2.1.1 The Drivers of NPL Formation within Chinese Banks

Banks have completely dominated China’s credit system from the start of the era of “reform and opening up” in 1978 until shadow banking began to take off in the 2000s. Banks have inherited a privileged position in China, shored up byFootnote ³³ :

1. 1 a captive customer base, since almost all lending until 2008 was done through banks, giving them a huge existing customer base, especially with the SOEs that dominated the economy;

2. 2 competition restrictions giving very few legal alternatives to bank deposits or bank loans in China’s highly controlled economy;

3. 3 implicit guaranteesFootnote ³⁴ —banks benefitted from a free implicit guarantee on their deposits, despite the lack of formal guarantees; this safety allowed banks to gather large deposit volumes despite paying low rates;

4. 4 controls on deposit and loan rates, as banks were subject to regulatory limits on the maximum deposit rate they could pay and the minimum loan rate they could charge; interest caps and ceilings could ensure healthy profit margins.

At the same time, the state imposed significant constraints. In principle, the most important constraints have been the followingFootnote ³⁵ :

1. 1 Controls on loan volumes—the PBOC set limits for total loan volume by banks in aggregate and these limits were then apportioned out to individual banks. Given bank dominance of the financial sector controlling the volume of bank loans meant controlling overall money supply, which was an effective way to run Chinese monetary policy. These restricted bank lending to the SME sector.

2. 2 Micromanagement of lending that was often directed by important government or party officials made things worse for the SMEs sector. This type of directed lending is rarer today, except for smaller banks that are closely tied to regional or local governments, though overall strategic directing of credit remains.

3. 3 Strict loan-to-deposit ratios that do not allow banks to lend funds over 75% of their deposits. This has proved to be a serious constraint on loan growth, especially as the growth of bank deposits has slowed down sharply in the face of competing investment products. As a result, regulators recently allowed some non-traditional deposit inter-bank market lending to be counted towards the 75% rate to relax the lending cap a little. A key function of fractional reserve banks is to offer liquidity on demand by employing deposits to make loans, keeping only a fraction of them in reserve.Footnote ³⁶ But bank loans (assets) can also be financed by more expensive means of funding, such as common equity and, in some cases, longer-term debt. If the cost of lending adjusts accordingly, then restricting bank lending by calculating it over a fraction of bank deposits is an inefficient means to restrict lending, especially as credit supply can migrate to the shadow sector.

4. 4 The high reserve requirements imposed by the PBOC, which seem to boost shadow lending and regulatory arbitrage. Banks are required to hold substantial reserves as deposits at the PBOC, earning rates far below what banks could earn on their normal business and even below their own cost of funds. This level varies, but currently about 19.5% of deposits must be placed with the PBOC. Deposits remain the major source of bank funding, so the reserve requirements represent a substantial cost in aggregate.

5. 5 State ownership—until 2004, the major banks in China were 100% owned by the state—which, by means of majority shareholdings, gives the government effective control of all major banks, particularly in personnel decisions, where the top management come from senior state positions and have not competed in a free labour market. So, their outlook is different from executives who have had private-sector careers. It is thus unsurprising that, even when directed lending was relaxed, banks kept favouring SOEs in their lending policies.

6. 6 Strong SOE lending bias due to risk factors—SOEs are also favoured because of implicit state guarantees and privileged favoured market positions, which make them safer borrowers. This tendency is reinforced by internal reward and punishment systems. A failed loan to an SOE is unlikely to be punished severely, while bad loans to the private sector can lead to job loss. Similarly, bank-manager career considerations make lending to entities run by powerful party members attractive.

7. 7 SME funding—a reverse bias. PRC banks have consistently shown an aversion to lending money to private-sector entitiesFootnote ³⁷ and especially SMEs. This bias is of critical importance for both the historical development of the shadow-banking system and the PRC’s future economic growth. More recently, SMEs have been the backbone of Chinese growth, reportedly providing 70% of employment and 60% of China’s GDP [gross domestic product] in 2012, while they only received about 20–36% of bank lending during the same year.Footnote ³⁸ While there has been a gradual improvement in terms of SOE-lending ratios, banks’ lending to the private sector still lags substantially behind lending ratios in other large economies—a bias that also affects to a significant degree bank profitability. The lack of long-standing and deep links between lenders and SMEs explains, in part, why the bulk of SME deposits go to the inter-bank market, where rates are not capped. But this places banks at a disadvantage when it comes to on-balance-sheet lending, given the aforementioned loans-to-deposits caps.

The combined effect of aforementioned factors and especially the bias to extend loans to SOEs without adequate credit controls has meant that the PRC banks accumulated a large number of NPLs in both the pre-2008 and the post-2008 periods. Although China’s economy was less affected than other countries in the region during the Asian financial crisis, due to the relatively closed nature of its economy and financial sector back then, PRC banks accumulated a substantial number of number NPLs during that period.Footnote ³⁹ These necessitated a number of reforms in 1998, which included: (1) recapitalization of state-owned banks; (2) adoption of international classification standards for NPLs; (3) enforcement of commercially viable loans to avert strategic defaults, whereby debtors default because the lender cannot effectively enforce the loan via the legal system; and (4) banning local governments from influencing lending decisions.Footnote ⁴⁰

Bank recapitalization was funded by the government’s issuing bonds valued at RMB 270 billion (Chinese yuan).Footnote ⁴¹ In 1999, four state-owned asset-management companies (AMCs) were established to transfer NPLs from the banks over a ten-year period.Footnote ⁴² Transfers of NPLs took place during 1999–2000, amounting to a staggering RMB 1.4 trillion—about 20% of the banks’ combined loan book, or 18% of GDP.Footnote ⁴³ It has been estimated that this was less than half of the total NPLs.Footnote ⁴⁴ After implementation of these measures, NPLs dropped to around 20% in 2000, but it took until 2004 for them to fall to a more manageable 13.2%; eventually, they fell to 2.4% in 2008Footnote ⁴⁵ as a result of the strongest GDP growth.

In 2004, the government decided to list the state-owned banks, starting the so-called “equitization” phase. To bolster the balance sheet of the Bank of China (BoC) and the China Construction Bank (CCB), which were floated first on the Shanghai and Hong Kong Stock Exchanges, the PBOC transferred RMB 320 billion in NPLs from the BoC and CCB to Asset Management Companies for 30–40% of the book value.Footnote ⁴⁶ The government also injected $45 billion of capital, which boosted capital-adequacy ratios and supported new lending that could offset the impact of existing NPLs on bank profitability.Footnote ⁴⁷

The original AMCs are still operating today, maintaining distressed asset sales as their core business. China Huarong Asset Management (CHAM) is the largest AMC, having absorbed 55% or 264 billion yuan-worth of bank NPLs in the first half of 2016, and expects the second half of the year to yield similar results.Footnote ⁴⁸ The Chairman of CHAM has noted that the quality of NPLs is deteriorating, making it harder to earn a profit, with the average disposal time being from one to three years.Footnote ⁴⁹

2.1.2 Tackling Bank NPLs Post 2008—a “Bottomless Pit”?

Tackling NPLs has become an increasing concern for the PRC in recent months. As discussed below, the central government had to essentially mount yet another expensive and covert bailout of banks in the course of 2016. This new period started after the 2008 financial crisis, when the central government decided to implement a RMB 4 trillion economic stimulus and loosen up monetary policy, in contrast with earlier periods when the government controlled tightly the money supply. The loosening-up of the monetary base was transmitted through the banking system in the form, inter alia, of lower interest rates and new lending.Footnote ⁵⁰ Such measures eventually became so widespread as to deviate from the original purpose, which was to use them as a monetary stimulus, and evolved into a replacement for a fiscal stimulus.

