The Banks Behind the China Shock

Distortions in China’s financial system are at the root of today’s global trade imbalances. Capital (along with labor) is the fundamental variable in production, and if it is made artificially cheap it alters the course of economic nature. Beijing has not delivered on its promised market reforms to its banking system and related financial industries. In fact it has done the opposite since WTO accession, tightening control over banks, molding them into policy tools, and imposing consequences on both China’s economy and trading partners. Chinese banks are saddled with ballooning, perpetually rolled-over debt piles that foist deindustrialization on others. This model is fated to decay, but China’s banking system will limp along longer than most global industries can withstand.

Chinese banks are policy actors

Banks play an unusually large role in allocating capital across China’s economy. China’s total banking assets reached roughly RMB 474 trillion in November 2025, equivalent to more than 50% of global GDP. Debt financing in China remains overwhelmingly intermediated through banks rather than capital markets (Figure 1).

In its accession to the WTO, China committed that commercial banks’ credit decisions would be based on their own evaluation and taken at their own risk. It also pledged that state-owned enterprises, including banks, would operate on a commercial basis and be responsible for profits and losses.

Twenty-five years later, China’s banking system falls far short of that commitment. Two mutually reinforcing forces shape the allocation of credit. First, Beijing and local governments steer lending directly through ownership, Party control, laws, supervisory incentives, cheap central-bank funding for priority sectors, and ad hoc policy intervention. Second, implicit guarantees embed the same political bias in banks’ own risk calculations: State backing makes favored borrowers appear safer, while government support for banks reduces the consequences of bad lending.

Beijing and local governments control banks’ behavior

Widespread state ownership gives Beijing direct influence over lending decisions (Figure 2). The six largest state-owned commercial banks (the “Big Six:” Agricultural Bank of China, Bank of China, Bank of Communications, China Construction Bank, Industrial and Commercial Bank of China, and the Postal Savings Bank of China) held, as of end-2025, 44% of all commercial bank assets in China. The Chinese government, through the Ministry of Finance, Central Huijin Investment (a state-owned investment company), and other state-owned companies, remains the majority shareholder in these six institutions. Chinese authorities also retain a controlling stake in joint-stock commercial banks, even when they are not majority-owned by the government.

The Communist Party reinforces this ownership control from inside the banks. With reportedly more than one million Communist Party members in the finance sector tasked with aligning financial activity with Xi Jinping’s political priorities, Party presence has increasingly been embedded in banks’ governance and management structures. In parallel, senior bankers have publicly pledged loyalty to Party guidance and reorganized internal decision-making to reflect Party hierarchies, affecting both state-owned and private financial institutions.

Beijing also states its expectations openly. As Xi Jinping put it in a 2017 meeting of the Central Financial Work Committee, “CPC leadership over finance must be strengthened,” while subsequent Party economic meetings have called on the financial sector to follow political guidance in its service to the economy and S&T ambitions. The Party’s control over banks is also inscribed in Chinese laws and regulations: For example, the Bank Law states that “Commercial banks shall conduct their business of lending in accordance with the needs of the national economic and social development and under the guidance of the industrial policies of the State.”

Beijing has replaced the blunt credit plans of the pre-1998 era with a set of incentives, assessments, and guidance to influence both the allocation of credit and its volume. The PBOC combines formal instruments with informal “window guidance” to influence lenders. Since 2016, its Macro-Prudential Assessment framework has evaluated banks not only on financial risks but also on credit growth and compliance with national credit policies. More recently, relending facilities have supplied cheap central-bank funding on the condition that banks lend to designated sectors or borrowers. Initially used to support rural areas and micro- and small businesses, these facilities have expanded rapidly since 2020, both in scale and in the range of policy objectives they are designed to serve. A RMB 500 billion facility launched in 2024, for example, provides one-year PBOC funding at 1.75% to support technology innovation and equipment upgrades. In May 2025, the PBOC added facilities for technology investment and the services sector. While data on some relending programs is not updated, the PBOC continues to describe relending as an active tool used to guide credit structure.

Banks themselves describe national policy as a core business objective. In its 2024 annual report, the Agricultural Bank of China (state-owned at 84%) said it had implemented national policies and supported “major national strategies and security capacity building in key areas.” The Industrial Bank (state-owned at 42%) mentioned that “The Company has actively integrated into and served the new development paradigm and actively transformed the “country’s most fundamental interests” into the ‘top priority of the Bank.’”

Authorities also intervene directly when political or financial stability requires it. Under the real-estate “White List” launched in January 2024, for example, local governments identified more than 5,000 projects and encouraged banks to finance them, including stalled projects run by distressed developers. Authorities have applied the same approach to troubled local government financing vehicles. In December 2022, banks extended RMB 15.6 billion in loans to Zunyi Daoqiao, Guizhou’s largest local government financing vehicle, by 20 years and cut the interest rates to as little as 3%—while around the same time, investors demanded more than twice that rate on the company’s bonds.

