China’s historically high rate of economic growth has gone hand in hand with even faster financial sector development, and its financial system now contains the world’s largest banks, second-largest stock market, and third-largest bond market. Ten years ago the system was relatively conservative and almost completely dominated by banking, but major fiscal stimulus after the global financial crisis and an innovative and competitive financial sector have changed that picture.

China’s historically high rate of economic growth has gone hand in hand with even faster financial sector development, and its financial system now contains the world’s largest banks, second-largest stock market, and third-largest bond market. Ten years ago the system was relatively conservative and almost completely dominated by banking, but major fiscal stimulus after the global financial crisis and an innovative and competitive financial sector have changed that picture.

Yet, with the wide range of financial products that has emerged, the system has become much harder to understand. And the increasingly complex financial system calls for constant effort to upgrade financial sector regulation and supervision.

This chapter reviews the financial system overall, scrutinizes the banking sector, and looks at the rapidly growing nonbank sector and products. It then discusses the equity, bond, and insurance sectors, in that order. The chapter concludes with policy recommendations.

Overview of the Financial Sector

Despite the changes of recent years, China’s banking sector continues to dominate the financial system. Credit is still primarily channeled through banks. In addition, many years of high economic growth, a high savings rate, and a closed capital account (see Chapter 8), have supported the sector’s rapid expansion. Between 2010 and 2015 the size of banking sector assets doubled from RMB95 trillion (234 percent of GDP) to RMB199 trillion (295 percent of GDP) (Figure 10.1). By asset value and Tier 1 capital, China’s banks now include four of the world’s five largest.

Within the sector, large, state-owned commercial banks dominate, while state-owned commercial banks’ dominance has declined amid ever-increasing competition for funding. The following are the main types of institution:

Figure 10.1.
Figure 10.1.

Financial System Structure

Sources: CEIC; China Banking Regulatory Commission; Morgan Stanley Research; People’s Bank of China; WIND; and IMF staff calculations.
  • State-owned banks. The big four (four largest banks) are the Industrial and Commercial Bank of China, China Construction Bank, Agricultural Bank of China, and Bank of China. The big four continue to dominate the banking sector, although their combined share of total sector assets declined from 50 percent at the end of 2010 to 40 percent at the end of 2015. All four are publicly traded onshore and in Hong Kong SAR, but the Chinese government—through either the Ministry of Finance or Central Huijin Investment, a state-owned investment company—remains the majority shareholder in all of these institutions. In addition, the Bank of Communications is structurally similar, but given its somewhat smaller size, observers normally refer to the big four, rather than the big five.

  • Joint-stock commercial banks (JSCBs). JSCBs constitute the next tier down. Twelve such banks jointly accounted for 18.6 percent of total bank assets at the end of 2015.1 JSCBs, with a wider range of controlling shareholders, have a more distant relationship to the central government than the big four. Most of these banks have expanded beyond their initial regional home bases to operate nationally.

  • Development and policy banks. China Development Bank and two policy banks—Agricultural Development Bank and the Export-Import Bank of China—accounted for roughly 10 percent of total bank assets at the end of 2015. These banks provide loans to rural and urban areas on projects related to economic, trade, and infrastructure development. They are fully state owned.

  • City-commercial and rural banks. The banking system includes an array of smaller, third-tier city-commercial and rural banks. The 145 city-commercial, 468 rural, and 122 rural cooperative banks have assets amounting to 25 percent of the banking sector. Except for those in big cities, such as the Bank of Beijing and Bank of Shanghai, they are generally small but have strong local networks. The ownership of these banks also varies: many have extensive, or even majority, private ownership, but local governments often play a key role through key ownership stakes or indirect influence.

  • Foreign banks. Forty-two locally incorporated foreign banking institutions operate in China, but their combined assets are small, accounting for less than 5 percent of total bank assets.

Meanwhile, nonbank financial intermediaries—trusts, fund management companies, and fund subsidiaries—have also grown rapidly in number and assets in recent years. These intermediaries have collaborated with banks to channel credit to sectors or borrowers that usually face difficulty gaining credit through bank loans. The ownership of these firms varies, with many privately and some publicly owned or with controlling shares held by local or provincial governments. They include the following forms:

  • Trusts. Sixty-eight trust companies collectively managed RMB16 trillion in assets as of the end of 2015 (KPMG 2015). They offer three main products: single fund trusts, assembled or collective trusts, and property management trusts.2 Trust companies still dominate nonbank financial activities, although the growth of their assets under management moderated recently (Figure 10.2).

  • Securities companies. As of the end of 2015, 125 securities companies collectively managed RMB6 trillion in assets, mostly on behalf of clients, and with relatively small balance sheets. Aside from their normal brokerage activities, they also provide credit to the financial system through two types of asset management plans (directed or collective). The size of directional asset management plans, in particular, has grown in recent years.

  • Fund management companies. Also as of the end of 2015 the 101 fund management companies provided products that are largely divided into two types: public offering funds (mutual funds), with RMB8 trillion under management, and special accounts, with RMB4 trillion.

  • Fund management subsidiaries. First established in 2012, fund management companies’ subsidiaries have expanded rapidly, with 78 fund management subsidiaries having RMB8.6 trillion by the end of 2015, up from less than RMB1 trillion in 2013. Fund management subsidiaries offer one-to-one (directional) and one-to-many (multiclient) products.

  • Asset management companies. The four asset management companies (AMCs)—Cinda, Huarong, Great Wall, and Orient—were established in 1999 to purchase nonperforming loans from large banks. They have expanded on their core business managing distressed assets to provide a wide range of financial products. In addition, concerns about rising nonperforming loans have led the government to issue provincial-level AMC licenses in more than 20 provinces (Box 10.1).

The links among these entities have become increasingly tight and complex. Competition for deposits and regulatory arbitrage have motivated banks to cooperate with nonbank financial institutions in search for higher yields.3 Accordingly, banks directly or indirectly finance a large portion of investment products created by nonbank financial intermediaries. Midtier listed banks and smaller unlisted banks have particularly high exposure to investment products issued by trust companies (such as trust beneficiary rights), as well as asset management plans (such as directional asset management plans) issued by securities companies and fund management companies and their subsidiaries. Banks also have exposure to shadow investment products through their off-balance-sheet wealth management products (WMPs)—investment vehicles that offer fixed rates of return well above deposit rates (that were previously regulated). WMPs also allow banks to get around “window guidance,” which restricts banks from extending loans to certain risky sectors.

Figure 10.2.
Figure 10.2.

Financial System Structure

Sources: CEIC; China Banking Regulatory Commission; China Insurance Regulatory Commission; Morgan Stanley Research; People’s Bank of China; WIND; and IMF staff calculations.Note: SSE = Shanghai stock exchange; WMP = wealth management product.

