Severe AI downturn could put 32% of SGX-listed firms at risk: MAS

Smaller firms faced greater exposure due to lower profit margins and thinner cash buffers.

About 32% of Singapore-listed companies were assessed as at risk under a severe downturn in artificial intelligence (AI)-related investment, according to a Monetary Authority of Singapore (MAS) stress test.

The companies assessed as at risk account for 16% of overall corporate debt. The stress scenario assumed a sharp pullback in AI-related investment and revenue losses across the AI supply chain, against a backdrop of heightened geopolitical tensions.

“The increase in at-risk listed firms was concentrated amongst highly leveraged, capital-intensive firms, and those reliant on working capital financing,” the regulator said.

MAS subjected SGX-listed companies to revenue shocks of up to 30% and interest rate shocks of up to 400 basis points to capture wider credit spreads under stress.

Smaller companies also accounted for a disproportionate share of firms assessed as at risk, reflecting their lower profit margins and thinner cash buffers.

Rapid expansion in AI infrastructure has become an increasingly important driver of global capital demand. Higher global interest rates, rising semiconductor and electricity costs, and greater reliance on market financing have also raised the hurdle rate for AI investments.

AI infrastructure investment is increasing demand for global capital, whilst greater reliance on bond markets, private credit and other financing channels is creating more interconnected financial risks.

Higher interest rates and rising semiconductor and electricity costs are also raising the cost of AI investment.

A slowdown in AI earnings or returns could trigger a reassessment of AI-related valuations and spread losses across equity, bond and private credit markets.

The report said that higher global real interest rates could weigh on Asian economies through capital outflows, currency depreciation and tighter financing conditions, particularly in economies with weaker growth and external positions.

The risk comes as global financial conditions remain mixed. Following a brief tightening in financial conditions after the outbreak of the Middle East conflict, equity markets rose and investment-grade credit spreads remained tight, supported by buoyant investor sentiment and optimism around AI.

However, global sovereign bond yields have increased in recent weeks amid fiscal concerns, strong AI-related investment demand and higher energy prices. The rise in borrowing costs has sharpened the contrast with still-buoyant risk asset markets.

Domestic financial conditions have remained generally accommodative, with Singapore borrowing costs easing over the past year. The three-month Singapore Overnight Rate Average (SORA) continued to decline, whilst Singapore investment-grade credit spreads tightened.

The Straits Times Index rose 33% year on year in Q3, whilst bank credit growth remained firm amid strong economic growth momentum.

Singapore’s corporate sector has otherwise remained resilient, with stable earnings and lower borrowing costs supporting debt-servicing capacity.

Corporate indebtedness stood at 117% of GDP in Q1, whilst most firms reported stable earnings that helped cushion the impact of external shocks such as higher energy prices.

The household sector also remained resilient. Whilst household liabilities increased, financial assets grew at a faster pace, contributing to a further rise in household net worth.

Debt-servicing capacity improved as interest rates, including mortgage rates, declined over the past year. Credit quality remained strong, whilst households maintained sufficient liquidity buffers against potential shocks.

Household financial assets increased 9.5% year on year in Q2, compared with an 8.3% rise in liabilities. Buoyant financial markets and continued savings supported the growth in financial assets.

More accommodative financing conditions and supportive labour market conditions also contributed to increases in personal loans and mortgage borrowing. The rise in mortgage borrowing came alongside a moderate increase in property prices.

The household debt-to-personal disposable income ratio rose modestly from 1.04 times in Q2 2025 to 1.08 times in Q2 2026. The ratio remained below its 10-year average and came alongside a stronger household net worth position.

Household debt-servicing capacity improved from a year earlier, supported by lower mortgage rates and stable income growth.

However, MAS said ongoing energy supply disruptions and elevated oil prices are putting some pressure on household budgets.

A sharp retrenchment in AI-related investment could also weigh on global growth, business investment, and semiconductor demand, potentially affecting household incomes and employment.