Capital Economics chief economist Vicky Redwood warned in a new note that investment in artificial intelligence will slow next year and begin to fall in 2028, bringing both market losses and a slowdown in economic growth. She forecasts that the S&P 500 index will fall 12% below its current level by the end of 2027, which would mean about 17% below the peak reached in early October. The decline in technology-oriented indices, she said, could be even larger.
Redwood assumes that the largest cloud and technology companies will spend about $800 billion on AI investment this year. In 2027, by her estimate, these expenditures will exceed $1 trillion. At the same time, she forecasts that investment will still grow in 2027, but at a slower pace, and will decline by about 10% in 2028. Last year, AI investment added about 0.3 percentage points to U.S. GDP growth. That is why its slowdown is significant for the macroeconomic picture, not only for the technology sector.
According to the economist's calculations, a reduction in capital expenditure could subtract about 0.1% from U.S. GDP. This is a moderate effect for the American economy, but for some Asian economies it could be considerably stronger. Taiwan's GDP growth, by her estimate, could be about 3 points lower next year, and South Korea's about 1 point lower. Redwood also notes that last year's market rally raised GDP by roughly 0.4%, and a market decline of the same scale would have a similar negative impact in the opposite direction. Thus, the link between the stock market and the real economy, in her logic, works both ways.
Among the risks that could deepen the correction, the economist cites higher energy prices and rising term premiums, which could make financing tighter. Such conditions could reverse investor sentiment faster than expected in the baseline scenario. At the same time, she emphasizes that her baseline scenario does not envisage a serious downturn in the U.S. economy. The slowdown in growth in 2028, in her assessment, is likely to be moderate rather than sharp.
Redwood also believes that AI is already contributing about 1% annual growth in labor productivity in the U.S. This is one reason she does not view current investment as a superficial boom without a real effect on the economy. However, the level of capital expenditure, in her view, may remain relatively high due to the need to replace outdated equipment and competition among companies that cannot afford to fall behind one another.
The economist reminds that recessions in the past often followed investment downturns, but capital expenditure itself rarely causes a slump. This means that a serious deterioration requires additional factors, not just cuts in AI spending alone. The key question, in her assessment, remains uncertain: whether revenues from AI can justify the current level of spending. For this to happen, by her calculations, AI revenues would need to grow roughly tenfold by 2030.
Redwood's forecast is an important signal for investors who in recent months have bet on the technology sector as the main driver of market growth. Her assessment shows that the future fate of the stock market is closely tied to whether major players can turn AI investment into real profits. For now, that answer remains open, and scenarios diverge depending on how quickly the market prices in the risks.