Bad Loans in China

Not all troubled debt in China is eventually classified as nonperforming. Chinese banking supervision recognizes several categories of loan quality, with nonperforming being the lowest, though these categories are to some extent qualitative. In addition to nonperforming loans, some loans with credit risk are described as “special mention” loans, some of which are eventually reclassified as nonperforming loans (NPL). Loans to companies for which recovery is deemed to be impossible are classified as NPL, but this is the end of a long process, and not all loans remain on the balance sheet for this classification.

To avoid this process, banks employ various techniques for minimizing at-risk exposures on their balance sheets. Regulations on bank lending are relatively strict, but those on trust companies and some other lenders are significantly looser. In some cases, banks will set up a single-unit trust that lends money to the company. The bank can then invest itself in the trust, booking it as an investment. While regulations forbid this procedure with a bank’s own loans, banks can invest through wealth management products in collective trusts, which they can then buy.

Once a loan is nonperforming, banks have limited options. But over the past decade, an important option has been selling NPLs to asset management companies (AMCs). During the banking crisis beginning in the 1990s, China established the four large AMCs discussed earlier in this chapter. These companies still buy troubled assets from bank balance sheets, but are reluctant to do so at face value, while banks are reluctant to recognize losses, reducing the flow of assets into AMCs. To cover their own cost of capital and operations absent the influx of new troubled assets, AMCs have increasingly moved into other industries, such as fund management. The perception that competition in this sector is insufficient is one reason why provincial AMCs have been established in recent years. The ownership and capitalization structure of these AMCs differ across provinces, with some having large ownership shares by local governments or their holdings companies, and others having more private sector participation. While flows so far have been small, these AMCs increasingly bid against the four national AMCs for troubled debt, but the long-term viability of their business model is not yet clear.

Two more transparent methods for addressing at-risk debt gained increasing prominence in 2016. First, while it is illegal for banks to directly hold shares of companies, this rule has been relaxed in an increasing number of cases in which debt has been swapped for equity. In other countries this has been a successful way to improve the management of companies and work out the stock of debt. In China’s case, however, many potential participants are state-owned enterprises (SOEs); changing management would thus have to be part of overall SOE reform. Second, official securitization of troubled loans has been discussed, though there are fewer cases as yet. Successful securitization depends on transparent pricing and finding a market to purchase such debts. The experience of securitized low-quality mortgage debt in the United States before the global financial crisis shows that this process carries substantial risks.

Capital markets have grown in recent years, although their role in the financial sector is still relatively small compared with other major countries. Bond and equity markets have undergone recent phases of rapid growth and their importance in the financial sector has increased over time. The total combined market capitalization of China’s mainland exchanges more than doubled between 2010 and 2015, from RMB26 trillion to RMB53 trillion, and is now the world’s second largest. Similarly, the outstanding balance of corporate bonds increased from RMB4 trillion in 2010 to RMB14 trillion in 2015. Banks hold the majority of bonds, both on balance sheet and through WMPs.

Finally, insurance companies are also growing rapidly—total insurance industry assets increased from RMB5 trillion in 2010 to RMB12 trillion by the end of 2015. In addition, insurance companies have been shifting their assets from bank deposits or bond investment to equity or other investments.

New forms of lending—such as peer-to-peer (P2P) lending platforms and internet banking (called fintech)—are emerging rapidly. Development of these new lending platforms has stirred technological innovation while alarming regulators after a number of large failures. Each platform is discussed below:

  • P2P. P2P platforms have grown very rapidly since 2007, although growth has slowed recently due to a number of fraud cases and stricter regulations. The number of P2P platforms increased from 10 in 2010 to more than 2,500 in 2015. The loan balance increased from RMB100 billion at the end of 2014 to RMB440 billion at the end of 2015. Borrowers include both small and medium-sized enterprises (SMEs) and retail customers. Aside from interest charges, P2P platforms also charge borrowers a transaction fee, which depends on the borrower’s credit risk. P2P lenders have offered very high returns—initially above 20 percent—to attract investors. However, yields on offer have fallen since mid-2013, and reached about 10 percent by the end of 2015. While some P2P platforms have institutional investors, most platforms work with retail investors. The average duration of P2P loans was about seven months at the end of 2015. Most P2P platforms, in principle, do not take any credit risk (as they simply match borrowers with lenders), although they have often worked with guarantee companies and other lenders to protect investors’ interests.

  • Internet banking. Several leading e-commerce companies have created subsidiaries to provide financial services to their consumers and SMEs that provide products. Chinese e-commerce giant Alibaba has created its financial subsidiary—Ant Financial—to provide internet and mobile-phone banking services. Alipay—a commonly used payment service in China—allows consumers to shop online, transfer money, buy services, and even make investments with a small amount of money through Yu’ebao, a wealth management fund affiliated with Alipay. The growth of Yu’ebao demonstrates the speed China’s financial system can develop. Following its inception in mid-2013, within one year Yu’ebao had 85 million customers. Today, Alipay overall has more than 450 million users (the largest in the world), and Ant Financial was valued (after a private fundraising round) at about $40 billion in early 2016. Similarly, Chinese e-commerce company JD.com established a financial subsidiary—JD Finance—to extend financial services including consumer finance for online purchases; loans to the suppliers of products, which are often SMEs; crowdfunding; and wealth management. JD Finance had raised RMB7 billion in financing as of early 2016.

Banks: Lending and Deposits

The composition of assets and loans varies among different types of banks:4

  • State-owned banks. The big four banks have a larger share of mortgage and corporate loans and a smaller share of interbank loans than other types of banks (Figure 10.3). These banks have traditionally provided loans to large and medium-sized enterprises, especially state-owned enterprises (SOEs) (see Chapter 11). A large share of loans goes to the transportation and utilities sectors and the manufacturing industry. Industrial Commercial Bank of China’s loans are particularly highly concentrated in the transportation and utility sectors (roughly a quarter of its loan portfolio).

  • JSCBs. These as a whole have a smaller share of corporate and mortgage loans than the big four banks, but have a larger share than city-commercial and rural banks. The shares of consumer loans and investment in securities have increased. These banks generally provide a larger share of their corporate loans to SMEs compared with the big four banks. The loan portfolio of JSCBs seems to be more evenly distributed across industries than that of the big four banks—JSCBs generally a provide a higher share of loans to the real estate and construction sectors, and wholesale and retail trade.

  • City-commercial and rural banks. City-commercial and rural banks have a smaller share of corporate, mortgage, and consumer loans than their bigger counterparts. Instead, they hold a larger share of their total interest-generating assets—roughly one-third—in securities. City-commercial banks, in particular, have invested more aggressively than their larger counterparts in nonstandard credit assets to expand their businesses amid fierce competition. These banks play a more vital role than their bigger counterparts in making loans to local micro and small enterprises. Having an advantage in knowing local businesses, their loan portfolios have a relatively high exposure to wholesale and retail trade.

Several other salient considerations deserve attention:

Figure 10.3.
Figure 10.3.

