Will China’s new financial regulatory reform be enough?

Executive summary

Effective financial supervision plays a crucial role in maintaining financial stability and a healthy financial system. China’s leadership has made financial risk a core priority, and in reforms approved in March 2023, it reassigned regulatory responsibilities, creating a new supervisory body that will take over some responsibilities from the central bank, the banking and insurance regulator, and the securities regulator.

The aim is that a change to the financial supervisory architecture (who does what in financial supervision) will make China’s system more effective and stable. In this policy brief, we argue that this incremental reform will not solve the core issues China faces in financial supervisory effectiveness.

We provide an overview of China’s large and complex financial system, including its largely state-owned banks (some of which are the largest in the world by assets), securities markets and other financial intermediaries.

Traditional divisions between different types of activities and institutions have been blurred by the rise of large financial conglomerates, risk-transfer techniques and internet-based finance.

Reforms in 2018 to China’s supervisory architecture did not eliminate perceived shortcomings, including failures to effectively regulate financial conglomerates, fintech and regional banks.

We then survey global benchmarks against which China’s financial supervisory architecture can be compared, including the United States and European Union. China’s supervisory system is already more streamlined, at least on paper, than either of these most comparable global counterparts. Like them, China’s system does not correspond exactly to any of the three textbook archetypes of supervision: sectoral, twin-peaks or integrated supervision.

Ultimately, the effectiveness of China’s financial supervisory architecture suffers from excessive state intervention in the financial system through other channels, including through the unique and pervasive influence of the communist party, which hampers supervisory independence and makes it difficult to establish accountability for regulatory failures.

While the recently announced reform may improve coordination across supervisory bodies, coordination within the new quasi-integrated supervisor, across central departments, and between them and local branches, will remain a challenge.

1 Reform looms for China’s financial system

President Xi Jinping’s determination to exert more control over China’s government and economy faces a new challenge: overhauling a system of financial regulation that oversees a highly complex web of banks, nonbanks, shadow banks and competing interests in local and national governments and the party bureaucracy.

As China’s leaders have acknowledged, the country’s economy has had trouble returning to its growth rates of past decades1. These difficulties pre-dated the pandemic and include a slumping real-estate sector, weak private investment, feeble consumer demand and deteriorating local government finances.

In the face of these and other issues, China’s leadership is pondering deeper questions of how to optimise its state and party organisations and their roles in financial regulation2.

Before the latest announced changes, China’s financial supervisory architecture was restructured five years ago, merging the authorities in charge of banking and insurance into the China Banking and Insurance Regulatory Commission (CBIRC), and giving the People’s Bank of China (PBOC) a greater role in financial-sector oversight.

On 7 March 2023, a new reform was announced, with the CBIRC renamed the National Financial Regulatory Administration (NFRA) and acquiring some competences previously located in other agencies.

Further supervisory integration is being considered soon under the authority of the Chinese Communist Party (CCP)3.

Financial supervision is a multifaceted public-policy task with several objectives, most prominently financial stability (addressing systemic risk), financial consumer protection (addressing information asymmetries) and financial market integrity (addressing fraud and criminal practices).

Even in regimes that do not share China’s party-dominated features, achieving these diverse and sometimes mutually misaligned tasks is difficult, and financial supervisory architecture choices – who does what among public entities with a financial supervisory mandate, or supervisors4 – have been a matter of animated debate in many jurisdictions, often in the wake of a financial crisis.

This policy brief aims to inform the discussion in relation to China with accounts of experiences in other jurisdictions, especially those with a large and complex financial sectors, that may serve as reference points. It also aims to inform readers outside China about Chinese financial-sector evolution and policy developments.

It focuses on supervisory architecture, stopping short of a comprehensive consideration of current financial stability challenges and financial services policy reform in China. It argues that although the financial system needs reform, no fundamental change of supervisory architecture is presently necessary.

Over the last four decades, a time of tremendous economic growth, China’s financial sector has grown much more complex, a complexity compounded by the pervasive role of the CCP spanning all organisational structures of the government, supervisory agencies and most financial firms.

Reforming the financial supervisory system to avoid major bank failures and system-wide instability is an ongoing challenge, to which the responses of the Chinese authorities have been broadly effective so far, while largely aligning with the letter of applicable international financial regulatory standards.

On paper, China’s current financial supervisory architecture is more streamlined than the equivalents in the United States and European Union, where the architecture of financial supervision is exceedingly complex because of burdensome historical and political legacies.

China’s recently announced reshuffle, like previous changes in supervisory architecture, appears incremental rather than radical. It will not, however, resolve fundamental challenges hobbling China’s financial system, which are not linked to specific choices of supervisory architecture but rather to excessive CCP and state intervention, and the lack of supervisory independence resulting from China’s CCP-dominated governance system.

Despite the growth of private-sector financial firms, China’s banking system remains dominated by a handful of gigantic institutions that are majority-owned by the central government. Similarly, several of the larger insurers are central state-owned enterprises. Many of the largest securities firms are mixed-ownership enterprises, with state entities holding significant stakes.

The party-state structure applies heavy-handed control over capital and credit allocation decisions, subject to political or government priorities or favouritism. Some Chinese scholars have criticised the system as reflecting excessive state intervention5.