However necessary the stimulus was, relaxed standards of bank supervision and loan underwriting meant that the PRC did not avoid a new crisis of NPL accumulation, even though it had just finished dealing with the last one. To this effect, Arner et al. note that “as growth rates have started to decelerate and the levels of private and corporate indebtedness have substantially risen in the post-2008 period, this has led to a simultaneous and substantive increase in the level of assets prices and NPLs.”

With NPL ratios among China’s commercial banks increasing for 19 consecutive quarters, debt to GDP standing at approximately 225% and increasing, and corporate debt at approximately 145%, the IMF among others has raised concerns calling for rebalancing measures.Footnote ⁵¹ The IMF conservatively estimates that, in 2016, Chinese bank corporate loan portfolio losses could equate 7% of GDP.Footnote ⁵² Other sources note that official data show that NPLs at Chinese commercial banks jumped 51% in 2015 to a decade high of RMB 1.27 trillion amid a stock-market rout and the worst economic growth in a quarter of a century.Footnote ⁵³ This is more or less the official figure, as disclosed by the China Banking Regulatory Commission (CBRC) at the end of May 2016, which puts NPLs at RMB 1.4 trillion, or 1.75% of total bank lending, on the back of 18 consecutive quarters of NPL growth.Footnote ⁵⁴ In fact, according to the IMF’s April 2016 GFSR, more than 15.5% of a total USD 1.3 trillion PRC corporate debt may be at risk of default.Footnote ⁵⁵

However, even these figures are disputed by credit-rating agencies, which place capital shortfalls for commercial banks at 11–20% of PRC GDP.Footnote ⁵⁶ The information gap is in part due to two factors. First there is the matter of lending transparency and transparency of exposures when lending is sourced outside of the traditional banking system remains.Footnote ⁵⁷ Moody’s Investor Services has estimated that, in 2015, China’s shadow-banking system accounted for nearly 80% of GDP.Footnote ⁵⁸ The second related problem is that loans to the shadow sector are invariably mischaracterized as investments and thus, when they do not perform, they are not accounted as NPLs, which means any restructuring would not appear on balance sheets as loan workouts, even if the IMF’s recommendations about banks moving to an explicit recognition of corporate workouts are adopted.Footnote ⁵⁹

Either way, these are large figures even for the world’s largest economy and a continuous risk to the stability of the regulated sector. Thus, the government is addressing the mounting NPL and bad-debt problems through the adoption a multi-pronged strategy. First, for the debt related to/generated by state-implicit guarantees, it actively encourages mergers or bankruptcies of technically insolvent companies, introducing a debt-for-equity swap programme and allowing provincial governments to establish AMCs. With regard to steel and coal companies defaulting on their debt, the CBRC has issued draft regulations to set up a debt-to-equity swap programme. Each swap involves three parties: the originator (creditor) bank, the SOE debtor, and a third-party executor.Footnote ⁶⁰ Executors will consist of the big four banks, which will sell NPLs to AMCs. Upon receiving approval by the CBRC, distressed SOEs will exchange debt for equity, thereby strengthening their balance sheet at the expense of the purchasing bank’s balance sheet, before being sold to the AMC. Effectively, the debt is retained by the government until disposed of by the AMC, while, with each transaction through the chain, the NPL’s value is diminishing, spreading the risk among government agencies. Technically insolvent companies will be excluded from the programme.Footnote ⁶¹ The programme began its operations on 25 October 2016.

However, the success of this multi-pronged strategy is still unproven, for a variety of reasons. First, in practice, the transfer of NPLs to AMCs amounts to an expensive bailout of the four main banks and a shifting of bad debt from the provinces to the centre. Second, the suggested debt-to-equity swaps run the danger of keeping alive so-called zombie SOEs,Footnote ⁶² as the IMF calls them. These could accentuate the misallocation impact of debt overhang and deprive from new credit viable companies. Thus, the IMF’s report on the recent PRC programme for NPLs suggests a robust test of corporate viability, which would have as an effect to push unviable SOEs to bankruptcy rather than artificially shoring them up through debt-to-equity swaps.Footnote ⁶³ Of course, given the political clout exerted by SOEs, leadership and their regional government owners implementing the IMF suggestions in practice might prove a daunting task. Third, the debt-securitization solution may not be working due to a dearth of private buyers for bank NPLs, which means that sold assets end up with state-buyers or state-owned banks, essentially recycling rather than resolving the problem.Footnote ⁶⁴

Finally, given the aforementioned structural problems that also create uncertainty about the true value of NPLs, such as the difficulty to enforce contracts especially against SOEs, the development of a true secondary market for distressed debt, as the IMF rightly suggests, may be unattainable for now. That would mean that the central government would eventually have to absorb AMC losses into its budget. Therefore, in the face of slowing growth and mounting public debt and with banks haemorrhaging deposits, which entirely rules out the possibility of implementation of a bail-in regime, PRC authorities have clearly to shift attention to the prevention of NPLs.

This shift in policy becomes ever more urgent given the rising volume of bad loans in the corporate sector, the escalating inability to serve them due to falling profits (Figures 1 and 2),Footnote ⁶⁵ and the multiplication of housing loans granted by PRC banks in recent years. The latter have aided the significant rises in real-estate prices that clearly indicate a developing asset bubble in specific regional spots, rather than the country as a whole.Footnote ⁶⁶ Even so, the impact on bank solvency of a fall in housing prices and a string of defaults due to the slowing economy is expected to be significant.