Bank Funding

Sources: CEIC; China Banking Regulatory Commission; Morgan Stanley Research; People’s Bank of China; WIND; and IMF staff calculations.Note: WMP = wealth management product.1 Sample covers all products issued by banks covered in WIND.

The average rate of return on assets is higher for smaller banks. In the first half of 2015 average return on assets ranged from 1.1–1.3 percent for the big four banks, and was about 1 percentage point higher for a select group of JSCBs and city-commercial banks. Higher average return on assets for these banks are produced by several factors: (1) higher rate of return for loans, (2) higher rate of return on investment (including financial assets held under resale agreement and receivable investment), and (3) lower share of deposits at the central bank (which has the lowest yield).

Bank funding depends heavily on deposits, supported by high domestic savings, although the share of deposits in total bank funding has declined recently. China’s share of deposits in total banking liabilities is high, at above 70 percent at the end of 2015, or 185 percent of GDP, and is among the highest in the Group of Twenty countries. Total deposits in the banking system increased from RMB68 trillion in 2010 to RMB125 trillion in 2015. At the same time, the share of deposits in banking liabilities declined from 80 percent in 2010 to 73 percent in 2015. In particular, interbank funding, including nonbank financial intermediaries, increased from 12 percent of total funding in 2010 to 17 percent in 2015, concentrated mostly in JSCBs and other smaller banks. This has been supported by the rapid development of an interbank market, discussed below.

Large state-owned commercial banks generally have an advantage over joint stock and city-commercial banks in securing lower-cost demand deposits. The funding sources for the big four banks are largely retail and corporate deposits, while JSCBs rely less on retail deposits and more on corporate and interbank and nonbank funding. The access to cheaper retail funding sources has allowed the funding costs for the big four banks to remain below the costs for JSCBs and other smaller banks.

Because the big four are able to attract deposits at lower cost, joint-stock and some other banks have relied on WMPs to fund growing balance sheets. To compete with the big four, JSCBs and city-commercial and rural banks have used WMPs to attract funding. Before the formal liberalization of deposit rates, WMPs allowed those smaller banks to offer rates above deposit ceilings on their short-term investments. Those banks, in turn, have targeted higher-yielding and often unsecured lending to offset the higher cost of funding. Even though the liberalization of bank deposit rates in late 2015 allowed banks to compete for deposits more freely, banks seem to continue to take advantage of WMPs to invest in high-risk and high-return sectors, since window guidance and higher capital and provisioning requirements work to discourage direct lending.

Nonbank Loans and Other Products

The 2008–09 global financial crisis saw the volume of nontraditional financial assets explode in China. Shifts in regulatory and supervisory priorities, caps on interest rates, and regulatory arbitrage as the authorities clamped down on risky products and excessive credit growth in nonfavored sectors had the effect of shifting the financial system away from traditional lending (Figure 10.4).5 This has accelerated more recently as banks have become more concerned about the rising number of defaults in the bond market and restructurings of bank loans, further driving weaker firms toward the nonbank financial sector. For these reasons, much of the new debt created in China in recent years has either come from nontraditional lenders or is bank lending that has been transformed through financial engineering into investment assets. These assets are held either directly on bank balance sheets or on the balance sheets of special purpose vehicles financed by investment products sold by banks.

Figure 10.4.
Figure 10.4.

Nonstandard Lending and the Nonbank Financial System

Sources: Bloomberg L.P.; CEIC; national authorities; WIND; and IMF staff calculations.Note: WMP = wealth management product.

The result is a complex and opaque system linking borrowers and lenders. Much of the borrowing under this system has gone to local government investment projects, overcapacity sectors, and real estate. These borrowers are generally rationed from direct bank lending, due to either restrictions on lending to unfavored sectors, capital, or provisioning requirements, or (in the past) limits on loan-to-deposit ratios. These incentives have pushed lending to these industries off of banks’ loan books. In many cases, loans to these industries have been repackaged into investments, which banks either sell to clients or keep on their own balance sheets. Funds for these investments until recently largely came from WMPs, but wholesale funding through China’s rapidly growing interbank market has grown in recent years. Between borrowers and lenders are a web of complex intermediaries, whose financial links to banks and borrowers vary widely.

An important systemic vulnerability of this arrangement is its direct link to the banking sector. Banks are estimated to have as much as RMB16 trillion of these nonstandard assets on their balance sheets, almost one-third of their overall assets, and off-balance-sheet exposures are also very large. While supervisory rules require banks to treat these assets for most purposes as carrying the same risks as the underlying products, this is not always straightforward (for example senior tranches of products with minimal credit history or assets with implicit guarantees). In some cases, this can result in underprovisioning or inadequate capital coverage. Liquidity risks from such products are also a serious concern, as the nontransparency of the system, widespread perception of implicit guarantees, and the strong presence of retail investors, who do not expect losses, raise the risk of runs. Deposit insurance can mitigate this risk to bank balance sheets,6 but a loss of confidence in WMPs or other products indirectly linked to banks could lead to a run in these areas, with unforeseeable results. Finally, the complexity and opacity of these products present a serious challenge for bank supervisors, who must ensure that increasingly complex structured products are adequately provisioned for.

Lower-quality creditors appear to be the main borrowers under this system. As in many countries, the highest-quality borrowers issue bonds and tend to rely on larger banks. In China, state-owned and other very large enterprises are seen as the highest-quality borrowers, with joint-stock and city-commercial banks instead lending to local government SOEs and private companies with good records. Nonbank borrowers tend to be companies rationed away from this system, either due to credit quality or due to regulations that limit lending to unfavored sectors, such as real estate, overcapacity industries, or infrastructure. These form the bulk of borrowers in the nonbank system.

A key group of borrowers under this system is special purpose vehicles established by local governments to finance infrastructure projects. A major part of China’s stimulus package during the global financial crisis was infrastructure investment pursued at the local government level. China’s local governments have the main spending responsibility, but the central government has the main authority for collecting taxes (see Chapter 6 on local governments). As a result, during the global financial crisis, local governments were forced to find new ways to mobilize resources for infrastructure investment. So-called local government financing vehicles (LGFVs) were a primary tool. These included many variations, but the most common was special purpose vehicles, which raised funds through bank WMPs and lent the funds to an infrastructure firm that paid off the loans after being paid by the government. Regulations gradually clamped down on this system beginning in 2014, and some provisions of the 2015 Budget Law (aimed at eliminating local government guarantees of special purpose vehicles and at shifting infrastructure investment toward public-private partnerships that take their risks) were intended to end this practice.

Many types of firms have moved into this disintermediated financial system. As deposit rate liberalization squeezed bank margins and the slowing economy reduced profits elsewhere, firms looked to the rising equity market and continued rapid growth of credit, and more firms began to move into financial services. The expanding availability of wholesale funding has reduced the importance of bank deposits in financial intermediation, catalyzing this process. The range of firms now involved in the system includes the following:

  • Fund subsidiaries of banks, which have about 10 percent of GDP under management, and issue WMPs to investors and then purchase various products.

