Kaitu Macro's chief economist pointed out that AI is a bubble that is doomed to burst, but it will continue to expand until then.
Kathleen Macro believes that the S&P 500 index (SPX) has set more than 30 new historical highs this year, and the AI craze will continue to boost the stock market until 2025.
The institution's chief economist, Neil Shearing, predicted in a report this week that by next year, the benchmark index will reach a peak of 7,000 points, which means it will rise another 27%.
Although early predictions about AI suggested the bull market could last for years, Shearing believes this is a bubble that will inevitably burst - but it will keep expanding until then.
"If we are correct that there is an asset price bubble emerging around AI (just as there was around other transformative technologies), then it may well expand further before it bursts, Shearing said. But it will burst. Bubbles come and go; it's an immutable fact of the market."
Although the AI boom is compared to the internet bubble of 2000, some do not agree with the analogy, pointing out that today's AI beneficiaries are not as overvalued as those in the early days of the internet bubble.
"I think [the AI sector's] profit growth will continue to be robust, but investor enthusiasm may have gotten ahead of that growth," Shearing said.
In other words, the market is too optimistic about AI and it is too early.
Although investors are flocking to the field, believing that it will greatly enhance productivity in the next few years, Kathleen Macro's report suggests that these benefits will not begin to appear until the end of this decade.
"At present, AI is following a 'Gartner hype cycle,' which suggests that people's views on the impact of new technology evolve over time," Shearing wrote. "At least as far as AI's impact on the macroeconomy is concerned, we are surely close to the expected peak."
Before AI technology truly begins to work its magic on the economy, pessimism sets in.
In recent months, there has been new doubt about the actual impact of AI.
A study by Daron Acemoglu of MIT showed that AI's contribution to productivity is greatly overestimated - in fact, AI will only contribute 1% to GDP growth over the next 10 years.
Last month, Goldman Sachs warned that AI investment returns could disappoint companies: although large companies are expected to invest $1 trillion in the technology in the coming years, AI's performance may not offset these costs.