These features risk a collision with the normal functions of financial regulation, such as formulating minimum capital requirements for banks and insurers, cleaning up failing or failed borrowers, ensuring regulatory compliance, conducting stress tests and crafting disclosure rules to protect investors.

2 China’s financial sector and current supervisory architecture

The starting point for China before Deng Xiaoping started the reform era was a so-called mono-bank system under the planned economy, in which the People’s Bank of China played the roles of central bank, regulator and monopolistic commercial bank all in one (Lardy, 1998).

2.1 China’s financial sector has unique features and has become very large and complex

Since 1978 several commercial banks have been carved out of the PBOC, and the creation of other banks and financial firms has been allowed, resulting in China now having a very large and complex financial sector, by most measures among the largest in the world. The Chinese banking system is the world’s largest in terms of aggregate assets, ahead of the euro area and well ahead of the United States (Figure 1)6.

Its public equity market is second only to that of the United States in terms of total market capitalisation. Its bond market has also become the world’s second largest, behind the US and ahead of both Japan and the euro area7.

Figure 1. Geographical distribution of the world’s top 100 banks by total assets, 2005–21

Source: Bruegel based on The Banker.

Since 2019 the world’s four largest banks by total assets have been Chinese: the Industrial and Commercial Bank of China, China Construction Bank, Agricultural Bank of China and Bank of China8. Many of the smaller Chinese banks – some of which are very large by international standards – have diverse shareholder structures that includes a mix of public and private-sector entities.

Similarly, several of the larger insurers are central state-owned enterprises, with the significant exception of Ping An, a private sector company9. Several of the largest securities firms, including CITIC Securities and Haitong, are mixed-ownership enterprises, with state entities holding significant stakes, though not the majority of equity.

The high degree of state ownership and intervention in the financial sector is a defining feature of China’s financial system, including through the mechanisms associated with the involvement of the CCP. State and CCP channels of influence include ownership, personnel appointments and more, all of which complicates financial supervision.

The party-state interferes in multiple ways in the operational management of financial firms, through detailed regulations but also direct nudging (or heavy-handed direction) of capital and credit-allocation decisions.

The CCP Central Organisation Department is the main institutional player for the appointment of the top executives of the largest state banks. These executives double as government officials with vice-ministerial rank (Heilmann, 2005) and often revolve in their careers between state financial firms and supervisory bodies.

For example, the current top banking supervisor, Guo Shuqing, was China Construction Bank chair from 2005 to 2011, and his successor there transitioned to the bank chairmanship from leading the PBOC’s anticorruption body.

To be sure, some of these features are not entirely unique to China. It is natural that the state selects the executives of banks in which it holds majority ownership. ‘Revolving doors’ between government and the financial industry exist in many other countries, including the United States. And neither financial repression nor directed credit are exclusive to China.

Still, the role of state-owned financial firms is much greater in China than in any other of the world’s very large financial jurisdictions, and the CCP has no functional equivalent almost anywhere else.

The complexity of China’s financial sector results in part from the country’s extraordinary burst of entrepreneurship since the 1980s, as new types of private financial firms have emerged, including asset managers, leasing firms, peer-to-peer (P2P) lending platforms and specialised insurers.

The four largest state-owned banks are no longer as dominant, falling from 95 percent of the total assets of Chinese banks among the world’s 1,000 largest banks in 2002, to 56 percent in 202210.

The 16 other Chinese banks in the 2022 ranking have diverse shareholding structures, and half of them are headquartered in places other than Beijing or Shanghai. The smallest, Bank of Ningbo, had over $316 billion in assets as of end-2021, equivalent to the twelfth-largest bank in the United States.

Like the US and EU, but unlike almost any other jurisdiction in the world, China now has multiple financial centres. In addition to Beijing, where the largest banks and other state-owned financial giants are headquartered, these include: Shanghai, a hub for the equity market with many large branches of banks headquartered elsewhere; Shenzhen, the most vibrant centre for startup finance and venture capital; Hong Kong, a major venue for international finance despite its loss of stature in recent years; and Dalian, the location of China’s main futures exchange and commodities market.

By contrast, in the next largest jurisdictions (other than China, the EU and the US), a single financial centre dominates: Tokyo in Japan, London in the United Kingdom, Toronto in Canada, Sydney in Australia11, Seoul in Korea and Zürich in Switzerland.

China consolidated its financial supervisory architecture in 2018, but even after these reforms, supervisory failures have persisted

2.2 China’s evolving financial supervisory arrangements and challenges

China’s financial rulemaking has converged substantially with relevant international standards in recent decades.

In particular, China’s accounting standards are largely consistent with the International Financial Reporting Standards. China has also adopted the international standards defined by the Basel Committee on Banking Supervision.

On the latter point, China is better aligned with the internationally accepted norms than the EU; on the former, China is better aligned than the US, which maintains a purely national accounting framework12.

The Chinese government established the China Securities Regulatory Commission (CSRC) in 1992, the China Insurance Regulatory Commission (CIRC) in 1998 and the China Banking Regulatory Commission (CBRC) in 2003.

This ‘one bank and three commissions’ model involved specialised supervisors for the different types of financial firms and markets, with the central bank, the PBOC, sometimes playing a coordinating role.