  • Trust companies, which establish entities that invest in the securities market. These exposures are largely off the balance sheets of trust companies and are financed through wealth management products set up by banks.

  • National AMCs, which have moved into asset management as the profitability of their traditional business of disposing of troubled bank assets has declined.

  • Insurance companies, which invest in complex products to increase returns.

Despite this diversity and range of financial entities engaging in lending, a number of common threads emerge:

  • Nonbank intermediaries in general assume no credit risk for the loans they make. Most intermediaries act only as channels for investment, taking funds from institutional investors, banks, or high-net-worth individuals and investing them in products. However, some high-risk products are directly sold to retail investors, or are repackaged or tranched into assets that can then be sold, with senior tranches going on bank balance sheets and equity tranches to high-net-worth individuals or institutional investors. Some (for example, trust companies and securities companies) have their own balance sheets, but these tend to be dwarfed by off-balance-sheet exposures that the nonbank financial institutions do not guarantee.

  • Complexity and opacity add to regulatory challenges. Banks’ WMPs in many cases turn over rapidly; banks’ on-balance-sheet investment products can include a wide range of direct loans, restructured fixed-income products, or equity investments in special purpose vehicles; and interbank borrowing is short term, over the counter, and complex. While supervision aims to minimize maturity mismatches and ensure adequate capital coverage for banks’ risks, this complexity is highly challenging.

  • Mispricing of credit risk. Companies with weak profitability or in overcapacity sectors and local governments constrained by the 2015 Budget Law would still like to (or need to) borrow. The expectation that the central government will bail out local governments creates a perception that they are low-risk borrowers. Moreover, companies deemed strategic by local governments can also continue to borrow, despite poor financial prospects.

  • Regulatory arbitrage is an important motivation. As banks move from investing deposits in loans to investing funds raised through WMPs or the wholesale market into investment products, the regulatory framework can shift. In some cases, capital or provisioning coverage may be insufficient, or, for off-balance-sheet WMPs, perceived guarantees can raise reputational risks that may lead to losses. Even in cases where banks do not carry these exposures on balance sheet, as in the case of WMPs or assets held by bank subsidiaries, the fact that retail investors are almost never made to bear losses can create a self-fulfilling equilibrium where banks, for reputational purposes, may feel obligated to bail out investors in nonguaranteed products.

Supervisors in China are aware of these risks, and have moved to contain them (Table 10.1). These measures have aimed to improve risk management by banks, reduce leverage and complexity in structured products, limit the risks from WMPs, reduce maturity mismatches, and eliminate the implicit guarantees (perceived or otherwise) that pervade the sector. But the sector is extremely innovative and keeping up with the constant pace of transformation, and creative exploitation of loopholes has been challenging.

Table 10.1.

Supervisory Measures Aimed at Reducing Shadow Banking Risks

article image
Source: Moody’s Investors Service.

Intermediating Products

Trust Loans and Products

Trust companies are important providers of loans. They began to grow rapidly following the global financial crisis, as banks’ ability to lend was constrained and concerns rose about infrastructure investment. Trusts were seen as a way to conserve bank capital but continue to support investment.

Trusts were a crucial link to the first phase of shadow banking in 2013. They often were established to support LGFVs and other special purpose entities that allowed financial activity to migrate off bank (or trust) balance sheets. These items, generally financed by banks’ wealth management products, could invest in infrastructure, equity, bad loans, or other products, and often provided guaranteed rates of return well above regulated deposit rates. Many of the excesses of these products were quickly reined in, however: guaranteed products are now required to be on banks’ balance sheets,

Here, too, supervisors have moved to rein in risks. Beginning in 2013 regulations on capital requirements for trust companies—even on assets for which they assume no formal credit risk—began to fully bind. Additional regulations on trust companies’ liquidity have further reduced their ability to lend. This, unfortunately, occurred as the stock market began to surge rapidly, encouraging trust companies to shift their investments into equities.

Nevertheless, trusts are still important players in the nonbank lending system. With balance sheets increasingly tightly supervised, most trust-related exposures are off-balance-sheet products. These are generally single-unit trusts, which match a single investor (a high-net-worth investor or institutional investor, though WMPs are often classified as institutional investors) with a pool of loans. These account for about 57 percent by value of total trust company products. In these cases, credit risks may be high if the quality of the portfolio is low, but with only a single investor, the risks of contagion are contained. On the other hand, trusts also sell collective trusts, in which many investors buy into a pool of loans or other assets. These products—constituting about 32 percent of trust company products—may also carry high credit risk, but here liquidity risk is more salient than with single-investor trusts, since investors in a nontailored product may not roll over funding, or if permitted, may withdraw their investments in a stress episode. The trust company may then have to step in, which, if such withdrawals were widespread, could lead to systemic liquidity problems.7

Investment Receivables

Investment receivables is a broad category that includes many products for which the owner has residual profit rights. These are normally held as an “investment” item on banks’ balance sheets, but are occasionally held and classified as securities repurchased through the interbank market or as reverse repos, with bankers’ acceptances used as collateral. Two of the most common investment receivable items are trust beneficiary rights and directed asset management plans. These products are similar, but trust beneficiary rights are investment products sold by trusts, while directional asset management plans are sold by securities companies.

To construct these products, a trust company or other nonbank financial intermediary makes a loan to a borrower, occasionally with credit enhancement provided by a bank. The nonbank financial institution packages these loans into an investment product and sells it to a bank, potentially the same one that provided credit enhancement for the original underlying. If the resulting trust beneficiary right/directional asset management plan is entered on a bank’s balance sheet, supervisory rules require it to carry the same risk weight as the underlying product, though provisioning requirements are significantly lower. Investment products also did not count against the implicit quota set by loan-to-deposit ratios; these are now gone, but it is not yet clear how new liquidity requirements will affect incentives to hold them.

However, a second step of financial innovation with investment receivables can further reduce capital coverage. By structuring receivables and having the tranches rated, nonbank financial institutions can sell the banks products with a lower risk weight than a package of unstructured corporate loans. Corporate loans themselves—and thus under supervisory rules, any asset composed of such loans—would attract a risk weight of 100 percent on banks’ balance sheets. But a nonbank financial institution can repackage the loans into trust plans, asset management plans, or other innovative financial products. If the senior tranche of these repackaged products can be rated at AA- or higher, the resulting investment receivable can be brought on banks’ balance sheets at a risk weight of only 20 percent. This may be a reasonable level of capital coverage, but the experience of subprime loans in the United States shows that even products with a long history of default can be underpriced. In China, there is no such history, and a growing role for market forces means that defaults should rise in the future. The equity tranche can then be sold to banks’ off-balance-sheet WMPs—which do not have capital requirements—with a high return and no explicit guarantee by the bank, to high-net-worth, or to risk-seeking institutional investors.