Academic evidence shows that these government pressures affect the allocation of credit. One study covering 2000–2016 found that banks increased lending to zombie firms by 178%–228% during the final year of a prefecture Party secretary’s term, while they cut lending to healthy firms by roughly 87%. Another study of 300,000 loans from a major state bank found that lending jumped by 92% at the end of each month, when branch managers faced reporting targets. SOE loans issued during those periods defaulted at roughly three times the rate of comparable non-SOE loans. A 2024 IMF study also found that banks with greater state ownership or weaker asset quality responded more strongly to industrial-policy announcements—consistent with institutions demonstrating policy loyalty when they depend more heavily on state support.

Data also shows that the PBOC’s formal and informal guidance affects lending volumes. When ordinary loan demand weakens, banks also discount bankers’ acceptances (buying short-term corporate bills in transactions that PBOC statistics count as loans) to inflate their reported balances before month-end. Discount rates have repeatedly fallen below banks’ own short-term funding costs and collapsed around reporting dates. In 2026, traders reported month-end rates as low as 0.01%, prompting regulators to tell some banks not to offer rates below 0.5%. Banks sometimes inflate their lending through “phantom loans” to meet targets. A government audit found that six state-owned financial institutions issued RMB 516.7 billion in such loans in 2023.

Widespread guarantees surrounding both borrowers and lenders amplify the distortions created by Beijing’s control over banks. Although Beijing has allowed somewhat more market discipline since its 2016 deleveraging campaign, it continues to largely shield SOEs and local-government financing vehicles from disorderly failure. Banks themselves also benefit from an implicit state backstop, which allows them to lend more than they would in a market economy. The PBOC’s 2019 takeover of Baoshang Bank marked a rare departure from the assumption that authorities would protect financial institutions. The resulting deterioration in funding conditions for other small and medium-sized banks showed how strongly the system depends on that assumption. More recently, the Chinese government injected roughly RMB 290 billion (€37.2 billion) into its banks to preserve their capacity to keep lending and support economic activity. Because banks expect Beijing to protect both them and many of their largest borrowers, the threat of failure imposes little discipline on either side of the transaction.

Banks are a leading cause of China’s trade imbalances

The political role of Chinese banks, combined with the implicit guarantees surrounding many borrowers, biases both the price and allocation of credit, creating major distortions across the economy. Favored firms borrow more cheaply than their commercial fundamentals justify and receive more credit than a market-based system would allocate to them. These distortions shape the structure of the economy, channeling savings toward manufacturing and state-owned firms, and away from services, private companies, and household consumption. The result is a persistent imbalance between domestic supply and demand, which translates into falling prices and growing trade surpluses.

Below-market borrowing

Cross-country evidence shows unusually large and persistent distortions in Chinese borrowing costs. Many Chinese companies borrow at rates well below those that their risk and commercial fundamentals would normally command. This does not mean that borrowing costs in China are low in absolute terms. Rather, banks distribute credit on preferential terms according to political priorities, giving favored borrowers an advantage over other firms in China and directing capital toward the sectors Beijing wants to support. OECD analysis shows that such distortions in China are far more pervasive than in the OECD on average (Figure 3).

That borrowing cost advantage can amount to large financial support to companies. Rhodium Group analysis, using a methodology similar to the OECD, found that Chinese state support to listed firms in the battery industry through below market borrowing amounted to RMB 12.1 billion in 2022, before slightly declining to RMB 8.6 billion and RMB 7.1 billion in 2023 and 2024 (Figure 4). Overall, the 2025 IMF study estimated the equivalent fiscal cost of IP in China through subsidized credit to be around 0.4% of GDP.

Banks do not distribute this advantage evenly. State-owned enterprises are obviously favored: In 2025, the median interest rate of central SOEs was 0.40 percentage points lower than for private companies, and that gap has widened from 0.10 pp in 2015 (Figure 5). The manufacturing sector also benefits from lower interest rates. A 2025 IMF study found that manufacturing firms benefit from lower effective borrowing rates, and this discount remains after controlling for profitability, growth prospects, and risk proxies, suggesting it is not solely explained by observable fundamentals. Data from listed companies confirms this analysis: The median interest rate of manufacturing companies was 0.85 percentage point lower than for non-manufacturing companies, and that gap has widened over the past decade, unlike observed international practice (Figure 6).

Below-market lending is becoming more systemic as banks struggle to find enough borrowers to sustain credit growth. Years of overinvestment, weak returns, and mounting debt have sharply reduced commercial demand for new loans, especially among private companies. Yet banks still face pressure to expand their loan books, refinance existing debts, and support economic growth. They therefore cut rates and increasingly turn to SOEs and other state-backed borrowers that remain willing to take on more debt regardless of expected returns. In December 2025, banks extended an astonishing 58% of new loans at or below the 3% Loan Prime Rate (LPR), the benchmark rate they quote to their best corporate customers (Figure 8). The PBOC has also guided corporate lending rates steadily closer to DR007, a benchmark for banks’ short-term funding costs, in part reflecting authorities’ efforts to maintain the flow of credit to priority sectors (Figure 7).

Misallocation of credit

Bank-led distortions in China do not only arise from below-market interest rates: The volume of lending is also subject to distortions and misallocation. The abundance of credit for certain firms and sectors can effectively operate as a subsidy, by removing incentives for companies to stay profitable and isolating them from market forces that would otherwise lead to their restructuring or bankruptcy.

