Economists are not the best judges of whether a stock market bubble exists, according to David Rees, Head of Global Economic Research at Schroders Investment. However, when looking ahead to the future impact of artificial intelligence on the global economy, two potential trajectories are foreseen: one is an "AI boom," where the technology proves genuinely disruptive and rapidly proliferates; the other is an "AI bust," where the stock market bubble bursts. From an investor's perspective, closely monitoring these developments is paramount. When the potential scenarios for AI diverge so significantly—capable of igniting a major boom or triggering a collapse—complacency and being too comfortable with the status quo might be one of the biggest risks. Both scenarios are based on Schroders' core assumption: that a robust macroeconomic environment, combined with massive investment plans from hyperscale cloud companies, will sustain capital expenditure and drive stock market gains for most of 2026. A critical juncture is then assumed to arrive towards the end of the year, when markets begin to question whether tech companies can deliver on their promises. The pivotal issue will be whether these technologies can be sufficiently monetized to generate returns on investment. At this crossroads, the scenarios begin to diverge. In the "AI bust" scenario, the stock market bubble bursts, leading technology companies to scale back their rapid investment pace, which negatively impacts the overall economy. Conversely, in the "AI boom" scenario, following market volatility, clear evidence emerges showing that AI technologies—not just large language models but also robotics and autonomous vehicles—are indeed transformative and profitable. This, in turn, would incentivize rapid adoption of the technology and potentially spawn new market leaders. Drawing on historical precedents of financial market bubbles bursting, Schroders infers that a market crash would have an immediate negative effect on private market activity. As it becomes clear that tech companies cannot monetize their AI investments, their spending plans would be shelved. It is further hypothesized that the market would experience a two-year investment downturn similar to the period following the dot-com bubble burst in the early 2000s. Falling stock prices and rising unemployment could dampen financial market sentiment and consumer spending, potentially pushing the US into a mild recession. Rising unemployment and weakening consumer demand would help alleviate capacity constraints in the US economy, allowing the Federal Reserve to lower interest rates below neutral levels. This move, coupled with some fiscal stimulus, would lay the groundwork for a consumer-led cyclical market recovery starting in late 2028. In such an environment, the overall stock market would perform well again, but with broader market breadth led by different market leaders. The "AI boom" scenario is deliberately framed as an extreme case. It is assumed that after a market shakeout in late 2026, there will be a brief pause in tech capital expenditure as winners and losers in the AI arms race become clear. Subsequently, the core hypothesis is that as AI's profound transformative power becomes evident, companies will scramble to deploy AI infrastructure and services, driving exponential growth in capital expenditure. This is expected to support robust growth in US Gross Domestic Product (GDP). However, the consumption outlook is less clear, particularly as technologies like robotics and autonomous vehicles begin to displace labor. It is assumed under the "AI boom" scenario that US productivity growth will surge to pre-dot-com bubble levels and remain around 3.5% annually. Assuming population growth and labor force participation rates remain stable, such strong productivity growth could lead to rising unemployment. Inflation in the US under this scenario is also prone to a two-speed development. Rising unemployment, coupled with pressure on incomes and consumer spending, sounds deflationary, especially for areas like housing and core services. On the other hand, labor displacement could lower costs in other service industries. However, the scramble to rapidly adopt transformative AI technology could create pressure across various sectors of the economy. If tech companies struggle to meet strong demand, it is reasonable to assume inflationary impacts could emerge in the goods sector. Significant attention is also paid to the potential energy demand triggered by AI, particularly through power-hungry data centers. With roughly half of US electricity generated from natural gas, increased demand could drive up natural gas prices. Given natural gas's importance in fertilizer production, this could also begin to exert upward pressure on food prices. Ultimately, rising layoffs and falling inflation would pave the way for significant interest rate cuts. The prospect of "jobless growth" could also have profound implications for US public finances. Approximately three-quarters of US federal government revenue comes from labor taxes, with only about a quarter from corporations. On the spending side, a large portion of federal expenditure is allocated to social welfare. This implies that the US, and governments globally, might need to collect more taxes from corporations, or perhaps even overhaul their entire tax and spending frameworks. But stepping back, the most obvious question raised by this scenario is: would governments allow AI to proliferate so unchecked?
Comments