In 2015, China established a deposit insurance agency, initially hosted directly by the PBOC (Desai, 2016), and entrusted since 2019 to the Deposit Insurance Fund Management Co. Ltd, a PBOC subsidiary13.

The rise of shadow banking, for example in wealth management products and trusts that also conducted lending, especially after 2008, blurred lines between the regulatory silos and led to a rethink.

The International Monetary Fund (IMF, 2017, page 34) found in its last published Financial Sector Assessment Programme report on China that “oversight of risks is hampered by a regulatory architecture that can leave significant gaps in functional supervision” and that incentivises regulatory arbitrage. Financial firms that performed the same functions but took a different form could face vastly different regulatory requirements and oversight.

Since that IMF report was published, the Chinese authorities have taken steps to contain the risks of shadow banking, clamping down on some of the regulatory arbitrage like banks’ off-balance-sheet lending, but have struggled to keep pace with some financial-sector developments.

Online P2P lending illustrates the pitfalls of a supervisory architecture in which supervisory authority is determined by the type of financial firm. Most P2P lending platforms were effectively underground banks masquerading as tech companies (Chorzempa, 2018).

None of the supervisors had been given explicit authority over the P2P segment and it grew to massive scale, at which point none wanted to touch it and risk being blamed for the eventual implosion. Chinese officials estimated that 50 million investors were involved, with around 800 billon renminbi ($115 million) outstanding when authorities shut down the entire P2P industry in 2019–2014.

The CIRC was long widely viewed as captured by the insurance industry, and in 2017 its chairman Xiang Junbo was arrested for corruption. It failed to police risky behaviour, like the sale of risky short-term investments disguised as insurance, which led to the high-profile collapse of large insurers, requiring the government to step in at enormous cost and effort to restructure them (eg. Anbang Insurance Group in early 2018).

With the stated aim of improving coordination among financial authorities, in the wake of the 2017 National Financial Work Conference, China established the Financial Stability and Development Committee (FSDC), headed by a vice premier (Liu He) who outranked the heads of regulatory agencies, and with a small secretariat hosted by the PBOC.

China consolidated its financial supervisory architecture in March 2018, merging the CIRC into the CBRC to form the China Banking and Insurance Regulatory Commission. The PBOC took over some of the CBIRC’s policymaking functions related to overall financial stability and systemically important financial institutions. The CSRC remained mostly untouched in that round of reform15.

Even after these reforms, supervisory failures have persisted. A failing bank in Inner Mongolia, Baoshang Bank, was taken over by authorities in May 2019 and later sent into bankruptcy, the first Chinese bank in two decades to do so.

Authorities blamed its controlling shareholder, the Tomorrow Group, a conglomerate whose founder Xiao Jianhua was swept up in a corruption probe, for treating Baoshang as a ‘piggy bank’ through lending to companies associated with the parent16.

Such lending within the same group of entities, known as related-party lending, and corruption also played a role in serious issues at the Bank of Jinzhou in Liaoning Province, Hengfeng Bank in Shandong and several other relatively small local financial institutions.

In response to the challenge of supervising financial conglomerates, the PBOC created a financial holding company regime in November 2020, through which it supervises at the group level companies that control banks or multiple financial firms, or surpass certain thresholds for financial assets17.

The lack of effective coordination, especially between the CSRC and the PBOC/CBIRC, and despite the creation of the FSDC, played a role in the last-minute cancellation of Ant Group’s blockbuster initial public offering (IPO).

The PBOC and CBIRC had not decided on a stable regime to regulate Ant Group, a complex financial technology firm, at the time the CSRC approved its IPO. When the risks posed by Ant Group’s size and business model were revealed late in the process, authorities opted to hastily cancel the offering.

A more effective approach would arguably have had the CSRC coordinate with the PBOC and CBIRC to ensure their approval of the IPO of a firm under their authority (Chorzempa, 2022, page 219).

The coordination challenge is not only horizontal, across central authorities in Beijing, but also vertical, between authorities in Beijing and those in local governments – and also in the PBOC, CBIRC (and future NFRA) and CSRC, and between their head offices in Beijing and their local offices.

While the CBIRC has supervised all banks and insurers, other financial firms not subject to the CSRC’s authority have typically been supervised by financial services bureaux at the provincial and/or sub-provincial levels18.

Traditionally these have included small loan companies, local asset-management firms and financial leasing firms, but more recently also fintech firms that provide nationwide services through the internet.

The rise of such nonbank financial firms supervised at the local level poses challenges for supervision, as local regulators lack both the authority and capacity to effectively oversee such firms’ activities.

These challenges are compounded by generally insufficient resources allocated to financial supervision in China, at least at the central level. The IMF (2017, page 39; also Figure 11 on page 40) noted that supervisory resources “are insufficient to adequately oversee a large and complex financial system, and need to be substantially increased.”

There is no indication that this shortcoming has been substantially addressed since. The new reforms may actually make the situation worse as they involve significant pay cuts for supervisory staff and other civil servants, which is likely to impede talent attraction and retention, and to create even more avenues for supervisory capture and corruption.

3 Experiences from other jurisdictions

The unique features of China’s financial sector call for a highly tailored policy and supervisory architecture. There is no reason for China to replicate any model from abroad, but knowledge of relevant experiences in other jurisdictions can usefully inform the Chinese policy debate, if only to avoid repeating mistakes made elsewhere.