Beyond trust beneficiary rights and directional asset management plans, similar products can be sold by other financial institutions. Insurance companies and fund management companies increasingly engage in direct lending or invest funds in structured products with underlying loans, while banks operate increasingly important fund subsidiaries that hold equities and invest in such products. The proliferation of direct lending and asset management by almost the full range of financial firms in China raises the stakes for supervisors, who not only have to understand complex products held on balance sheets, but also must assess the potential risk from off-balance-sheet products that could be covered by entities if there is a run or contagion from liquidity or credit shocks.

Entrusted Loans

Since companies in China cannot generally provide loans to each other, banks sometimes broker “entrusted” loans between firms. In China these loans are generally made to companies that have been somehow rationed out of the formal banking system, either because they are in overcapacity or otherwise troubled industries or because they are perceived to be too high risk. The stock at the end of March 2016 was RMB11.6 trillion, up by 19.6 percent from March 2015. Much entrusted lending happens between firms in the same conglomerate, but about one-quarter is between unrelated firms. In general, loans between unaffiliated enterprises tend to bear higher interest rates.

Evidence exists that some entrusted lending comes from firms borrowing from banks, only to onlend to unrelated firms through entrusted loans. Analysis conducted by the Hong Kong Monetary Authority shows that Chinese firms’ financial assets tend to be positively correlated with their financial liabilities, rather than negatively, which one would expect if firms borrowed only to finance investment, rather than to finance lending of their own. This relationship is particularly strong among SOEs, but the China Banking Regulatory Commission (CBRC) has become increasingly concerned about this practice, and is working to reduce it.

Bankers’ Acceptances

Chinese banks also use bankers’ acceptances to provide credit to borrowers. In most countries, bankers’ acceptances provide borrowers with financing in a manner analogous to working capital.8 In China, supervisory rules require them to be tied to a real underlying transaction, such as an international trade transaction or an investment. However, as the cash position of Chinese corporations has deteriorated but the web of implicit guarantees surrounding many sectors has remained, bankers’ acceptances have become a way for banks to finance the continued operation of firms with low profits but that are nevertheless seen as unlikely to go out of business.

The growth of bankers’ acceptances in China was partly motivated by regulatory arbitrage. Since recipients of bankers’ acceptances must place a deposit (generally for a significantly smaller amount) at the issuing bank, they become a way for banks to increase deposits while also providing credit to high-risk enterprises; bankers’ acceptances thus took up less space under banks’ loan-to-deposit limits. In the past, bankers’ acceptances were valued at a discount close to benchmark one-year loan rates. However, as the practice has become more widespread, and as bankers’ acceptances have come to be traded among banks and companies, discounts have grown and differentiated based on issuers and borrowers.

As with other products, financial innovation and speculation have led to increased demand for bankers’ acceptances. Banks will provide bankers’ acceptances to firms, which then sell the acceptance to a nonbank financial institution at a discount for cash. The company can then invest the cash in a trust product, earning the spread between the rate of return on the trust product and the rate implied by the discounting. In these cases, while the original bankers’ acceptance may have been priced at a relatively high level to cover a low-risk trade transaction, the actual underlying transaction—nonbank financial institution purchase of potentially speculative assets—can be substantially riskier.

Bankers’ acceptances are also occasionally repackaged by nonbank financial institutions to evade supervisory requirements. Bankers’ acceptances themselves are categorized as loans for supervisory purposes. However, if a bank discounts a bank bill and sells it to a nonbank financial institution, that institution can in turn sell the bank a structured product with the acceptances as an underlying asset. This product, since it contains loans, carries the same risk weight as loans, but, since it is classified as an investment, does not count against the bank’s loan quota or loan-to-deposit ratio, so the transaction frees loan quota for the bank.

Other Products

Beyond these products, there are other forms of nonbank finance that are generally of less importance. Finance companies, leasing, and consumer credit companies are formal parts of the nonbank financial system that also provide financing either for consumers or investment projects. Beyond the formal financial system are pawnshops, informal lending, and traditional forms of lending, which are still used in some circumstances by SMEs or firms rationed out of the formal financial sector. Finally, emergent growth areas such as P2P and Internet banking are increasingly important.

Nonstandard Funding Sources

Wealth Management Products

When China still had ceilings on deposit rates, banks established WMPs to invest in riskier projects that could pay a higher rate of interest to investors. Following the publication of the CBRC’s Regulation #8 in 2013, which limited the ability of WMPs to invest in shadow bank products, and the full lifting of deposit ceilings in 2015, their role began to diminish. However, the infrastructure remains convenient for banks (and other financial institutions) to channel savings into particular investments. They are thus analogous to the money market mutual funds that developed in the United States to circumvent interest rate ceilings, capital charges, and Federal Deposit Insurance Company fees in the 1970s, but have remained an important part of the financial system.

CBRC regulations state that guaranteed-principal WMPs must go on bank balance sheets, while nonguaranteed ones (where owners take the credit risk) can be put off balance sheet. As recently as the end of 2013 guaranteed products accounted for slightly less than half of the total, but by the end of 2015 this had fallen to only one-quarter, with floating return products now accounting for 73 percent of the total stock. Nevertheless, most WMPs are held on balance sheet, and many hold relatively low-risk assets, such as bank deposits and bonds. However, the stock of off-balance-sheet WMPs is large, and these do raise important concerns. In practice, investors have in some cases been compensated by the banks that issue the WMP for losses on such products. Here too there are similarities with the U.S. financial system, where reputational risks pushed banks to compensate investors in nonguaranteed products during the global financial crisis. The share of off-balance-sheet items (including non-WMP items) in total liabilities is particularly large for joint-stock banks.

Despite the lifting of deposit ceilings, WMPs continue to grow very strongly. At the end of 2015 there were nearly 61,000 WMPs in China, with a total value of RMB23.5 trillion. WMPs are issued by all types of banks, but the largest and fastest-growing share is from joint-stock banks, which account for 42 percent of the total stock.

Traditionally, a majority of WMP investments came from retail investors, but the share has been gradually falling (to half of December 2015 flows), with “interbank” investments becoming more and more important (to about 12 percent), and institutional investors comprising about one-third. Traditionally, WMPs were “closed,” with a time-specific payoff, which was generally linked to a particular investment. However, “open” WMPs, where buyers can buy in and out with fewer restrictions, now account for slightly less than half of the total.

At the same time, the average maturity of WMPs has been increasing, but remains short term. Products with maturities greater than one year account for only about 10 percent of the outstanding stock. Since some WMPs are invested in longer-term assets, this can lead to mismatches. While such mismatches are an inherent part of the banking business, banks can direct income from one maturing asset to cover liabilities invested in another. WMPs, however, match investors’ funds to specific investments, meaning that funds have to be rolled over each time to avoid illiquidity, raising the risk of a potential default on WMPs. So far, no WMP has failed to repay investors, however, as banks or other investors have stepped in to cover potential losses. But this blurs the line between on-balance-sheet guaranteed products and off-balance-sheet products that are not explicitly guaranteed, but for which investors may expect the same treatment that others have received.