Figure 1 IMF 2016 Source: IMF Working Paper (2016)

Figure 2 IMF 2016 Source: IMF Working Paper (2016)

2.2.1 Mapping the PRC Shadow-Banking Sector

There is a wide range of estimates about the size of the shadow-banking sector in China, depending on the definition and the kind of activity that should be included in the total.Footnote ⁶⁷ The latest reports put it at around 80% of PRC GDP.Footnote ⁶⁸ But there are doubts whether these moderate estimates are accurate due to lack of transparency. On the other hand, the PBOC has devised a key index: the Total Social Finance Statistic, which may provide an up-to-date measure of bank and non-bank financing, though it lacks in granularity.Footnote ⁶⁹ Yet figures may still be exaggerated due to the entrusted loans phenomenon. These are in the main loans made by parents of non-financial firms to subsidiaries that are run through banks for legal reasons.Footnote ⁷⁰ The banks are indemnified for the credit risk of the borrower by the non-financial firm.Footnote ⁷¹ The motive of those loans can be clouded. A large number of entrusted loans appear to be lent within a corporate group.Footnote ⁷² But some entrusted loans are funded by SOEs as a way of profiting from their ability to borrow cheaply and in large volume by on-lending the funds they obtain. It seems that all entrusted loans to non-affiliates are motivated by the pursuit of short-term profit and they surge at periods when the flow of credit from the regulated sector is restrained.Footnote ⁷³

In principle, lending within a corporate group would not have to be considered as a shadow-banking activity and indeed most loans to affiliates are to support a member of the group rather engage in interest-rate arbitrage.Footnote ⁷⁴ Yet loans to unrelated parties are on the rise, with large corporations using their stronger balance sheet to borrow money via Hong Kong banks to on-lend to smaller firms within China that face restricted access to PRC bank lending.Footnote ⁷⁵

Overall, it is difficult to separate intra-group and extra-group loans to unrelated parties.Footnote ⁷⁶ Moreover, even lending within the “group” in China may still be between entities that are only loosely connected and where the borrower may be allowed to fail without the failure of other members of the group. This classification problem leads to most calculations treating all entrusted loans as shadow-banking activity, even though this doubtless exaggerates the extent of shadow banking in China. At the same time, the true size of corporate on-lending may be grossly understated, as this activity is unregulated and unreported.Footnote ⁷⁷

Another development that impacts on the measurement of the PRC shadow-banking system is the new wave of Internet-based providers of payment servicesFootnote ⁷⁸ —essentially an integrated payment-collection service provided by the likes of AliPay and others,Footnote ⁷⁹ initially funded by merchant fees. What brings Internet providers within the boundaries of shadow banking is that payment-service providers have leveraged their knowledge of users’ credit profile and they now offer Internet-based peer-to-peer (P2P) lending,Footnote ⁸⁰ WMPs, and micro-lending to individuals and small loans to SMEs, alleviating the liquidity problems that the latter often face.

These firms are at the heart of the FinTech revolution that has disrupted the way that financial services are provided.Footnote ⁸¹ Given the vast chunks of data that Internet finance providers can control and mine, and the cheaper cost base compared to traditional banks, combined with the higher yields they offer, such a development may seem inevitable. But, as we see below, there are risks, especially with respect to P2P lending. Moreover, if FinTech firms do indeed grow to become the principal providers of retail funding, their systemic importance may increase very significantly.

2.2.2 Risks, Benefits, and Shadows

As already mentioned, Chinese shadow banking is largely the result of interest-rate and regulatory arbitrage, and is thus a system that remains dominated by banks, which use it to avoid capital requirements while making substantial returns on loans masqueraded as investment products or other forms of financing that are classified as receivables. Elliott et al. calculate that perhaps two-thirds of business into shadow banking is effectively “bank loans in disguise,” where a bank is at the core of the transaction and takes the bulk of risks and rewards, but pays non-banks to participate in order to avoid regulatory constraints and costs. According to the same study, the other third is due to cost advantages that non-bank institutions enjoy because of relaxed regulatory constraints and their willingness to lend to smaller players.Footnote ⁸² Essentially, a very large number of shadow-banking transactions such as the already discussed private-wealth products, loans and leases by trust companies, entrusted loans, bankers’ acceptances, leasing finance, guarantees through specialized firms, and TBRs involve the mediation of banks, and could even become a claim on the banks, depending on the circumstances. For example, in certain securitizations, the bank may substitute it for due payments to avert reputation risk.

We explain below how, apart from the wealth products and TBRs discussed in the introduction of this article, other forms of shadow-banking activity can also create a long chain of interdependence and interconnectedness between not just shadow operators and regulated banks, but also between regulated banks themselves,Footnote ⁸³ even though the exposure either does not appear on bank balance sheets or the risk of the transaction is not accurately reflected. These transactions expose regulated banks not only to the risk of counterparty insolvency, opening the gates for credit-risk migration to the regulated sector, but also to the risk of counterparty illiquidity that could, in the absence of a liquidity backstop, change into a generalized liquidity shock.Footnote ⁸⁴

In the next few paragraphs, we identify specific shadow-banking operations and practices that are particularly responsible for the generation of links of interconnectedness between the regulated and the shadow sector.

2.2.3 Interconnectedness, Credit-Risk Migration, and Liquidity (Refinancing) Risk

1. 1 Trust loans and leases are financial transactions undertaken by trust companies—a separately regulated type of firm that combines elements of banks and fund managers. Before a recent tightening-up of the rules of their operation, trust companies had wide latitude to operate across the financial sector. The key question in this case is how the trust company has sourced the loans it provides. If the money comes from the banks and the trust company is merely performing interest-rate arbitrage or acts as a bank subsidiary, there is ample scope for credit and liquidity-risk migration, though such (contingent) exposures may not be reported.

2. 2 Bankers’ acceptances are certificates issued by banks that promise unconditionally to make a future payment, usually within six months, and are generally backed in part by a deposit from the party desiring the bankers’ acceptance to be issued. In a way, they are no more than a certificate of deposit used to back commercial transactions such as purchases of inventory, where the seller receives the acceptance as a substitute to cash collateral. Acceptances are negotiable and may be sold by the holders at a discount rate prior to the maturity date, often back to the issuing bank. The fact that the acceptances are transferable can lead to multiplication of transfers of the same acceptance, transforming a modest deposit into a much larger one. Thus, borrowers can create money-multiplication spirals by using the acceptance to obtain a loan based on the discounted value of the acceptance and then use these funds to make a new deposit to back a further, larger acceptance. In this way, “acceptances” build leverage in the financial system and the economy, since the initial deposit has gradually turned into a large, off-balance-sheet loan from a bank. Since acceptances can be used as quasi-money, this practice leads to the formation of a parallel system of fractional reserve banking whose net exposure is hard to calculate, as it would require concurrent netting and set off of all acceptances.

3. 3 Interbank entrusted loan payments: these are loans that one bank makes to another bank’s client on the second bank’s behalf. As the entrusted payment matures, the first bank will receive principal and interest paid by the second bank. Naturally, this form of financing creates large levels of interconnectedness, making the first bank totally dependent on the solvency and liquidity of the second bank and not merely of the borrower.