Chinese officials have in the past asked for advice and studied foreign models, including those of the United Kingdom and United States, when considering reforms19.

3.1 Varieties of financial supervisory architecture

In most countries, the specialised public agencies tasked with the supervision of financial firms and markets are only decades old (Hotori et al 2021), with the result that there is less depth of accumulated comparative experience in this than in other policy areas for which China has looked abroad for inspiration. For example, Germany created a securities regulator only in 1994, two years after the establishment of the CSRC.

A commonly held categorisation identifies three main archetypes:

  • A sectoral supervisory architecture (also referred to as institutional or functional) in which separate agencies supervise, for example, banks, insurers and securities firms. This is the main organising principle of financial supervisory architecture in China.
  • An integrated architecture entails a single authority in charge of most or all supervisory roles, as is the case with Japan’s Financial Services Agency (FSA) or Germany’s BaFin20.
  • A ‘twin peaks’ architecture distinguishes between prudential supervision, aimed at mitigating systemic risk and preserving financial stability, and conduct-of-business supervision, aimed at mitigating information asymmetries and protecting savers, investors and other consumers of financial services, as well as the integrity of the system as a whole21. Australia, Belgium, the Netherlands, South Africa and the United Kingdom all operate under a twin-peaks architecture. In the UK, the Bank of England is the prudential peak and the Financial Conduct Authority the conduct-of-business peak.

Realities are always more complex than any such taxonomy can capture, and each category comes with significant variations, based on the circumstances that led to its adoption. Intersecting the three archetypes is the central bank’s involvement in financial supervision, a question of relevance for China, as previous debates over architecture have included suggestions to integrate supervision under the PBOC22.

In a sectoral framework, it is common but not universal that the central bank or a body under its direct authority is the banking supervisor. The prudential authority is under the central bank in most twin-peaks jurisdictions (but not in Australia).

There is more variation in integrated supervision countries; the integrated supervisor is either the central bank itself (eg. in Hungary, Ireland, Russia, Singapore), or a separate institution (eg. in Germany, Japan, South Korea, Switzerland).

Some maintain mutually independent bank examination channels at the central bank and the integrated supervisor (eg. Japan), while others have organised a division of labour (eg. Germany, where the Bundesbank performs most operational banking supervision, which feeds into BaFin’s decision making23).

Resolution, a hot topic in China following the recent bank failures, is an additional point of differentiation. The creation of a dedicated resolution authority – thus avoiding what has often been described as ‘constructive ambiguity’ as to how failing banks may be handled – has happened only very recently in most jurisdictions other than the US, where the Federal Deposit Insurance Corporation (FDIC) was established for that purpose in 1934.

In most cases, bank-resolution authority is vested in the main banking supervisor – for example, BaFin in Germany, the FSA in Japan (jointly with the national deposit insurer) and the UK Prudential Regulatory Authority (part of the Bank of England). The main outliers are the United States with the FDIC, and the EU, as detailed below.

3.2 China’s benchmarks and their limitations

Because of the massive size and complexity of China’s financial system, for matters of financial supervisory architecture, the most meaningful comparison points are the US and the EU (or euro area)24.

For some aspects of the discussion on supervisory architecture, Japan and the UK can also provide useful reference points; both, however, have smaller financial systems, and their complex financial activities overwhelmingly occur in one location, respectively Tokyo and London, a considerably simpler setup than in China, the US or the EU, where competition among financial centres is associated to some extent with rivalry between the corresponding local governments. Other jurisdictions are generally too small for a direct comparison to be useful.

A common feature of the US and EU financial supervisory architectures is their considerable complexity and related challenges of turf delineation and overlap – much greater, on the face of it, than in China.

The US has four federal prudential supervisors for deposit-taking financial firms: the Federal Reserve, FDIC and Office of the Comptroller of the Currency (OCC) for banks, and the National Credit Union Administration (NCUA) for the separate system of credit unions. The FDIC is the resolution authority for banks, and the NCUA for credit unions.

In addition, each US state has its own autonomous banking supervisor, although in practice there is significant coordination with their federal peers. There is no US federal insurance supervisor; even large nationwide insurers are supervised only at state level.

The Securities and Exchange Commission has a broad mandate over securities markets, but must share the turf of derivatives markets with the Commodities Futures Trading Commission, a division of labour that is a historical legacy with no apparent justification in substance25.

There are separate supervisors for publicly sponsored specialised financial institutions, anti-money laundering supervision and macroprudential oversight. Thus, the US supervisory architecture has many elements of a sectoral architecture, but is considerably more complex.

In the EU, arrangements at member-state level are much more variable than at state level in the US, let alone at provincial level in China. To start with, the European Central Bank (ECB) is the central bank for most but not all EU countries26, most of which have several financial supervisory authorities under different models.

The 20 countries of the euro area, together with Bulgaria, have in the last decade pooled banking supervision in a Single Supervisory Mechanism (SSM) that brings together the ECB as central decision-making institution and the respective national bank prudential supervisors27.

For these countries, the Brussels-based Single Resolution Board plays a central but not exclusive role in resolving larger banks. Smaller banks in the banking union, and all banks in other EU countries, are resolved by national resolution authorities, if not through a court-ordered bankruptcy process (Gelpern and Véron, 2019).

