Why the AI boom spells short-term labour gloom

Technological innovations boost stock returns while shrinking labour, but this is not simply a workforce replacement

Featured faculty: Ying Chao
Written by Sally Ho

Capital markets have been on a tear this year, driven by a massive bet on AI investments. For instance, Chinese chipmaker CXMT’s recent debut raised US$8.6 billion to deliver China’s largest listing since 2010, while a month earlier, SpaceX made history with a record-breaking US$75 billion IPO.

On top of that, two rival AI pioneers, OpenAI and Anthropic, have announced their plans to go public in the near future, each aiming to become a trillion-dollar company. Older giants like Alphabet, Amazon, and Microsoft have also seen their stocks soar following their AI ventures.

Ironically, the labour market has taken the opposite path, marked by layoffs and hiring freezes across big tech. While overall market adjustments later often bring the job market back on track, a clear pattern emerges: waves of innovation bring short-term crises for workers despite market optimism.

“There is a glaring disconnect between financial markets and the real economy. For example, stock prices sometimes surge when hiring is weak, and stock returns often move in the opposite direction of physical investment,” says Ying Chao, Assistant Professor of the Department of Finance at the Chinese University of Hong Kong (CUHK) Business School.

“When a technological breakthrough happens, markets immediately price in the expectation of future productivity, but the real-world gears, like hiring, production, and physical investment, take much longer to catch up,” he adds.

Professor Ying’s new paper, Innovation‑driven contractions: A missing link for asset pricing puzzles, co-authored with Li Zhonghao of Nanjing University and Gill Segal of the University of North Carolina at Chapel Hill, examines why technology investment is often accompanied by short‑run contractions in the real economy.

Good news for tech, bad news for workers?

Having examined financial data on US companies in Compustat from 1953 to 2019 and measured their patent values, Professor Ying and the team find that, despite boosting productivity and efficiency, technological advances shrink the workforce with a negligible impact on capital expenditure on physical assets.

This is not simply creative destruction or technological unemployment, as it doesn’t fit the typical notion that when superior businesses overtake traditional ones, both jobs and physical investments dwindle simultaneously. “Nowadays, new technology initially reduces labour input while moving the needle only slightly on capital growth,” he says.

Professor Ying and the team then grouped companies by earnings, asset values, and investment levels. After observing how often companies adjust their product prices and tracking stock returns across the entire economy, they find that when companies rarely change their product prices, higher hiring activity and capital expenditures are associated with lower stock returns.

Companies often keep prices steady for extended periods for many reasons, such as staying competitive, retaining loyal customers, and so on, even as economic conditions shift. The more rigid or “sticky” product prices are, the more likely the stock market is to react in the opposite direction of real economic activity.

This price stickiness acts as a hidden force that explains why financial markets disconnect from the real economy and why it sometimes intensifies.

“When a technological breakthrough happens, markets immediately price in the expectation of future productivity, but the real-world gears, like hiring, production, and physical investment, take much longer to catch up.”

Professor Ying Chao

“The negative contemporaneous relationship between stock returns and investment growth or labour-market surprises becomes stronger when prices are stickier,” says Professor Ying. “Stock prices are forward-looking, reflecting expected future profits from innovation. Real economic activity, like capital expenditure or labour usage, can temporarily contract or lag behind.”

“Therefore, when a company’s valuation skyrockets before a corresponding surge in real economic activity, it doesn’t necessarily indicate an irrational market bubble. Instead, it often reflects the market rationally pricing the long-term benefits of innovation, while the real economy is still undergoing short-term adjustments.”

How sticky prices drive higher stock prices

To better understand the dynamics of technological breakthroughs affecting the real economy, the team built an economic model comprising two sectors: one producing consumption goods, and the other producing machinery and equipment for other firms. Firms in both sectors have some control over their product prices, but these prices are rigid even when costs fluctuate.

After a technology shock, firms can produce more with the same level of labour and materials. But since their prices are sticky, firms charge the same price even as production costs fall, allowing them to make larger profits on each sale. “Because product prices adjust more slowly than marginal costs after a positive technology shock, markups temporarily rise. These higher markups increase valuations while restraining short-run hiring,” says Professor Ying.

A technological breakthrough also tends to lower the price of new machinery with more efficiency. Although companies want to buy more of these advanced tools and hire more workers, it takes time for product prices and wages to catch up. This stickiness across the broader economy temporarily reduces demand in real economy activities.

Put together, workers might feel the pinch of fewer jobs in the short term, while investors see the potential for higher profits and stock returns. If product prices could change easily, the stock markets and hires would move in the same direction.

“Keep in mind that whether a specific company can sustain high stock prices depends on longer-term fundamentals, including market demand, how fast they expand capacity, and how well they fend off competitors,” Professor Ying adds.

A technology-based signal for future returns

For shareholders, the disconnect between the stock market and the real economy can make it harder to predict how a stock will perform in the short term. Therefore, based on the above model, the team develops a mathematical formula called the investment-based dividend yield to track how technological innovations, rather than stock market data, influence economic dynamics.

This investment-based dividend yield is a broad economic indicator derived from the above model, rather than a measure of the cash-flow potential of an individual company’s investments. “This tool is a macro indicator that helps predict the overall stock market’s future risk, not a tool for short-term trading on individual stocks. This indicator strongly and negatively predicts future aggregate stock returns,” says Professor Ying.

Looking ahead, Professor Ying hopes to determine whether this tool can be applied with more specific information, such as data from individual industries, and with more precise ways to track innovation, especially those related to AI. “With AI specifically, the biggest challenge now is untangling genuine productivity shocks from other noise, like government subsidies, hype-driven demand, financing conditions, or geopolitical shifts,” he adds.

“Going forward, we also need to look more closely at cross‑sectional differences: how these tech shocks impact different types of workers, the role of intangible capital, and how the ultimate gains and losses of technological change are distributed across the broader economy.”