Governments struggling with rising borrowing costs and mounting budget pressures could benefit temporarily if the artificial intelligence investment boom loses momentum, as investors shift money back into government bonds. But a sharp correction could also destabilise financial markets and damage the global economy, according to warnings from the International Monetary Fund.
The extraordinary investment frenzy surrounding AI has helped drive stock market gains, boost economic activity and increase demand for computing infrastructure. However, it has also intensified concerns about inflation, stretched company valuations and competition for investment capital at a time when governments need substantial financing.
Signs of a possible shift emerged this week after a selloff in AI-linked chipmakers prompted investors to move towards government bonds, pushing yields lower. The movement suggests that even a modest disruption to the AI boom could revive demand for government debt, offering some relief to countries facing increasingly expensive borrowing.
Yet the potential benefits come with considerable risks. A sudden collapse in AI-related investments could trigger broader financial losses, weaken economic growth and expose vulnerabilities across equity and credit markets.
Stay ahead of the geopolitical week.
MD Briefing delivers expert analysis across five global fronts — the Indo-Pacific, energy, geoeconomics, European security, and the Middle East — every Monday morning. Free.
How the AI Boom Has Complicated Government Borrowing
The AI investment surge has affected government bond markets in several ways. Strong stock market performance, increased economic activity and heavy spending on data centres and computing infrastructure have supported demand and contributed to inflationary pressures.
For central banks, persistent inflation can mean keeping interest rates higher for longer, increasing the cost of borrowing throughout the economy, including for governments refinancing existing debt or financing budget deficits.
The boom has also intensified competition for investment capital. As investors pour money into technology companies and AI infrastructure, government bonds must offer attractive yields to compete for funds.
At the same time, governments are confronting domestic fiscal pressures, political uncertainty and the effects of higher energy prices linked to the war in Iran.
Together, these factors have contributed to concerns about sovereign debt markets, where rising yields increase financing costs and can make already strained public finances harder to manage.
A slowdown in AI investment could reverse some of these pressures. If investors become concerned about technology stock valuations or the profitability of AI spending, they may redirect capital towards government bonds, particularly those considered relatively safe.
That shift could push bond prices higher and yields lower, reducing borrowing costs for governments issuing new debt.
Chipmaker Selloff Signals Growing Investor Caution
Recent market movements have offered an early indication of how quickly sentiment could change.
Earlier this week, a sharp selloff in AI-related chipmakers spread across the broader technology sector after Broadcom chief executive Hock Tan said some AI investors had limited understanding of the technology and described their behaviour as resembling gambling.
South Korean semiconductor giants Samsung Electronics and SK Hynix each fell almost 10% on Tuesday as concerns spread through the sector.
The market turbulence came ahead of another important quarterly earnings season for major technology companies and semiconductor manufacturers, whose results are closely watched for evidence that substantial AI investments are generating sustainable returns.
The decline in technology shares coincided with falling US Treasury and other government bond yields, suggesting that investors were beginning to reassess the balance between riskier assets and fixed-income investments.
The movement remains an indication of changing sentiment rather than evidence that the AI boom has reached its peak. Nevertheless, it illustrates how doubts about one sector can quickly influence broader financial markets.
BlackRock Shifts Towards US Treasuries
The change in sentiment has also been reflected in investment strategies at some of the world’s largest asset managers.
BlackRock, the world’s largest asset manager, reduced its recommendation on US equities from overweight to neutral for the first time in more than six months, citing the risk of AI-related volatility.
At the same time, the firm upgraded its recommendation on US Treasury securities, favouring shorter-dated bonds to secure income at prevailing yields.
In a note, BlackRock Investment Institute head Jean Boivin, chief strategist Wei Li and fixed-income chief investment officer Rick Rieder described the shift as a way to manage near-term risks. They remained broadly optimistic about AI but expressed concern about market volatility and expectations for Federal Reserve interest-rate increases.
The adjustment is significant because BlackRock has been broadly positive about the AI investment story. Its decision suggests that even investors who remain convinced of AI’s long-term potential are becoming more cautious about the risks surrounding market valuations and financing conditions.
Federal Reserve officials have also warned that AI-related investment could contribute to inflationary pressures. Heavy spending on data centres, computing equipment and electricity infrastructure may increase demand across several parts of the economy.
Meanwhile, high equity valuations and ambitious earnings expectations have raised questions about whether current investment levels can be sustained.
Political opposition to the construction of data centres could add another source of uncertainty as the US midterm elections approach.
Investors Debate Whether Bond Yields Have Peaked
Bank of America’s monthly global fund manager survey has highlighted the tension between continued optimism about AI and growing interest in government bonds.
Investors still identify long positions in global semiconductor stocks and bets against US Treasuries as two of the market’s most crowded trades.
The survey nevertheless revealed differing views about how much further government bond yields might rise.
On average, respondents expected the yield on the 10-year US Treasury to increase by another 65 basis points before reaching its peak. However, a sizeable minority believed investors should already begin shifting towards bonds rather than waiting for yields to climb further.
More than a third of respondents expected the opportunity to switch from equities to bonds to emerge only when the 10-year Treasury yield reached 6%. Almost one-fifth believed the market had already reached a turning point.
If expectations change rapidly, investors could rush to purchase government bonds at current yields before prices rise and yields fall.
Such a shift could benefit governments by lowering the cost of issuing new debt. However, an abrupt reversal of crowded investment positions could also generate significant volatility, particularly if financial institutions have become heavily exposed to the same market assumptions.
The danger is that a relatively small shock could trigger a much larger adjustment if investors attempt to exit similar positions simultaneously.
Why an AI Correction Could Be a Double-Edged Sword
For governments, a cooling AI market could offer some short-term financial relief. Lower bond yields would reduce borrowing costs at the margin, while weaker investment demand could ease some inflationary pressures that have complicated central bank policy.
But the consequences would depend on the scale and nature of the correction.
A gradual slowdown in AI investment could allow financial markets to adjust without causing widespread disruption. A sudden collapse in valuations, by contrast, could erode household wealth, damage corporate balance sheets and undermine business confidence.
The economic consequences could extend beyond technology companies. Semiconductor manufacturers, data centre operators, electricity providers and businesses investing in AI infrastructure could all face weaker demand or financing difficulties.
If the resulting shock weakened economic growth, government revenues could also suffer as corporate profits and household spending declined. Governments might then face greater borrowing needs even as bond yields fell.
The International Monetary Fund has warned about the risks that a sharp correction in AI-related markets could pose to financial stability and the global economy.
That warning underlines the limits of any potential benefit to governments: cheaper borrowing would offer little comfort if a market crash simultaneously weakened economic activity and increased pressure on public finances.
What Happens Next?
The coming months will test whether AI investment can sustain its current pace and justify the expectations embedded in technology stock valuations.
Quarterly earnings from major technology companies and semiconductor manufacturers will provide important evidence about demand, profitability and the financial returns from expanding AI infrastructure.
Central bank decisions will also matter. If inflation remains persistent, government bond yields may stay elevated even if enthusiasm for AI weakens. Conversely, a significant deterioration in economic prospects could strengthen demand for government debt as investors seek safer assets.
For governments facing strained budgets, the prospect of lower borrowing costs may be welcome. But relying on an AI market correction to ease fiscal pressures would be a risky strategy.
The central issue is not whether governments would benefit from a temporary decline in bond yields, but whether financial markets can absorb a reassessment of AI valuations without triggering a wider economic shock.
An orderly adjustment could give policymakers some breathing room. A disorderly collapse, however, could turn a potential fiscal reprieve into another source of economic and financial instability.
With information from Reuters.