4. 4 Guarantee companies participate in shadow-banking activities first by having as their core business the provision of financial guarantees, transferring the credit risk to the guarantee firm to reduce capital requirements for banks that make such guaranteed loans to the shadow-banking sector. Second, many guarantee companies make direct loans themselves, lacking a licence to do so. This practice increases interconnectedness between regulated banks and the shadow sector. As the final health and robustness of guarantee companies are unknown and it is unclear how they manage their exposures, any concurrent call on their guarantees, in the event of a string of defaults, will probably mean collapse of the guarantee companies and risk repatriation for the bank that tried to offload risk by using the guarantee, weakening its financial base.

5. 5 Inter-bank market activities act as a substitute for formal deposits using the inter-bank market, although the majority of participants are not banks, but large corporations using finance-company subsidiaries to participate. Under such arrangements, corporations can lend money to banks in deposit-like arrangements without being subject to the caps on deposit rates and without forcing banks to incur many of the regulatory costs of deposits, such as triggering the minimum-reserve requirements. However, any temporary withdrawal of such “loans” can cause a refinancing risk in the regulated sector, leading to calls for LoLR support.

Furthermore, the regulated sector may maintain invisible links with the shadow-banking sector, creating interconnectedness and channels for migration of liquidity and bankruptcy risk. In particular, the already discussed WMPs, which are predominantly short-term,Footnote ⁸⁵ can be the perfect conveyor belt for repatriating bankruptcy risks to the regulated sector and not necessarily through contractual or other explicit legal links,Footnote ⁸⁶ but even more likely through the contagion triggered by reputation risks and loss of confidence.Footnote ⁸⁷

Another invisible, in principle, risk-transmission channel relates to P2P-lending platforms and it is much less debated than the widespread frauds perpetrated by P2P operators in the first few years of the emergence of these platforms.Footnote ⁸⁸ P2P platforms in China currently operate under three models.Footnote ⁸⁹ First, the platform is used to simply match lending and borrowing information. Second, the platform may be supported by operator (or its affiliates) guarantees or credit-enhancement facilities or may merely sell such products. Third, the platform may offer repackaging and sale-of-credit assets through securitization and other financial-engineering techniques, whereby the P2P platform issues “WMPs” to raise money for purchasing credit assets from affiliates of P2P platforms. Fourth, the platform may facilitate the transfer of credit assets created by its operator or its affiliates to end investors—whereby the P2P platform or its affiliates extend certain loans to borrowers first and then sell these loans through the P2P platform to end investors.

The second, third, and fourth models amount to the provision of straightforward credit-intermediation services and, similarly, the third and the fourth form of P2P platforms may also offer covered or uncovered loans. Interconnectedness could be further accentuated if P2P platforms are used to sell bank-credit assets. In addition, the four types of P2P business models involve liquidity risks, as there may be liquidity mismatches if the platform takes in short-term investments from clients on a rolling basis to fund long-term assets in the pool. In this case, any confidence crisis may spill over to the regulated sector, even if there are no explicit and implicit links between the P2P operator and regulated-sector institutions.

Thus, it is paradoxical that the P2P platforms that intermediate credit, rather than just matching interested borrowers and lenders, are not licensed/regulated like banks. Simply put, the mere imposition of capital requirements on P2P operators is possibly insufficient, as it is unclear whether the practice of P2Ps’ extending loan guarantees and loan securitizations has abated even after the requisite prohibition of these activities in November 2016.Footnote ⁹⁰

Another development that may link Chinese banks with lending channelled through the shadow sector is the emergence of so-called “directed asset management plans [AMPs].” The new shadow channels work like trusts, with a bank investing proceeds from its WMPs in a directional plan that then lends to a borrower chosen by the lender. This extension of credit can circumvent restrictions on certain borrowers, like local government financing vehicles, as well as capital requirements. The amount of money placed in such products jumped by 70% last year to RMB 18.8 trillion (USD 2.9 trillion), outpacing the 17% growth for trust assets, according to data provided by industry groups.Footnote ⁹¹

Finally, the rise of money-market funds (MMFs) like the aforementioned Yu’e Bao is not free of concern and systemic stability implications. As discussed, given the close connection between China’s banking and non-banking credit sectors, it is not surprising that nearly 90% of Yu’e Bao’s portfolio consists of bank deposits, short-term debt paper issued in the inter-bank market, and cash deposited with depositary and clearing agencies which settle and clear money-market instruments.

While limiting their participation in securitization deals and other money-market instruments, many other MMFs also invest the majority of their portfolios with banks to earn more interest. This is because MMFs’ deposits are not subject to the aforementioned interest-rate caps and thus banks can offer MMFs a relatively higher interest rate, in particular at times when the inter-bank market is short of liquidity. In fact, some investors may have withdrawn their bank deposits to invest with MMFs in order to obtain a higher return.Footnote ⁹²

MMFs are a classic example of how the liquidity (and credit risks) borne by the shadow operators may migrate to the regulated sector. Chinese MMFs typically have a constant net asset value (NAV).Footnote ⁹³ But an MMF’s constant NAV may not reflect the real value of the MMF when the number of redemption requests is significant, since, in such a case, the MMF would be required to fast realize its investments (at possible discount prices) to meet the cash payout requirement. This redemption structure poses a liquidity risk for the regulated sector as well as the MMFs themselves. MMFs currently place the bulk of their portfolios with banks either as deposits or with the short-term bank paper. Moreover, funds placed in other (than deposits) money-market investments are maintained in the same pool. Yu’e Bao and many MMFs offer T+0 redemption for any amount not exceeding a specific threshold.

Even setting aside the real risk of firesales to meet a sudden surge of repayments/redemptions, a large MMF investor exit during any panic scenario, such as when they realize that the value of their investments in MMFs is far from guaranteed, can create a sudden withdrawal of MMF deposits from commercial banks to cover their liquidity needs. In these cases, MMFs can transfer unexpected liquidity shocks to the regulated sector. The best example of an MMF investor panic is the US Reserve Primary money-market fund debacle—the so-called “break the buck” incident.Footnote ⁹⁴ Therefore, systemic risk may increase due to MMFs’ higher susceptibility to “runs,” because they present a significant retail investor base and, at the same time, investors treat their investments with MMFs like principal protected deposits (as a result of constant NAV and T+0 redemption features).Footnote ⁹⁵ One solution for the risks that MMFs pose would be to adopt a floating NAV, since the fund’s assets are also floating at redemption. Another is to impose generalized “gating” restrictions at the discretion of the fund’s management. In fact, both measures should be adopted to stabilize the PRC MMFs even if they lose business. However, above all, what ought to be done is interest-rate liberalization to make bank deposits more competitive vis-à-vis MMF products.