SOEs benefit most from abundant credit. A 2024 IMF study covering 137 Chinese commercial banks (2004–2021) found that industrial policy announcements were associated with larger credit expansion in SOE-dominated sectors than elsewhere, crowding out non-SOE dominated sectors. This pattern is also visible in year-on-year changes in the total corporate debt stock over the last decade: New borrowing among non-SOE firms stagnated from 2018 to 2023, while SOE debt stock grew from 2.3 to 8.5 trillion RMB over the same period (Figure 9).

While Beijing’s industrial policy priorities shapes credit allocation, Chinese banks are also increasingly locked into the consequences of their past lending. Between 2019 and 2023, batteries, specialty chemicals, and machinery (all priorities under Made in China 2025) ranked among the four industries with the fastest debt growth, alongside agriculture. But banks also increasingly lend to large established borrowers in traditional and increasingly saturated industries. In 2025, for example, the 20 largest additions to corporate debt were concentrated among capital-intensive borrowers in sectors already marked by substantial excess capacity, including construction and infrastructure, property, power generation, petrochemicals, automobiles, and airlines. Out of this group of 20, 18 were state-owned.

Excessive lending results in large capacity increases unrelated to profits or demand. Across the chemical industry, between 2021 and 2025, operating profits fell by 55% and revenues barely grew, yet liabilities increased by 43% according to data from the National Bureau of Statistics (Figure 10). By June 2026, the industry’s capacity utilization had fallen below 70%, lower than during the worst of the COVID-19 shock, while chemical prices continued to decline.

Chinese banks keep “zombie firms” alive through repeated loan rollovers, creating a deflationary spiral across the economy. Chinese chemical company Chengxing, which produces material used in electronics and batteries is a classic example. In 2025, its core business, phosphoric acid, only used only 46% of its capacity. Yet after three consecutive years of losses, the company secured a syndicated loan facility of up to RMB 500 million at a 3% rate (equivalent to the 1-year LPR) to refinance its debt. Xinjiang Bayi Iron & Steel, a state-owned producer controlled by China Baowu, provides another example. By the end of 2025, after four consecutive annual losses, the firm secured RMB 8.89 billion in new borrowing, effectively rolling over all of its outstanding loans. Bayi increased production and announced plans to raise it again in 2026, even while consistently selling at a loss. Cases like Chengxing and Bayi are rife and affect the entire economy. Chinese academics found that zombie firms cause prices to fall for everyone in the market and reduce capacity utilization even among healthy companies in the same industry and along their supply chains.

Implications

China’s model has become an “industrial policy of everything,” both by choice and by default. Banks and their underpriced credit are the heart and bloodstream of this man-made economic body. The banking system sustains a vast and rising share of the industrial base, concentrating support on the largest and most indebted firms, whose failure would impose the greatest losses on banks. Even if Beijing wanted to taper off this intervention—which it has not signaled whatsoever—doing so would force banks to recognize losses, trigger defaults and factory closures, and threaten employment and local government finances.

These systemic banking system distortions drive deflationary pressures and trade imbalances. When demand and profits fall in a normal market economy, companies reduce investment, cut production, or exit, allowing supply to adjust. By endlessly financing low-return firms, keeping legacy capacity on life support, and even funding further expansion, China’s financial system obstructs each stage of that necessary and healthy adjustment. It sustains and expands industrial supply without generating an equivalent increase in household income, demand, and welfare.

This model can persist long after it ceases to generate adequate economic returns, and for much longer than foreign industries can withstand the resulting pressure. Chinese authorities control most major banks and their largest borrowers, allowing Beijing to spread losses across banks, SOEs, local government balance sheets, and the wider economy. China’s financial distortions result in a slow decay marked by declining productivity, weaker investment returns, and prolonged deflation, but they will likely fall short of an immediate crisis, due to the autocratic system of political control over Chinese society. Other countries cannot wait for those internal costs to force China to adjust. As we argued in recent research, advanced market economies, particularly Europe, risk losing entire industrial ecosystems and becoming more dependent on Chinese supply in the years to come.

The WTO would offer a legal route to challenge these practices, but it cannot provide a sufficiently rapid or comprehensive remedy. Ample evidence shows that Chinese banks misprice and misallocate credit and frequently make lending decisions on non-commercial grounds. WTO members could bring a violation complaint through the WTO dispute-settlement system, invoking China’s accession commitments and, where relevant, GATT Article XXIII. If a panel upheld the complaint, China would need to bring its measures into conformity. If China failed to comply within a reasonable period, complainants could seek authorization from the Dispute Settlement Body to suspend concessions of equivalent effect. That process would take years. Even if Beijing reformed its banks, pervasive implicit guarantees would continue to distort their assessment of risk and their lending decisions.

Governments should place the distortions generated by China’s banking system (and China’s failure to honor the commitments designed to prevent them) at the center of their diagnosis of China’s economic and trade imbalances. A WTO case could establish the legal and evidentiary basis for that diagnosis, but governments need faster trade, competition, and economic security measures to protect their industries from the pressures that the system creates.