Three other sectoral EU-level agencies coordinate supervision, respectively for banking, insurance and pensions, and securities and markets. Aside from limited exceptions, however, they are not financial supervisors, which makes their names partly misleading28. The EU is also in the process of creating a central Anti-Money Laundering Authority.

3.3 Strengths and weaknesses of different arrangements

It is extremely difficult to evaluate the relative performance of supervisory frameworks. The direct costs and administrative burden of supervision should not be neglected, but cannot be the dominant assessment criterion, given the much greater magnitude of policy outcomes at stake.

Arguably the most important role is to avert financial instability, and to mitigate it when it happens, but financial crises are infrequent and tend to be caused by a multiplicity of factors that are impossible to fully disentangle.

As for conduct-of-business supervision, quantitative indicators are inherently ambiguous: a rise in the number of fines for noncompliance, say, may be caused by more widespread violations (bad), or greater strictness (good), or both.

There have been fads in this area, which in retrospect have often appeared unfortunate. For example, during the 2000s a number of jurisdictions followed the 1998 decision to establish the UK FSA as an integrated supervisor, a move that is now widely viewed as misguided and that the UK reversed in 2011 with the shift to a twin-peaks framework.

Special resolution regimes for banks, outside of the US, are a more recent development that remains largely untested, although major shortcomings are already evident in the case of the euro-area banking union (Restoy et al 2020).

The advantages and shortcomings of each archetype are well known. Sectoral supervision offers apparent legal clarity and skill specialisation, but it is undermined by the blurring of sectoral boundaries – not least because of financial innovations such as derivatives and other risk transfer techniques – and the emergence of diversified financial conglomerates.

Also, a purely sectoral framework may struggle to provide effective conduct-of-business supervision if there are perceived trade-offs with prudential objectives, as often happens.

Integrated supervision ostensibly eliminates overlaps and gaps, since everything is brought under a single roof, but it has to manage different kinds of supervisory responsibilities that entail different cultures.

In particular, discretionary risk assessment for prudential supervision contrasts with a more rules-based compliance mindset for conduct-of-business supervision. These are either effectively kept separate in the integrated structure, thus creating silos, or brought together, with the likelihood that at least one important responsibility may be neglected, with catastrophic consequences.

The UK FSA is generally considered to have failed in its prudential role because of lack of sufficient focus on financial stability risks, which allowed the fiascos of Northern Rock, the Royal Bank of Scotland and other British banks that were exposed as fragile or unviable in 2007 and 2008.

The twin-peaks option is favoured by many academics and independent observers, but it does not eliminate coordination issues since the same financial firms are subject to supervision by multiple authorities with possibly inconsistent requirements29.

Furthermore, there are many links between prudential and conduct-of-business challenges, making the distinction often debatable. For example, financial crime or the misleading distribution of risky savings products are conduct-of-business violations, but they can also have significant financial-stability implications.

The question of whether to place the prudential supervision of banks with the central bank or elsewhere is similarly contentious. There are synergies between central banking and banking supervision, particularly for liquidity policy and financial-stability analysis, but there is also a potential conflict of interest between the two roles.

For example, a central bank that is also a banking supervisor may be tempted to pursue excessively accommodative monetary policy to mitigate perceived weaknesses in the banking system, in extreme cases to hide its own supervisory failures.

The UK went full circle on this issue, separating the FSA from the Bank of England in the late 1990s, then reintegrating prudential supervision in the Bank of England in the early 2010s.

In the EU, the ECB (2001) argued forcefully in favour of synergies between monetary policy and banking supervision, was initially overruled with the creation of the European Banking Authority, and was eventually vindicated with the establishment of the SSM in 2012-14.

The US maintains a hybrid model in which the Federal Reserve System plays a key role in the prudential supervision of banks, but is far from the only agency involved.

As for resolution authority, separating it from the main supervisor (albeit with ‘backup’ supervisory authority, as is the case with both the US FDIC and the EU Single Resolution Board), has significant advantages in terms of eliminating perverse incentives for supervisors to wait too long before taking action (‘supervisory forbearance’). But this separation also increases organisational complexity and the need for interagency coordination.

A jaded view is that any framework is bound to be found wanting at some point, and that reforms of supervisory architecture are political reactions to inevitable supervisory failures. This view, however, does not entirely match the record. In many cases, the supervisory architecture was changed not merely because the supervisor failed, but because specific pernicious supervisory incentives needed structural correction.

This was the case, for example, with the replacement of the UK FSA with a twin-peaks architecture, and with the replacement of national prudential supervision of banks with the SSM in the euro area, both decided in the early 2010s. Conversely, there have been a number of cases in which supervisors have ostensibly failed in their prudential mandates, but the architecture was not subsequently changed in a major way.

For example, the Netherlands did not reverse its adoption of a twin-peaks framework following a series of bank collapses between 2008 and 2012. The US adopted only incremental architectural changes following the so-called subprime crisis of 2007-08 – mainly the elimination of the tainted Office of Thrift Supervision and the transfer of its role to the OCC.