2.3.1 Overview

China follows a silo approach to financial regulation, which has drawn a fair amount of criticism.Footnote ⁹⁶ Since 1993, China’s nascent financial sector was regulated by the PBOC, the central bank of China, which, in addition to monetary-policy powers, had the competence to formulate and implement fiscal policy, and to regulate and supervise the banking industry.Footnote ⁹⁷ Following the amendment of the Commercial Banks Law in 2003, the banking supervisory and regulatory responsibilities of PBOC were transferred to the China Banking Regulatory Commission (hereinafter “CBRC”). The move to create specialized regulatory bodies under the State Council in 2003 (the “three agencies one bank” system) signalled the Chinese government’s intention to modernize and consolidate regulatory structure in the financial sector by developing expert regulators. The PBOC remained the central bank and could still exercise considerable regulatory power over all financial institutions in China in order to maintain the overall stability of financial markets. It also retained monetary-policy competence, acting as LoLR.

The CBRC’s remit includes banks, financial-asset managers, trust and investment companies, and other depositary financial institutions. Its responsibilities include approving new banking licences, formulating prudential rules, and conducting compliance examinations. It also formulates supervisory rules governing the banking industry. The China Securities Regulatory Commission (CSRC) regulates and supervises the securities and futures markets, and enforces sanctions. The China Insurance Regulatory Commission (CIRC) regulates the insurance industry. It should be noted that the sectoral regulators do not enjoy hierarchical independence in the same mode as Western regulators, even if they enjoy operational independence. Rather, they are parts of the wider public administration machine and are often headed by career bureaucrats.

The sectoral regulators produce so-called second-tier laws consisting of administrative rules and regulations enacted by the State Council, the highest administrative authority of China. The third tier consists of CBRC’s Guidelines, notices, and rules. Most of CBRC’s regulatory initiatives fall into this category. In general, the third-tier regulatory initiatives serve as the base of China’s banking regulations and deal with contemporary regulatory issues.

2.3.2 Responsibility for Financial Stability and Fractured Monitoring Lines

Although overarching responsibility for financial stability resides with the State Council, the highest executive authority, the Law of the People’s Bank of China as amended in 2003, gives the PBOC responsibility to guard against systemic financial risk and maintain financial stability.Footnote ⁹⁸ During the global financial crisis (GFC), responsibility for financial stability was exercised by a high-level committee of key financial agencies, which was established to regularly assess conditions and consult on policy actions. Each of the agencies had contingency plans in place to respond to a crisis, including memoranda of understanding (MOUs) to promote co-operation. The efforts of Chinese authorities to rapidly deploy a range of macroprudential measures to cool off overheating property markets was testimony to the strong administrative capacity of the State Council and its ability to respond to near-term systemic threats. But, from an operational perspective, responsibility for financial stability is largely taken by the four agencies, though the centre-piece in this sometimes unwieldy institutional jigsaw is the PBOC. The Law on the PBOC specifically charges the bank with responsibility to guard against systemic financial risks and the maintenance of macro-financial stability. In addition, the PBOC implemented the Deposit Insurance Act in May 2015Footnote ⁹⁹ and various macroprudential policy tools. In particular, the PBOC put in place a Macro Prudential Assessment framework (e.g. differentiated reserve requirements) to guard against systemic risk. Moreover, the PBOC takes the lead in enhancing inter-agency co-ordination through the Financial Crisis Response Group directly under the State Council and the Financial Regulatory Coordination Joint Ministerial Committee (JMC).

Other government agencies, outside of the sectoral regulators that may have a role in safeguarding financial stability, are the Ministry of Finance, whose responsibilities include, inter alia, debt issuance and the management of state-owned assets, and the State Administration of Foreign Exchange, which has custody of foreign exchange reserves. Together, these two agencies provide the government with very substantial “financial backstops.” Currently, inter-agency co-operation and information-sharing in China are underpinned by a series of MOUs between the CSRC, CBRC, and CIRC. The PBOC also has developed Interim Rules on Information Sharing with the three financial regulators. In late 2013, these financial regulators set up a joint financial-stability co-ordination mechanism. Formal quarterly Joint Working Meetings of the regulators and joint reviews of financial intermediaries are held on a periodic basis. In addition, the Investor Protection Bureau was established in the PBC, CSRC, CBRC, and CIRC to strengthen the co-ordination of investor protection.

The institutional separation of functions among the PBC, CBRC, CSRC, and CIRC keeps, in theory, each institution’s focus on its mandate, helping to keep them accountable. In addition, the lack of institutional integration avoids creating a single dominant entity that over-concentrates powers and might miss key risks or fail to take important action due to “groupthink”—an accusation validly launched against the former single regulator in the UK, the Financial Services Authority (FSA). A key lesson of the GFC, however, is that inter-agency co-operation must be equally effective in normal periods, so that any build-ups of risks can be identified and addressed well before they evolve into a systemic event. In other words, an effective macroprudential policy framework is one in which financial agencies not only share their concerns on emerging risks, but also work closely together to prevent them.

China has made progress in improving its institutional arrangements for co-ordination in regulating the financial system during the past decade. Despite these steps, there remain gaps and regulatory overlaps. It is not clear that the existing protocol would always provide each agency with speedy access to the information from other sectors that is needed to assess the overall soundness of the financial systems. This is an important issue because legal hurdles and obstacles to information collection and sharing among agencies have proved to be major impediments to the early assessment of systemic risks in many countries.

With Circular No. 107/2013, PRC authorities clarified regulatory responsibility for the shadow-banking sector adhering to the sectoral or silo approach.Footnote ¹⁰⁰ But this approach is clearly flawed both vis-à-vis firm and product regulation multiplying oversight fragmentation with some products like WMPs being assigned to multiple regulators and innovative markets trading securitized debt having no clear regulator.Footnote ¹⁰¹ The most important risk of such fragmentation of oversight is, of course, that systemic-risk nodes and channels of interconnectedness between the shadow and the regulated sector may go unnoticed.

There are other key weaknesses in the silo approach. First, as the Financial Stability Board (FSB) notes, there is insufficient collecting and sharing of cross-sector granular data, which is a key shortcoming, since China has not established a national, credible data- collection system to gather more granular data. Second, there is insufficient cross-sector analysis on systemic risk. The PBOC devotes considerable resources to financial stability and publishes a comprehensive annual Financial Stability Report (FSR). But, as the FSB report notes, the FSR should include more analysis of the interconnectedness across sectors and markets in depth to explore the potential sources of risks. The key financial-stability issues should be discussed in depth to replace the list of financial developments and risks. Moreover, room exists for improvement in the comprehensive surveillance of systemic risk.Footnote ¹⁰²

Third, there is insufficient policy co-ordination. Several characteristics of the economy, financial sector, and policy framework in China make policy co-ordination a particularly important tool. For instance, the interest-rate policy is constrained by the relatively inflexible exchange-rate regime, as well as the gradual opening of the capital account. Fiscal policy—particularly government spending—also has limitations as a countercyclical tool. It is not always flexible enough to prevent credit booms and the build-up of systemic risk in the financial sector.Footnote ¹⁰³ Moreover, the importance of real estate as a major asset class for retail investment,Footnote ¹⁰⁴ the fact that underdeveloped financial markets provide limited risk-management tools, and the shortcomings in crisis-resolution frameworks highlight the need for policy co-ordination to identify and monitor systemic risk in the financial system. This is also the key theme of the FSB’s most recent peer review of the PRC financial system.Footnote ¹⁰⁵

