MNI INTERVIEW: China Property Bailout Would Take 10% Of GDP

Disposal of bad debt from real estate-related companies and financial institutions caught up in China’s real estate downturn would cost around 10% of GDP, a prominent economist told MNI, adding that such a rescue is necessary in order to address weak domestic demand and disinflationary forces now being reinforced by AI.

Extrapolating from Japan's experience in disposing of non-performing assets from 2002 to 2004 implies that China would need to spend on a significant scale to clean up non-performing debt as a result of the falling property sector, requiring tremendous determination and capacity for execution from policymakers, said Lu Ting, chief China economist at Nomura.

But the drain on residential and local government investment has exacerbated deflationary pressures, leading to a continuous fall in China's 10-year treasury yields over the past few years, even as yields overseas rise, reducing global investors’ interest in Chinese assets, he said in an interview. Lu acknowledged that he saw little room for either monetary or fiscal easing in the second half of 2026, but expects year-on-year GDP growth to slow to 4.1% in Q2, compared to market expectations of expansion of 4.7%. The country saw a 5% y/y GDP expansion in Q1, much higher than expectations. (See MNI: PBOC To Support Fiscal Efforts, Broad Easing Seen Limited)

The negative interest rate differential between China and other big economies has led to capital outflows and several rounds of yuan depreciation since the autumn of 2022, prompting the authorities to strengthen capital controls, Lu noted.

A failure to resolve real estate debts would leave China in danger of repeating the stagnation which befell Japan, while the fiscal burden on cities throughout the country grows ever larger, he said, estimating that the average decline in national housing prices from their peak has been about 35% to 40%.(See MNI INTERVIEW: PBOC Must Stoke Inflation, Target 4% CPI)

AI IMPACTS

This disinflationary pressure could increasingly be exacerbated by the development of AI, which could amplify the wealth gap between Chinese residents and across regions, further weakening consumption and investment demand and posing major challenges to policy making, Lu said.

While the AI boom is boosting fixed asset investment, exports and productivity, it is unlikely to rescue the economy from the current cycle of weak growth and prices, the economist said. AI-related industries should account for more than 1% of GDP in 2026, while AI-related fixed asset investment should contribute 0.3 percentage points to growth, but these shares are well below those of real estate and traditional infrastructure. AI investment is also significantly lower than peak levels in new energy vehicles, lithium batteries, and photovoltaic panels, he added, noting that the threshold for AI investment is high and that the sector still faces barriers from U.S. restrictions on export of key equipment.

AI accounted for half of China’s 19% export growth in dollar terms in May, though this was mainly due to rising prices, Lu said. Chip exports grew by 110% last month, but were up only 2% in volume terms, he noted.

A strong export performance does not anyway fully compensate for weak domestic demand, while the economic benefits of AI are highly concentrated in a small number of cities with high-end talent, he said. The marginal propensity to consume of the elite group benefitting from AI is low, while the middle- and low-income groups vulnerable to displacement by the technology, including many college graduates and flexible workers, have a strong willingness to consume, though their purchasing power is declining due to unstable employment and wage pressure, Lu said.

Coming on top of delayed marriage rates and low levels of childbirth, AI will add to sustained deflationary pressures, while any recovery in the real estate market may be limited to a few, mainly tier-one cities, according to Lu.

The economist argues that policymakers, while promoting AI investment, should take measures to avoid the continued widening of regional disparities and allow lower-tier cities to share in the AI dividend. They should improve social security provision to alleviate the effects of unemployment caused by AI substitution, and slow down the deployment of technologies such as autonomous driving that have an impact on blue-collar jobs, he said.