The US Financial Crisis Inquiry Commission (2011, page xviii), in its landmark report of January 2011, stated: “We do not accept the view that regulators lacked the power to protect the financial system. They had ample power in many arenas and they chose not to use it.”

The above-mentioned complexity of the US and EU financial supervisory architectures, the two most relevant benchmark jurisdictions for China, offers nuanced lessons.

One way to look at it is to recognise administrative and political inertia, and to observe that the streamlining of supervisory architecture in these two large jurisdictions has been extraordinarily challenging.

Another perspective is that the persistence of at least certain features of the supervisory architecture is positive for predictability and accountability, and that top-down disruption of existing structures would likely do more harm than good insofar as it undermines that predictability.

Table 1 summarises some key financial supervisory tasks in selected large jurisdictions, with much simplification. It highlights the complexity of the US and EU frameworks, relative to China and even more so to Japan and the UK.

As we have noted, it is improbable that China can beneficially adopt a supervisory architecture as streamlined as those of Japan and the UK, but it can aim to avoid the considerably greater complexity of the US and EU frameworks.

Table 1. Selected financial supervisory responsibilities in China, the US, the EU, Japan and the UK

Notes: CBIRC = China Banking and Insurance Regulatory Commission; CFPB = Consumer Financial Protection Bureau; CFTC = Commodity Futures Trading Commission; ECB = European Central Bank; FDIC = Federal Deposit Insurance Corporation; FSA = financial services agency; NCUA = National Credit Union Administration; OCC = Office of the Comptroller of the Currency; PBOC = People’s Bank of China; SEC = Securities and Exchange Commission; SSM = Single Supervisory Mechanism.

Source: Bruegel.

4 Policy considerations for China

China’s financial supervisory architecture should correspond to the specifics of its financial sector and broader policy system. No textbook architecture with theoretically clean divisions between all the different supervisory tasks will match all the financial-supervision challenges China’s authorities face. Streamlined frameworks that work reasonably well in smaller countries with less-complex financial systems would not necessarily function well in China.

Conversely, among jurisdictions of comparable size and complexity, neither the US nor the EU, both of which have multiple supervisory bodies, offer particularly useful templates for how to organise financial supervision in China.

China has modified its supervisory architecture in recent decades through incremental and tailored adaptation, driven by changes in its own financial system while taking into account the international context.

The 2018 reform made China’s framework more streamlined; as the original proponent of the twin-peaks concept noted, it “represents a further step towards the adoption of a Twin Peaks structure” in China (Taylor, 2021, page 31).

China’s reforms of shadow banking regulation in recent years, from shutting down P2P without broader financial instability, to reducing the risk of banks’ off-balance-sheet lending, are also indications that the existing setup can address supervisory challenges that cut across different types of financial institutions and markets. It remains to be seen, however, whether the steps announced in March 2023 will further improve supervisory effectiveness.

A full-fledged twin-peaks architecture would arguably be desirable, but it should be noted that the corresponding strengthening of the consumer protection task would constitute a significant policy inflection from the priorities of Chinese policymakers observed in the past.

Going further by consolidating all financial supervision in the PBOC as a single integrated supervisor would not solve the coordination challenges. Managing such a sprawling and unwieldy organisation with so many often-competing responsibilities would inevitably result in some tasks being undermined, as happened with the UK FSA in the early 2000s.

It is doubtful that such a setup would lead individual departments to coordinate better than recent practice between the PBOC and CBIRC.

Reshuffling the architecture in a major way may also have short-term downsides, especially at this juncture. It may add to already high uncertainty related to the coming renewal of China’s economic and financial leadership, an unclear growth outlook and continued stress in the real-estate sector.

Implementing a new architecture and completing the corresponding transition is likely to take several years. Known details about a new umbrella CCP organisation that would oversee all existing agencies, a change that has been signposted in addition to the reform announced on 7 March, are not specific enough for a confident assessment of how it might interfere with those agencies’ supervisory responsibilities and alter the incentives for better (or worse) supervisory consistency and effectiveness30.

This is not to downplay the scale of the challenges confronting China’s financial supervisors. As summarised in section 1, these include major governance concerns in supervised entities (such as oligarchic banks); operational coordination across different agencies or different departments within a single large agency; insufficiently clear divisions of responsibility that result in risk avoidance and blame shifting; corruption; and the fundamental difficulties of achieving good corporate governance and supervisory independence in China’s CCP-centred system.

None of these are clearly linked to a particular choice of supervisory architecture archetype, whether sectoral or twin peaks or integrated. Instead, to deal with such challenges, China should try to improve the operation of its financial supervision structure within a generally stable supervisory architecture.

Clarifying the responsibilities and mandates of the different supervisors and individual departments within them, by contrast, is a matter of high priority. There are too many competing and unclear mandates among China’s financial-sector authorities, leaving too much scope for blame shifting and blame avoidance.

In the event of a supervisory failure, it should be possible to identify unambiguously where the failure occurred. As for bank resolution specifically, experience in both the US and EU highlights the great advantages of a centralised, predictable system in which a single authority is in charge of decision-making, even for cases of failures of small banks (Gelpern and Véron, 2019).

Before announcing the supervisory reforms in early March, the Chinese authorities circulated a draft Financial Stability Law, which may be adopted in revised form later in 202331. While representing potential progress compared to the status quo, that text still suggested too many cooks in the resolution kitchen.