WMPs invest in a wide range of products. About half of WMP assets at the end of 2015 were bonds, with another 22.4 percent in bank deposits and cash. Equity accounts for another 7.8 percent of WMP assets, a total of which has been depressed both by the sharp stock market decline in 2015 and by the reallocation of investors away from equities toward fixed-income assets. Following concerns about excessive risk concentration in WMPs, investment in nonstandard credit assets, which can include distressed or suspect debt, is capped at 35 percent and now accounts for 15.7 percent of total assets.

However, not all fixed-income products held by WMPs are innocuous. As the stock market boomed in 2014, some WMPs bought equities, but offered high fixed returns to shareholders based on expected dividend income. These WMP investments were in some cases categorized as fixed income. In addition, private placement notes, which are cash bonds issued by companies (potentially including weak companies in overcapacity sectors or local SMEs with poor profitability) on the interbank market to qualified investors, have risen in popularity due to recent regulatory shifts. Private placement notes are both less transparent and less liquid than regular bonds, but since they are classified for supervisory purposes as bonds, they are not subject to the ceilings on nonstandard credit assets that would apply if, for example, trust loans directly made to those companies were put on WMPs’ balance sheets.

Finally, nonstandard credit assets held by WMPs now total RMB3.7 trillion and are in many cases complex and nontransparent. Among these, RMB1.1 trillion are in structured products such as trust beneficiary rights/asset management plans, while another RMB620 billion are trust loans. Other items—such as entrusted loans, repos, and other assets banks classify as “interbank market assets”—are smaller. These products are often repackaged corporate loans. In some cases, they are nonperforming loans restructured by nonbank financial institutions and then bought by WMPs, but many are simply corporate loans that banks have removed from their balance sheets to free space for lending under loan-to-deposit restrictions, to reduce provisioning requirements, or for other reasons. In some cases, these products are structured; this allows risks to be diversified, but undermines transparency. Senior tranches, as highly rated investments, are easily sold, but WMPs aimed at high-net-worth investors or risk-seeking institutional investors will also buy junior tranches. With poor risk pricing in the corporate sector, and the number of defaults rising, these tranches are particularly risky. Investors may not fully understand the potential downside risks, but since WMP investors have often been compensated by issuing banks, this may not be a concern for many.

Wholesale Market and Interbank Finance

The WMP infrastructure built up while deposit rates were regulated is gradually being supplemented with a more sophisticated system of wholesale funding through the interbank market. While China’s banks remain less dependent on wholesale funding than banks in other regions, this is gradually changing as the repo and bond markets deepen. Interbank borrowing in China happens largely on the China Foreign Exchange Trade System platform, which is also used for bond trading. The bond market is discussed in more detail later in the chapter, which focuses on bank liabilities incurred in the interbank market.

Banks and other financial intermediaries can easily access short-term finance through repo on the interbank market. The key link to the nonbank lending sector comes through collateralized repo, which constitutes the majority of trading in the interbank market. The terms of repos in this market are set bilaterally between buyers and sellers; that is, haircut, yield, and eligibility of collateral are determined largely on a case-by-case basis. Tenors on this market tend to be very short: the average tenor at the end of 2015 was 2.7 days, down from 4.2 days at the end of 2010. Interest rates on overnight repo in early 2016 were about 2.1 percent, about 100 basis points above three-month certificate of deposit rates. The range of accepted collateral has also widened in recent years, as occurred in the United States in the years before 2008, raising concerns about adequate valuation, though discounting (haircuts) are widespread. In addition, much interbank borrowing at short tenors is uncollateralized; this raises the same potential maturity mismatch risk—from banks that invest repo’ed funds in longer-term assets—as collateralized borrowing, but raises counterparty risks. Uncollateralized short-term borrowing is particularly common among those smaller banks most invested in unconventional products.

As in all wholesale funding markets, there are potential collateral and counterparty risks. In most cases, public sector bonds (including policy banks) are used as collateral, lowering potential risks. Counterparty risks, however, are a more serious concern, as are risks from the valuation of collateral underlying over-the-counter repo transactions. All these risks can be contained under tight supervision, but the rapid growth of the market itself is a strain on resources.

Repos and uncollateralized short-term borrowing are only one part of a very diverse spectrum of assets banks can buy and sell through the interbank market. In addition to short-term lending to other bank and nonbank financial institutions, banks sometimes classify as “interbank investments” holdings of special purpose vehicles invested in nonstandard credit assets. If financial intermediaries have their products structured and then rated, they become standardized assets and can be sold on the interbank market. These carry more favorable risk weights than the underlying products might attract if they were held without being repackaged, analogous to the highly rated senior tranches of structured mortgage products in the United States before 2008.

Nonbank financial intermediaries also use the repo market to add leverage. Banks can invest WMPs—or even reinvest investment receivables—with nonbank financial intermediaries. These nonbank financial intermediaries can invest in bonds, which can then be repo’ed out, with the proceeds further repurchased (haircutting, of course, reduces the amount that can be repo’ed out with each step). The resulting product can be tranched, with very high risk resulting for the equity tranche, but with senior tranches having sufficiently low leverage that they can gain high ratings. As corporate defaults rise, in a context of widespread underpricing of risk, uncertainty about bankruptcy proceedings, and a lack of clear methods for valuing firms and collateral, this structure carries substantial risks, for both credit and liquidity.

Other Sources of Finance

Many formal participants in the disintermediated financial system raise their own funds. For example, insurance companies collect premiums and securities firms invest funds on behalf of their clients; funds from both of these sources are occasionally invested in nonstandard credit assets. Entrusted loans are made by corporates, though in some cases, companies in less profitable industries that are still perceived as low risk are able to borrow from banks, and then onlend to riskier companies through entrusted loans. Trust companies also have off-balancesheet clients’ accounts that do not use WMPs. In some cases, these are single-fund trusts, where a single investor is tied to a single product. In others, these are collective trusts, where numerous investors are brought together on a portfolio of projects. While all of these businesses are growing rapidly, their contribution to the market for nonstandard credit assets has not dramatically increased in recent years.

Other nonbank entities—P2P lenders, pawnshops, and so on—also make loans to companies or individuals. However, they generally intermediate directly, and are discussed in detail below.

Bond Market

China’s bond market has grown rapidly in recent years and is now the third largest in the world. In reality, there are two markets: the exchange-traded market and the so-called interbank market.

China’s significantly larger interbank market is much more than a straightforward bond market. Participants in the interbank market include not only banks, but also trusts, insurance companies, and money market funds. Banks are net borrowers from the system, while investors seeking high returns, including highnet-worth individuals, institutional investors, and retail investors investing through WMPs and other intermediary products, are net lenders to the market. The disintermediated nature of the market makes it very difficult to avoid mismatches between banks’ short-term interbank borrowing and longer-term lending, and to accurately price credit and liquidity risk.

