The Bank for International Settlements (BIS) warned that if AI has a major contraction, similar to prior recessions, it could have a much greater impact on global stock markets than previous recessions. The reason for this concern is that, with all the uncertainty surrounding the financing of AI, many of the investments in AI projects internationally are highly concentrated among a limited number of investors, causing many of these investors to be dependent on a rapid deleveraging of their investment portfolio.
According to BIS’s recent annual report published on June 28, AI is one of three “pressure points” affecting the global economy (alongside inflation and sovereign debt pressures). These findings strongly suggest that technology companies will continue investing heavily in AI over the coming years. McKinsey had estimated that by 2030, approximately $6.7 trillion in cumulative capital expenditures will be required globally to meet the growing demand for compute power, including $5.2 trillion for AI-enabled data centers and $1.5 trillion for traditional IT infrastructure.

The firm also projects that global data center capacity could nearly triple by 2030, with about 70% of demand driven by AI workloads, underscoring the sustained need for investment in AI infrastructure, semiconductor technologies, power systems, and data centers. Furthermore, as stock prices continue to rise and investor confidence returns to equity markets, uncertainty associated with the possibility of not receiving an expected return on investment for AI will have potentially more dire consequences than historically experienced.
According to the BIS report, the majority of the money flowing into the AI sector has moved through hedge funds, private credit vehicles, and other types of non-bank financial intermediaries (NBFIs), which typically do not have the same amount of regulatory oversight as many regulated financial institutions. Because they lack sufficient regulation, there remain some areas within this segment of the financial services industry that could lead to significant amplifications in losses if the sentiment in the market turns against them.
A general concern about NBFIs is reflected in various studies conducted by the BIS over the past several years, with many of these reports warning about the rapid growth of NBFIs (often referred to as the “shadow banking sector”) and how financial risks are now being increasingly concentrated outside of banks as a result. Although banks have generally improved their balance sheets since the global financial crisis of 2008 by becoming better capitalized, leverage and liquidity risks have shifted into other types of unregulated financial services providers, such as investment funds, hedge funds, private-equity funds, etc., resulting in it being potentially more difficult for regulators to detect large-scale disruptions to financial markets.
According to Zhang Tao, the chief representative for Asia and the Pacific at the BIS, the interrelationship of worldwide financial systems could lead to an incredibly quick correction, unlike anything that has ever happened before. “The speed of a correction could be much faster than previous banking crisis episodes,” Zhang stated.
The BIS further noted that this situation bears a striking similarity to what transpired in the global financial crisis of 2008. A repricing of risk resulting from either rising interest rates or unfulfilling returns on AI has an equivalent potential to create disruptions in the credit markets comparable to those produced during the global financial crisis; however, the channels for transmission of those problems will vary between the two crises. In the case of the 2008 global financial crisis, the majority of the problems were caused by an excessive amount of leverage across the banking and residential mortgage loan sectors, while in today’s case, a significant portion of the risks is concentrated across several areas, which include the financing of interconnected technology, private credit markets and non-bank lenders.
The BIS defined the key risk economically rather than in purely financial terms. Should AI investments yield inadequate returns, corporations may defer making significant capital expenditures, thus transforming the current investment boom into what the BIS describes as a prolonged period of underinvestment and perhaps some indirect consequences to the availability of financial capital.
The report by BIS positions the current investment cycle for AI within broader historical trends of rapid investment in revolutionary technologies, including canal building in the 1830s, the railway construction boom in the 1840s in Britain, the electrification of the economy in the late 1920s, and the late 1990s dot-com boom. Each of these previous examples experienced sharp downturns when rapid investments exceeded the ability to generate sustainable returns on those investments. The BIS believes that fierce competition between developers of AI products and services and cloud technology may perpetuate similar conditions of excess investment to the detriment of long-term profits.
The report warns that a significant decline in tech equity valuations could trigger larger macroeconomic ramifications today than they did historically, due to the extent AI spending has become incorporated into corporate budgets, earnings projections, and overall growth projections.
Furthermore, the BIS reports that households actually have more exposure to the equity markets than they did in previous decades and therefore could experience a larger economic impact from a market correction caused by technology.
In discussing the fragile global environment, BIS General Manager Pablo Hernández de Cos stated that the 2022 inflation shock is fresh in the minds of all economic actors, which implies a heightened probability that renewed disruption of supply chains would again disanchor inflationary expectations.
The BIS has published its report ahead of the annual Sintra symposium hosted by the European Central Bank, where many of the same issues that pose risks to stability will be debated among global policymakers. For AI specifically, the most important factors will be whether corporate earnings from AI investments support the significant level of investments made recently, whether capital spending is sustained at current levels, and whether banks and regulators improve transparency concerning the non-bank funding sources that have supported AI growth. The BIS cautioned that the longer fiscal/financial oversight reforms are delayed, the more chaotic any subsequent market adjustment will be.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.