Whether the resolution authority is embedded in the PBOC or in the new NFRA, or is created as a new, separate institution, it should belong in one and only one central institution to avoid supervisory forbearance and to maximise efficiency in responding to future crises, which will inevitably happen even if only at local level.

The dominant role of unitary national authorities (PBOC, CBIRC/NFRA and CSRC) in China’s setup has some advantage over the more fragmented US and EU arrangements, in line with the objective of an integrated financial system that operates on a level playing field, and discourages supervisory arbitrage in which firms play different local supervisors against each other.

It would be unfortunate for China to jeopardise this advantage by assigning explicit responsibility to local authorities in resolution issues, even if such a move might help in terms of face-saving or expediency.

In that spirit, the draft financial stability law should be amended to assign clearer exclusive responsibility to central authorities in resolving the financial institutions they supervise, if they are determined to be failing or likely to fail.

The main challenge for China’s financial-sector policy remains its unfinished transition from a state-directed to a market-based financial system, and the way the CCP’s pervasive role creates obstacles to good corporate governance in individual financial firms and to the independence of supervisory authorities.

Too often, political authorities and sometimes the supervisors themselves intervene directly in financial firms’ capital and credit-allocation decisions, occasionally resulting in failures of risk control and risk management.

Chinese reformers should aim at a clearer and more rigorous division of responsibilities, in which financial firms manage financial opportunities and risks, and supervisors focus exclusively on their respective public-policy mandates.

No major changes to the current supervisory architecture, beyond incremental adjustments like that recently announced, are needed for that.


1. See for example Alicia García-Herrero, ‘Li Keqiang’s farewell points to employment as China’s major problem’, First Glance, 7 March 2023, Bruegel.

2. Xinhua, ‘The Political Bureau of the CCP Central Committee held a meeting and decided to convene the second plenary session of the 20th CCP Central Committee’ (in Chinese), 21 February 2023.

3. Reuters, ‘China to set up new financial regulator in sweeping reform’, 7 March 2023. In the previous two decades, comparable financial-sector reforms had been preceded by National Financial Work Conferences, held about every five years since 1997. In this cycle, a National Financial Work Conference was expected in 2022 (Wei and He, 2022) but was not held, and the decision was announced at the annual session of the National People’s Congress, China’s rubber-stamp legislature. See also Alicia García-Herrero, ‘China’s new regulator hints at a major clean-up of the world’s largest financial sector’, First Glance, 13 March 2023, Bruegel.

4. ‘Regulator’ and ‘supervisor’ are often used as synonyms in this area. We conform to international practice by referring to them as supervisors, while recognising that the extent and nature of their authority to enact binding rules (‘regulation’ in a narrow sense) vary considerably across jurisdictions. The administrative resolution of certain failing financial firms outside of the generally applicable court-ordered bankruptcy framework is a task that is in principle separate from both regulation and supervision, and has gained prominence in multiple jurisdictions in the past 10-15 years. ‘Supervision’ is occasionally used in this text as shorthand to encompass both supervision and resolution.

5. See for example Yiping Huang, ‘Remarks at the launch event of the 2022 CF40 Financial Reform Report’ (in Chinese), 25 April 2022.

6. Because US banks originate and distribute a lot of asset-based securities instead of keeping them on their balance sheet, the aggregate size of the US banking sector measured by total assets is significantly smaller than its peers in the euro area and China.

7. Including sovereign, sub-sovereign, agency and corporate bonds. Source: International Capital Market Association.

8. Source: The Banker annual Top 1,000 World Banks rankings.

9. The Shenzhen municipal government is among its largest shareholders, but with only a single-digit stake.

10. Source: The Banker database, authors’ calculations.

11. Although two of Australia’s four large banks are headquartered in Melbourne.

12. As documented by the International Financial Reporting Standards Foundation’s jurisdictional profiles (https:// www.ifrs.org/use-around-the-world/use-of-ifrs-standards-by-jurisdiction/), and the Basel Committee’s Regulatory Consistency Assessment Programme reports (https://www.bis.org/bcbs/implementation/rcap_jurisdictional.htm). The EU has been determined to be ‘materially non-compliant’ in the latter programme’s risk-based capital category, arguably the most important, and ‘largely compliant’ in three of the other four categories. By contrast, China has been deemed ‘compliant’ (the highest rating) in all five categories.

13. Wu Yujian, Zhang Yuzhe and Denise Jia, ‘PBOC Sets Up Deposit Insurance Fund Management Company’, Caixin Global, 30 May 2022.

14. Bloomberg News, ‘China’s peer-to-peer lending purge leaves $115 billion in losses’, 14 August 2020.

15. Pan Che, Fran Wang, Wu Hongyuran and Wu Xiaomeng, ‘China to Merge Banking, Insurance Regulators’, Caixin Global, 13 March 2018.

16. Wu Hongyuran and Han Wei, ‘How China Prevented a Local Bank Crisis from Snowballing’, Caixin Global, 17 August 2020.

17. Fitch Ratings, ‘China’s new rules on financial holding firms to curb systemic risks’, Non-Rating Action Commentary, 14 October 2020.