Most of the issuance in the bond market is by large companies and public sector entities. Bonds issued by the government and policy banks account for a large share of the bonds outstanding (about 54 percent). More recently, however, local governments have been allowed to issue bonds. Established only in 2014, over RMB3.8 trillion in municipal bonds was issued in 2015. During 2016 almost all provincial-level entities issued debt. These are subject to strict ceilings, but have added a new aspect to the market (see Chapter 6 on local governments).

The municipal bond market was established to reduce the overhang of LGFV debt as uncertainty about the ability and willingness of local governments to pay these debts rose. In 2013 the National Audit Office estimated about RMB18 trillion in outstanding debt either directly owed or contingent liabilities of local governments. The government planned to address the flow problem through the National Budget Law passed in 2015, mentioned above, which established a municipal bond market for local governments to issue debt, up to an annual quota, to finance infrastructure projects. The stock problem was to be addressed through a swap program under which LGFV debt could be swapped off bank balance sheets for new municipal bond issuance. However, the greater risk accruing to LGFV debt, with only a nontransparent potential implicit guarantee, than to municipal bonds meant that the former carried higher interest rates. Encouraging banks to participate in the swap thus required adding sweeteners to the deal. A key sweetener was determining that municipal bonds were eligible collateral for the various People’s Bank of China liquidity operations, such as the Medium-Term and Standing Lending Facilities, and the Pledged Supplementary Lending facility.

Corporate bonds now account for 27 percent of the outstanding volume of the bond market. In general, and as in most countries, bond yields are lower for large companies than bank lending rates, which has encouraged disintermediation. State-owned enterprises have traditionally dominated the market. In recent years, however, a wider variety of firms have issued bonds, encouraged to do so by falling yields, increased secondary market turnover, and greater appetite among investors, particularly following the stock market correction in mid-2015.

There are three classes of corporate bonds with maturities greater than one year, each supervised differently. So-called enterprise bonds, issued by state-owned firms, normally at maturities of three to 10 years, and under the supervision of the National Development and Reform Commission. Corporate bonds, mostly issued by nonfinancial corporates that are not national SOEs, are supervised by the China Securities Regulatory Commission (CSRC). Finally, medium-term notes, issued by nonfinancial corporates, are self-regulated by the National Association of Financial Market Institutional Investors, under the guidance of the People’s Republic of China. The term “corporate bonds” is generally used interchangeably for all three types.

Medium-term notes have certain advantages over the other two. These are effectively corporate loans from banks that are repackaged into bonds and can then be sold. This program was designed to relieve pressure from bank balance sheets, but it is not clear how much of this has happened: much of the stock of new bonds has been bought either by banks (which may reduce large exposure risks and help liquidity management, but retains the credit risk on banks’ balance sheets under different capital coverage and provisioning requirements), by WMPs (which may also be on balance sheet), or by trust beneficiary rights/directional asset management plans (which themselves may be held by WMPs).

Aware of perceptions of guarantees in the financial system, the authorities have gradually allowed more defaults in the bond market, where previously there had been very few. The pace of defaults, particularly among SOEs, began to rise in 2014, though so far, bankruptcies have remained extremely rare. This has made investors aware of potential risks within the market. By mid-2016 this had raised the overall yield curve, but also led to an improvement in risk pricing, with more differentiation by credit quality.

At the shorter end of the interbank market, collateralized repos dominate the interbank bond market, with around RMB2.1 trillion in daily trading in the first quarter of 2016, compared with less than RMB150 billion in uncollateralized repo. Some of this is bank lending but much is also between investors. Many investors repo out their bonds and use the proceeds to lever up their corporate bond portfolios. This maturity mismatch raises sensitivity to interest rate risk. This leverage is primarily taken by institutional investors (insurers, fund managers, securities firms). Individuals and corporates are often the lenders in such cases. The CSRC has tightened the haircut on the corporate bond repo to try to curtail this behavior, along with other measures.

Equity Market

China’s stock markets are now the world’s second largest. By both market capitalization and trading volume, the Shanghai and Shenzhen stock markets lag behind only the United States’ New York Stock Exchange and NASDAQ markets. As with other areas of the financial system, growth has been very fast and volatility is high. Share turnover—which in China is unusually fast, for reasons discussed below—on the Shanghai exchange passed London in 2007; at that time the market capitalization of the Shenzhen exchange was smaller than that of Johannesburg, while it is now larger than Toronto’s.

A wide range of companies are listed, but market capitalization is dominated by state-owned firms. Some of China’s companies—such as China National Petroleum Corporation, China Petrochemical Corporation (Sinopec), and China State Construction Engineering Corporation, and China’s large public banks—are among the largest companies in the world by market capitalization. However, in many of these cases, the government (or other large investors) share of equities is very high, leaving only a small share for free float. This adds to price volatility.

The dominance of retail investors in Chinese stock markets has amplified volatility. While less than 10 percent of Chinese people own stocks (compared with a majority of Americans), in 2014 about 80 percent of tradable shares were held by individual investors. During 2015, as the stock market rose rapidly before peaking midyear, 38 million new accounts (many opened by investors who already held at least one account) were opened. Many, though by no means all, were held by small investors who held shares for only a brief time. The importance of these investors in daily trading volumes increased during the boom, amplifying volatility.

The market is also largely closed to foreign investors. Chinese companies can, and do, list in foreign markets and in Hong Kong SAR. Foreign investors are subject to quotas under China’s Qualified Foreign Institutional Investor program (see Chapter 8 on the capital account). While these quotas have gradually been expanded, they remain binding in many cases. Concerns about these restrictions were one reason behind the decision in June 2016 to postpone inclusion of onshore Chinese equities in the MSCI Emerging Markets Index.

The rapid run-up of equity prices and subsequent correction in mid-2015 also raises questions about the stock market. A perception among investors, to some extent encouraged by the government, that China’s stock market was a one-way bet on the country’s future led to a 112 percent increase in the Shanghai Composite Index between January 2013 and June 2015. As prices began to correct in July, the authorities became very concerned about potential fallout. While direct linkages between the equity market and the real economy were believed to be small, widespread leverage among investors and uncertain risks from complex products were a serious concern. The authorities embarked on a wide-ranging effort to stabilize the market, which included a moratorium on initial public offering, large purchases of equities by the government entity China Securities Finance Corporation, and a lockup of shares held by large investors. The market correction continued for a few months after, but in a more orderly fashion. However, the measures taken at the time have not yet been fully unwound, underscoring the lack of a predictable framework for investors.

High returns in the equity market also attracted a large pool of risk capital, which has subsequently moved on. WMPs are not allowed to invest directly in equities, but can invest through asset management plans. Securities firms also grew rapidly as new investors moved their savings into the stock market. The stock market provided a cheap source of equity finance for firms as well, some of which took advantage of the rise in prices to raise equity finance to pay off debts. Insurance companies’ limits on equity investment were raised in mid-2015, and their funds also moved into equity markets. As prices fell in 2016, many firms liquidated these positions, with one of the main beneficiaries being the bond market.