As part of the reform debate and in order to streamline financial-sector regulation, several ideas have been mooted, including proposals for the creation of a super-agency for financial regulation placed under the PBOC. The super-agency is planned to co-ordinate monetary and financial supervisory policies and information-sharing between these regulators and to supervise legislation, without changing the roles of the incumbent institutions. Other ideas that have also been floated include the creation of a macroprudential council.Footnote ¹⁰⁶

In our view, super-agencies have at best a mixed record.Footnote ¹⁰⁷ Yet the observations above about inter-agency co-operation and information-exchange arrangements significantly strengthen the argument that the PRC needs a regulatory “big bang.” This in our qview should see the current silo approach replaced by a system that integrates the process of monitoring of financial-stability risks, microprudential risks, and investor-protection risks (and enforcement) without resorting to the establishment of a single mega-regulator—a solution that would carry serious shortcomings. For this reason, we suggest a configuration of PRC regulation along regulatory objectives that would be premised on the mixed approach rather than sticking to either the functional or institutional method of regulation. Our proposal model reflects a “twin-peaks”-plus model and it comes closest to the UK financial architecture, discussed in Section 3 below, although we envisage the macroprudential body as one with the power to order directly the different parts of the regulatory and financial system. In this context, we envisage that the augmented CBRC would become a microprudential regulator for the entire financial sector, with a merged CSRC and CIRC as conduct, market-abuse, and consumer-protection regulator.

In addition, the gaps in information-sharing necessitate the establishment of a body with overall responsibility for macroprudential policy in the PRC and binding powers and not just a cross-agency or even supra-agency monitoring mechanism, since the current arrangements in the PRC can easily open fractured lines.Footnote ¹⁰⁸ Thus, our model also includes a macroprudential body with full enforcement powers that sits within the PBOC (for the full proposal, see Section 5 below). An approach to financial regulation based on regulatory objectives like the one described above that would break down the silos would also offer regulators a much better idea of cross-sectoral connections and risks, including an accurate mapping of bank exposures in the shadow-banking sector. This way, regulators could gain a thorough view of shadow-banking risks. In addition, the consumer-protection body could pursue investor-protection infringements against both authorized and non-authorized persons. Finally, breaking down the silo approach would mean much less inter-institutional rivalry and turf wars that normally result in the blocking of important reforms.

3.3.1 Overview

3.3.2 A New Wave of Bank-Leverage Ratios

Partly due to the extended implementation timescale of the Basel III ratios and partly due to the fact that the Basel III leverage ratio of 3% of non-weighted assets is arguably very low, important financial centres have drawn up their own plans for imposing a leverage ratio. Many have chosen to go substantially higher than the Basel III 3% ratio.Footnote ¹⁴² The Federal Reserve was required to establish the enhanced risk-based and leverage capital requirements for “covered companies,” namely bank holding companies with total consolidated assets of at least USD 50 billion and non-bank financial companies that the Financial Stability Oversight Council has designated for supervision by the FRB.Footnote ¹⁴³ Banks and other financial organizations are required to report to the market and regulators on their leverage ratio at regular intervals under the PCA regulations. Given the striking failure of the previous American leverage ratio to warn of impending disaster, mostly because off-balance-sheet exposures went unnoticed, the Dodd-Frank Act has adopted a more sophisticated approach to bank-leverage measures.

The leverage ratio itself for all banking organizations is set at 4%,Footnote ¹⁴⁴ although certain large banking organizations are subject to a lower leverage ratio if they utilize the advanced internal ratings-based (IRB) approach to RWAs, which requires them to have a minimum total leverage exposure of 3%.Footnote ¹⁴⁵ However, a further supplementary rule has been introduced to cover the largest and most interconnected bank holding companies (BHCs), defined as those with more than USD 700 billion in consolidated total assets, or USD 10 trillion in assets under custody (covered BHCs). These banks would be required to maintain a Tier 1 capital leverage buffer of at least 2% above the minimum supplementary leverage ratio requirement of 3% (for banks which use the advanced IRB approach), for a total of 5%. Insured subsidiaries of BHCs (“Insured Depository Institutions”) must maintain at least a 6% supplementary leverage ratio to be considered “well capitalized” under the agencies’ PCA framework.Footnote ¹⁴⁶ Failure to exceed the 5% ratio would subject covered BHCs to restrictions on discretionary bonus payments and capital distributions. This rule would currently apply to the eight largest and most interconnected US banks, although it will not be implemented in its entirety until January 2018. Furthermore, there are additional requirements for leverage restrictions in times of crisis.Footnote ¹⁴⁷

The Canadian “assets to capital multiple” is a rather comprehensive leverage ratio, because it also measures economic leverage (to some extent). It is applied at the level of the consolidated banking group by dividing an institution’s total adjusted consolidated assets—including some off-balance-sheet items—by its consolidated (Tier 1 and 2) capital. Under this requirement, total adjusted assets should be no greater than 20 times the capital, although a lower multiple can be imposed on individual banks by the Canadian supervisory agency, the Office of the Superintendent of Financial Institutions (OSFI). This is more conservative than the previous US incarnation of the leverage ratio—and the inclusion of off-balance-sheet items strengthens the ratio even more. Indeed, the stringency of Canada’s leverage ratio has been cited as one factor—along with sound supervision and regulation, good co-operation between regulatory agencies, strict capital requirements, and conservative lending practices—contributing to the strong performance of its financial sector during the financial crisis.Footnote ¹⁴⁸

In 2008, the Swiss regulator FINMA (Financial Market Supervisory Authority), in strengthening capital-adequacy requirements, introduced a minimum leverage ratio under Pillar 2 of Basel II (supervisory review) solely for the country’s two biggest banks: Credit Suisse and UBS.Footnote ¹⁴⁹ The Swiss leverage ratio is based upon the Basel III definition of Tier 1 capital as a proportion of total adjusted assets and is set at a minimum of 3% at the consolidated level and 4% at the individual bank level.Footnote ¹⁵⁰ In fact, the Swiss government had considered raising the leverage ratio up to 10%—naturally attracting the wrath of the Swiss banking industryFootnote ¹⁵¹ —because it regarded the current 4% ratio as inadequate.