18. ‘Provincial’ is used here as shorthand for any mainland Chinese territory directly under the central government, namely the 22 provinces, but also direct-administered municipalities (Beijing, Chongqing, Shanghai and Tianjin) and autonomous regions (Guangxi, Inner Mongolia, Ningxia, Tibet and Xinjiang).

19. Michelle Price and Benjamin Kang Lim, ‘China Asks Britain for advice on Creating Financial Super-Regulator’, Reuters, 15 May 2016.

20. BaFin is the acronym for the Bundesanstalt für Finanzdienstleistungsaufsicht (Federal Financial Supervisory Authority), established in 2002 by the merger of several public bodies.

21. The expression “twin peaks” in this context was coined by Michael Taylor (1995) and is now widely used (Godwin and Schmulow, 2021).

22. Lingling Wei, ‘China’s Latest Plan for Market Control Involves the Central Bank’, Wall Street Journal, 14 July 2016.

23. See Deutsche Bundesbank, Cooperation with the Federal Financial Supervisory Authority, accessed 7 February 2023.

24. As a consequence of the UK’s departure in 2020, the euro area represents the overwhelming majority of the financial sector in the EU by almost any measure. For example, the euro area has represented more than nine-tenths of total EU banking assets continuously since early 2020, versus less than three-quarters when the UK was in the EU (source: European Central Bank Consolidated Banking Data series).

25. See CRS (2020) for a somewhat more detailed overview of the US supervisory architecture.

26. Seven of the 27 EU countries still have their own currency. Among these, Bulgaria is on a path toward euro adoption, with no certainty yet as to the final date. The 20 EU countries that have adopted the euro do not have an independent monetary policy; their national central banks exist as independent institutions that participate in the ECB-centred Eurosystem.

27. On the supranational integration of banking sector policy, known as banking union, see for example Teixeira (2020).

28. The three agencies are the European Banking Authority (EBA), European Insurance and Occupational Pensions Authority and European Securities and Markets Authority (ESMA), known collectively (and confusingly) as the European Supervisory Authorities (ESAs). ESMA has some direct supervisory responsibilities, for example for central counterparties from non-EU countries, credit rating agencies and trade repositories. One of the authors (Véron) is an independent nonexecutive director of DTCC Data Repository (Ireland), a trade repository directly supervised by ESMA.

29. Norman Blackwell, ‘Financial Regulation in the UK Is Ripe for a Serious Rethink’, Financial Times, 11 January 2023.

30. Keith Zhai and Lingling Wei, ‘China to Shake Up Financial System as Xi Jinping Installs Key Associates’, Wall Street Journal, 20 February 2023.

31. Fitch Ratings, ‘China’s New Financial Stability Law to Curb Contagion Risk’, Fitch Wire, 2 February 2023.


Chorzempa, M (2018) ‘Massive P2P Failures in China: Underground Banks Going Under’, PIIE China Economic Watch, 21 August.

Chorzempa, M (2022) The Cashless Revolution: China’s Reinvention of Money and the End of America’s Domination of Finance and Technology, New York: Public Affairs Press

CRS (2020) Who Regulates Whom? An Overview of the U.S. Financial Regulatory Framework, Congressional Research Service, Washington DC.

Desai, S (2016) ‘A Regional Comparison of China’s New Deposit Insurance System’, Pacific Exchange Blog, 24 February, Federal Reserve Bank of San Francisco.

ECB (2001) The role of central banks in prudential supervision, European Central Bank.

Financial Crisis Inquiry Commission (2011) The Financial Crisis Inquiry Report: Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States, Washington DC.

Gelpern, A and N Véron (2019) An Effective Regime for Non-viable Banks: US Experience and Considerations for EU Reform, study for the European Parliament’s Committee on Economic and Monetary Affairs.

Godwin, A and A Schmulow (eds) (2021) The Cambridge Handbook of Twin Peaks Financial Regulation, Cambridge University Press

Heilmann, S (2005) ‘Regulatory Innovation by Leninist Means: Communist Party Supervision in China’s Financial Industry’, The China Quarterly 181.

Hotori, E, M Wendschlag and T Giddey (eds) (2021) Formalization of Banking Supervision: 19th–20th Centuries, Singapore: Palgrave Macmillan.

IMF (2017) People’s Republic of China Financial Stability System Assessment, International Monetary Fund.

Lardy, N (1998) China’s Unfinished Economic Revolution, Washington: Brookings Institution

Restoy, F, R Vrbaski and R Walters (2020) ‘Bank Failure Management in the European Banking Union: What’s Wrong and How to Fix It’, Financial Stability Institute Occasional Paper 15, July, Basel.

Taylor, M (1995) ‘Twin Peaks’: A Regulatory Structure for the New Century, London: Centre for the Study of Financial Innovation

Taylor, M (2021) ‘The Three Episodes of Twin Peaks’, in A Godwin and A Schmulow (eds) The Cambridge Handbook of Twin Peaks Financial Regulation, Cambridge University Press

Teixeira, PG (2020) The Legal History of the European Banking Union: How European Law Led to the Supranational Integration of the Single Financial Markets, Oxford: Hart Publishing

Wei, H and J He (2022) ‘Prospects of China’s Latest National Financial Work Conference’, Briefing, 20 July, Anbound.

This article is based on Bruegel Policy Brief Issue no08/23 | March 2023.