit from increased production, others would experience disruptions. The rapid and uneven evolution of technology makes it difficult for markets to distinguish between winners and losers.
2. Why have stock prices declined for both AI “winners” and “losers”?
Indexes of businesses at risk and those anticipated to gain from AI have both decreased. This occurs as a result of investors’ frequent value adjustments due to their uncertainty about how AI would truly impact profitability.
3. What causes stock markets to frequently underestimate technological revolutions?
Markets frequently misprice significant technological advancements, as history demonstrates. The development of electricity, the demise of businesses like Blockbuster, and the difficulties faced by BlackBerry and Kodak are a few examples. Sometimes investors respond too quickly or too late.
4. Why have bond markets not responded to AI with vigor?
Interest rates should increase if AI is predicted to greatly accelerate economic development. Long-term bond yields, however, have remained steady, indicating that bond investors now think AI would have little or no macroeconomic impact.
5. Who will profit from artificial intelligence?
Nobody is certain yet. Possible uses for profits include:
Large tech firms
Startups and AI labs
New businesses not yet established
Customers by means of reduced costs
Profits could be distributed among numerous companies rather than concentrated in a small number of massive corporations if AI reduces entry barriers.
Conclusion
Although stock markets are meant to forecast the future, there is a great deal of uncertainty due to technology advancements like artificial intelligence. Investors are still unable to predict which businesses will succeed or fail in the end.
History demonstrates that markets frequently first underestimate disruptive innovations. Financial markets will probably continue to be perplexed about AI’s actual economic impact until its capabilities, costs, and business models become more apparent.
Disclaimer: This article is for informational and educational purposes only. It does not constitute financial or investment advice. Readers should conduct their own research before making investment decisions.