The UK’s Bank of England has announced a new leverage ratio, which will be binding on all banks regulated by the Prudential Regulation Authority (PRA) by 2018.Footnote ¹⁵² This followed on from work done by the Independent Commission on Banking (chaired by Sir John Vickers),Footnote ¹⁵³ which recommended that large UK ring-fenced banks ought to be subjected to a supplementary leverage ratio of approximately 4%.Footnote ¹⁵⁴ Accordingly, the important UK banks may have to comply with a minimum total leverage ratio of 4.95%, although this would only occur if the maximum CCB had been activated under the Basel III system. It is perhaps noteworthy that, notwithstanding these increases in the leverage requirement, there has been widespread relief amongst the UK banking industry that tougher regulations were not imposed.Footnote ¹⁵⁵

5.1.1 Overview

There is a widespread impression in the Western press that the PRC government either does not try or does not wish to radically tackle NPLs and diminish the financial-stability risks emanating from the shadow sector. Both views are wrong.Footnote ²²² The truth is much more complex. The PRC authorities have adopted a multitude of measures to tackle the financial-stability risks. A first wave of reform seems to have taken place between 1998 and 2005. But then, between 2006 and 2013, the pace of reform stalled.Footnote ²²³ It picked up again when, on 25 March 2013, the CBRC promulgated regulation on the use of wealth-management funds by PRC banks in order to strengthen risk management.Footnote ²²⁴ It requires commercial banks to match each wealth-management product with invested assets (underlying assets), and to set separate management, account, and book-keeping for each product, as well as a percentage limit on non-standardized credit assets in total underlying assets. Under Notice No. 8, commercial banks are required to match each wealth-management product with invested assets (underlying assets), and to set separate management, account, and book-keeping for each product, as well as a percentage limit on non-standardized credit assets in total underlying assets.

As regards the shadow-banking system, the CBRC has already implemented the following concrete measures: (1) banks are prohibited from providing guarantees or repurchase undertakings for the assets underlying the trust schemes; (2) trust companies have been subjected to capital requirements to ensure that their capital is compatible with the assets held under their trust schemes; (3) a 30% cap has been imposed on trust companies allocating assets under a bank-trust co-operation arrangement referring credit assets (loans and bonds); (4) all assets held by a bank through a bank-trust co-operation arrangement must be reflected in the bank’s balance sheet and therefore become subject to capital-adequacy requirements.Footnote ²²⁵

The CBRC promulgated the Regulation Governing Capital of Commercial Banks on 7 June 2012 and has moved to establish a living-wills (recovery and resolution plans) template for the largest institutions. Both measures mark a milestone for the regulation of the banking sector in China. As reported by the CBRC at the end of December 2013, the weighted average core Tier 1 capital-adequacy ratio of commercial banks (foreign branches not included) was 9.95%, increased by 0.14% year on year; and the weighted average capital-adequacy ratio was 12.19%, falling by 0.29% year on year. Moreover, in accordance with the Private Bank Guidelines, each wholly privately owned bank must have in place a “living will” for itself.

In addition, from July to October of 2013, a series of new policies in respect of cross-border RMB settlement were released by PBOC, including but not limited to the Notice of Simplifying the Cross-Border RMB Business Process and Improving the Relevant Policies and the Notice of Relevant RMB Settlement Matters for Overseas Investors to Invest in Domestic Financial Institutions. On 15 June 2015, the CBRC published the Amended Implementation Measures on Licensing Chinese-funded Commercial Banks, which were designed to further narrow the scope of licensing required for banking businesses, decentralize the power of the approval authority, and simplify the licensing procedures. This is regarded as a positive measure to invite private capital into the financial sector.

The CBRC Guidelines on Corporate Governance of Commercial Banks (hereinafter the Corporate Governance Guidelines) of 19 July 2013 require a sound organization structure, clear-cut responsibility boundaries, proper development strategy, good value criteria and social responsibility, effective risk management and internal control, an appropriate incentive, and disciplinary mechanism. Specifically, the Guidelines provide that the board of directors is accountable to shareholders and is ultimately responsible for the operation and management of a commercial bank. Thus, board members are required to have specific expertise and qualifications, and are subject to a remuneration structure that incentivizes proper risk management.Footnote ²²⁶ Moreover, the CBRC Guidelines place increased attention on internal controls within commercial banks, including effective risk assessment and effective procedures to safeguard assets and enable the bank to comply with applicable law. Commercial banks are required to establish independent risk-management departments, which shall be vested with sufficient power, resources, and a direct channel of communication with the Board.

Given that the discussed increase in levels of systemic risk is, in part, a by-product of the post-2008 loosening of credit standards to tackle the GFC, including a large expansion in credit that led to a worsening of average credit quality, the tightening of prudential regulations to force banks to build up reserves through higher provisioning for NPLs is a rather positive step. The same is the case with the adoption of more stringent capital-adequacy standards, placing limits on bank guarantees for corporate bonds, and tightening liquidity and diversification ratios. Banks have also been strongly encouraged to raise new capital.

To this effect, perhaps the most potent of this new wave of financial regulations is the promulgation of the Deposit Insurance Regulations. Under the Regulations, each bank shall maintain insurance coverage for money deposited with it and each depositor would be compensated for loss up to RMB 500,000 in the case that the accepting bank becomes insolvent or goes bankrupt. The Deposit Insurance Regulations came into force on 1 May 2015.Footnote ²²⁷

An even more potent example in the developing template for tackling and monitoring systemic risk is the macroprudential management framework (discussed in Section 2.2). The systemic-risk-monitoring framework includes the PBOC and the three regulatory agencies (CBRC, CSRC, CIRC) and consists of three layers: FCRG meetings to discuss and make decisions inter alia on financial-stability issues; risk surveillance, analysis, and prevention policies by individual agencies; and co-operation among them to arrive at an overall risk assessment.

5.1.2 Critique

Given this flurry of regulatory reform, the first question that may be asked here is why then the overall picture for financial-stability risks in the PRC has not markedly improved. A convincing explanation in this context may only be given by reference to fundamentals and these are the pace of innovation and the structure of financial regulation. Clearly, PRC authorities were caught unprepared to tackle the massive wave of innovation that marred the shifting economic paradigm, including the advent of electronic service providers and Internet-based businesses, and its spillover to a financial system that suffered the consequences of repression for at least two decades. The pace of the resulting transformation of the business paradigm and financial system structures and modes of service delivery and savings management has pushed regulators to the back-pedal, forcing them to engage reactively rather than proactively with emerging risks. Namely, unlike NPLs, where political pressure and policy preferences have historically played a large role in their accumulation, shadow banking developed at such a pace, with implicit bank guarantees probably playing a role for such acceleration, that the PRC regulators were left far behind.

Second, the silo approach to regulation and lack of clear lines of responsibility for financial stability prevent effective implementation of prescribed rules. On the contrary, they leave the gate open for restrictions to be continually circumscribed and that is not just as a result of a chase for yield, but also and very clearly due to the funding needs of certain economic sectors. Specifically, the FSB has identified several weaknesses to the financial stability, including to the followingFootnote ²²⁸ :

1. 1 A limited macroprudential toolkit (dynamic differentiated reserve requirements and window guidance) is rather limited. To this effect, as financial liberalization progresses further, the FSB raises the prospect of the PBOC turning into a frontline macroprudential regulator to fulfil its system-wide financial-stability mandate “by adapting its own tools and having a greater say in the application of other tools.”

2. 2 The sector-specific perspective of different regulatory agencies presents a considerable challenge to the overall calibration of macroprudential policy, “particularly as some measures may (for example) seek to loosen credit for countercyclical purposes while others seek to curb the growth in non-bank credit intermediation.”Footnote ²²⁹

3. 3 Information-sharing among regulators—the increased blurring of product lines across the financial industry with the concomitant potential for increased regulatory arbitrage underscores the need for enhanced information-sharing across agencies but, right now, the sharing of key information (e.g. results of stress tests and on-site inspections) is either restricted or shared if requested.

4. 4 Communication—the FSB notes that co-ordinating bodies like the FCRG and JMC do not have a separate communication policy and little information is publicly available on their deliberations. While there are good reasons to keep certain information confidential, on the other hand, public communication by these bodies can be a form of “soft” interventionFootnote ²³⁰ and also could form an important part of their accountability framework.

5. 5 Absence of clear policy framework on which market participants’ expectations about the policy could be anchored, intensifying occasional turbulence caused by policy uncertainty.

5.2.1 A New Regulatory Architecture

To remedy this piecemeal approach to systemic regulation and address the financial-stability challenges that the PRC currently faces, we suggest a four-pronged reform that is informed by the contemporary reform debate in China and the findings of impartial observers such as the FSB and IMF, and incorporates international reform experience (discussed in Section 3.1).

5.2.2 Evaluation of the Proposed Capital Reforms and the Question of Adverse Selection

This paper has explained why properly calibrated leverage ratios may prevent wasteful use of capital and make financial institutions more resilient. They control asset substitution and risk shifting. Essentially, banks have to increase the size of their capital cushions, no matter the risk-weighting of the relevant exposures and in proportion to the net increase in value of the bank assets. An unweighted and all-encompassing leverage ratio is simple to apply and monitor, and eliminates regulatory arbitrage, namely the bank’s ability to engage in RWA and the more particular forms of regulatory arbitrage in the PRC to avoid the already discussed lending and interest-rate restrictions. It would thus restore confidence in the financial health indicators of Chinese banks, putting a stop to frequent runs on their shares.

It is also argued that a binding and all-encompassing leverage ratio will sharply reverse incentives when it comes to bank lending, rather than the current ex post requirements, which, in an environment of lending biases and limited accountability for bank management, might lead to adverse selection that is discussed below. PRC banks became subjected to wide leverage ratios from 2013, with further overhauling of the ratio without raising capital requirements in 2015. So they are already able to meet higher leverage ratiosFootnote ²³⁶ but, even if they were not able to do so, raising private equity funding can only increase private-sector investment in PRC banks,Footnote ²³⁷ the quality of risk management, and governance of PRC banks, which are both state-dominated and state-manager-dominated, leading to severe lending biases. It will also provide a very strong incentive to bank leadership to appoint and promote executives from the ranks of private-sector professionals with strong backgrounds in credit-risk management. In addition, it will provide a check on smaller banks’ risk on management procedures, which seem to be not much different than the loose processes followed by shadow banks in an alleged race to the bottom.Footnote ²³⁸

Arguably, the proposed boosting of bank-capital structures and funding sources will incentivize: (1) shareholder controls, as it would be their equity that would be at stake if a loan becomes an NPL and not just creditors’ (depositors’ and bondholders’) money; (2) the shoring-up of management accountability for lending standards; (3) the implementation of sound risk-management procedures; and (4) the tightening-up of bank governance. These changes will augment and speed up implementation of declared regulatory objectives in the PRC banking sector, like those included in the aforementioned CBRC Corporate Governance Guidelines.

We have already explained (in Section 4) that empirical research on the implementation of leverage ratios, which entail higher equity buffers for banks, has not shown them to have an appreciable adverse impact on economic growth. There is no reason to believe that the results would be substantially different in the PRC, especially if any shortfall in funding due to higher bank-capital constraints is made up by an increase in risk-equity capital and other forms of equity funding. So the final major question that any cost–benefit analysis of the proposed introduction of a leverage ratio has to answer is the issue of adverse selection.

It is argued that higher leverage restrictions do not always reduce risk for the bank. Even if the risk of individual assets remains constant, bank managers may be incentivized into choosing assets with more highly correlated returns.Footnote ²³⁹ Accordingly, as this argument goes, leverage ratios may encourage banks to increase the riskiness of their asset portfolio, not decrease it—a classic Goodhart’s Law outcome. This concern was precisely the rationale for seeking risk-sensitivity in the Basel framework in the first place.Footnote ²⁴⁰ In many ways, reluctance to regulate leverage stemmed from a fear that it would affect economic growth without making banks safer, since banks would evade the “crippling”-profit consequence of a strict leverage ratio by focusing on high-return projects, which are riskier in most cases. Namely, as this argument goes, leverage ratios are bound to create perverse incentives.Footnote ²⁴¹

The fact that leverage ratios do not distinguish between different types of bank assets on the basis of their riskiness could mean that banks felt encouraged to build up relatively riskier balance sheets or to expand their off-balance-sheet activity.Footnote ²⁴² In addition, the lack of risk-weighting in the calculation of leverage ratios could penalize prudent banks holding substantial portfolios of highly liquid, high-quality assets. Accordingly, as this argument goes, because leverage amplifies gains from both risky and riskless investments, riskier assets generate higher returns, and RoE targets incentivize asset risk.Footnote ²⁴³ Research on bank performance during the GFC strongly suggests that asset write-downs were positively and significantly related to asset volatility and leverage.Footnote ²⁴⁴

Yet adverse selection does not seem likely, if PRC regulation of loan securitizations and issuance of asset-backed bonds are tightened. Securitizations can be a key form of loan refinancing; banks will strive to put on their books better loans, not worse, in order to be able to sell them. Second, if banks are forced by means of capital-adequacy regulation to keep “skin in the game” (e.g. a co-investment stake in the pool of varied size), their lending will be less reckless, even if they plan to slice and securitize the loans, since some exposure to the pool of underlying assets shall remain. Thus, good lending could be carried on, while underwriting standards are enhanced.

Finally, another argument against the possibility of adverse selection is that a binding leverage ratio will shift the attention of supervisors from dealing with ex post outcomes of bad lending to monitoring closely the behaviour of the ratio and thus increasing the chances of successful early intervention and preventive action when the signals turn negative. It also incorporates market discipline. Leverage ratios may remedy weaknesses of creditor monitoring of banks in the face of complexity and opacity of bank balance sheets, as a bank that will need to boost its capital buffers to meet the ratio would have to be profitable and/or of good repute so that it could do so (meeting the ratio) through the sale of quality assets, retained profit, or from the market.