Regulations on leveraged trading have continued, but remain a concern. The high degree of leveraged trading in China has aggravated price swings in a market already dominated by retail investors and momentum investing. There are a few mechanisms for leveraged investing in Chinese equity markets:9

  • Securities companies. Formal leveraged trading is allowed up to certain margin requirements (generally 50–80 percent of assets net of haircuts).

  • Umbrella trusts. These are tranched trusts in which junior investors (largely high-net-worth investors or institutional investors) borrow from senior investors (WMPs). This is more informal.

  • Over-the-counter fund matching. Fund providers set up and monitor stock accounts that investors use for leveraged trading.

The use of equities as collateral also raises risks. Most bank-loan collateral is in the form of real estate, and interbank trading, when collateralized, normally uses government debt. But equity is used in some cases. Haircuts on collateral are generally large, and supervisors in recent years have raised them. Nevertheless, this remains a potential concern for equity investors and for lenders’ stability.


China’s insurance industry, like the rest of the financial sector, has grown very rapidly in recent years. Total assets of the insurance sector reached RMB12 trillion at the end of 2015, with 32 percent of the Chinese population having access to insurance products in one form or another. About 80 percent of assets are held by life insurers. Around two-thirds of insurance company assets are in deposits or fixed-income securities.

Seeking return, China’s insurers have moved away from fixed-income and liquid investments. In 2010 bonds and cash accounted for nearly 80 percent of insurers’ assets. The China Insurance Regulatory Commission loosened rules on insurance investment in 2104, after which insurance companies began to actively move into the booming equity market. As noted above, in mid-2015, as equity prices began to fall, insurers were allowed to raise their equity exposure from 30 to 40 percent of assets if the difference was invested in blue chip stocks. By the end of 2015 the share of insurance company assets in cash and bonds had fallen to 58 percent. The biggest share of this shift, however, came from alternative assets. The composition of this part of the portfolio is not easy to assess, but investments in infrastructure projects—which generally carry implicit guarantees from local governments—have risen in recent years.

This rising riskiness in asset portfolios is matched by shifts in liabilities. A rising share of products have promised high guaranteed returns. Much of the rapid rise in premium income seen in 2015 came from universal life insurance products. Until February 2015 the return on such products was guaranteed at 2.5 percent. After this ceiling was lifted, average premiums rose to about 5 percent, with smaller companies selling products at even higher returns.10 Many of these products were also of very short duration, raising concerns about asset-liability mismatches and rollover risks.11

The China Insurance Regulatory Commission has responded to these risks. In March 2016 the commission required insurance companies to stop selling products in which most policies would have a duration of less than one year and assigned ceilings based on capital levels to limit the exposure of insurance companies to other short- and medium-term products. China’s new China Risk Oriented Solvency System supervisory framework is also being implemented to bring supervision closer into line with global norms.

Conclusions and Policy Prescriptions

Ensuring China’s financial system remains up to the task of underwriting rapid growth will require continued effort to ensure stability.12 Certain key areas of emphasis emerge:

  • The increasing complexity of financial systems makes supervisory coordination important. Increasing linkages of banks, insurance companies, trusts, securities firms, and other companies mean that systemic risk can jump rapidly from one sector to another.

  • The dispersion of credit risks across the financial systems, from transparent and closely supervised bank loans to more opaque investment portfolios, WMPs, and nonbank financial firms that may still have linkages back to banks, is also a concern. In the case of banks, supervision should ensure that capital coverage and provisioning are sufficient to cover credit and liquidity risks on bank balance sheets, whether these are in the loan book or in complex structured products in the investment book. In other cases, such as WMPs, insurance companies, or asset management plans, the issue is less clear cut, but ensuring a level supervisory playing field that does not allow credit risks to collect in one sector will be crucial.

  • The deepening of interbank and other funding markets also raises potential risks. As the funding models for banks and other firms diversify, ensuring sufficient systemic liquidity becomes more complex. In the case of banks, the move toward Basel III liquidity rules by mid-2017 is an important step. But better understanding of the linkages between banks and other financial firms and ensuring that collateral standards are high are also essential.

Overall, continued development of China’s financial system represents an important opportunity to support growth in what is now the world’s largest economy in purchasing power parity terms. But ensuring this growth is managed in a safe and sustainable way is equally important.


  • China Insurance Regulatory Commission (CIRC). 2015. Annual Report of the Chinese Insurance Market 2015. Beijing.

  • Hong Kong Monetary Authority (HKMA). 2015. Monetary and Financial Stability Report. Hong Kong SAR.

  • Hong Kong Monetary Authority (HKMA). 2016. Monetary and Financial Stability Report. Hong Kong SAR.

  • KPMG Huazhen LLP. 2015. “2015 China Trust Survey.Hong Kong SAR.

  • Morgan Stanley Research. 2016. “Asia Insight: Primer—Demystifying China’s Financial System.Hong Kong SAR.

Thanks to Simon Gray, Alfred Schipke, and Alison Stuart for comments. Zhe Qi, Angelin Oey, and Naixi Wang provided excellent research assistance.


The 12 JSCBs are China Merchants Bank, Industrial Bank, Shanghai Pudong Development Bank, China CITIC Bank, China Minsheng Bank, China Everbright Bank, Hua Xia Bank, China Guangfa Bank, Ping An Bank, Evergrowing Bank, Zhe Shang Bank, and Bo Hai Bank.


Single trust funds match single investors (normally banks or wealth management products, which can themselves have one or many investors) with specific projects, whereas collective trust funds raise funds from multiple investors and allocate them to multiple projects.


Originally, banks used trust companies to invest in restricted, high-risk, and high-return projects, but the landscape has changed continuously in response to various regulatory measures aimed to control shadow banking activities. In recent years, securities companies, fund management companies, and fund management companies’ subsidiaries have expanded in size to channel funds from banks and other investors to high-risk projects, equity, and bonds.


In all cases, however, foreign currency exposure tends to be small. Even among the big four, which are globally systemic banks, assets are overwhelmingly in domestic currency and overseas operations are relatively small.


“Traditional lending” in this case refers to loans held on the loan book of banks’ balance sheets. Lending has instead been shifting toward loans repackaged and held in the investments book of balance sheets, held in off-balance-sheet special purchase vehicles, or moved away from the banking sector entirely.


China’s deposit insurance system was formally established on May 1, 2015.


About 10 percent of trust company assets fit into neither category; these assets are highly varied.


The difference is that while working capital is a short-term loan provided to a borrower, bankers’ acceptances are promises of future payment by a borrower and guaranteed by the bank drawn on a deposit made by the borrower to the bank providing the banker’s acceptance.


Taken from HKMA 2015.


The ceiling was initially raised to 3.5 percent, but in September 2016 it was reduced to 3 percent.


Short- and medium-term insurance products in China are products with a duration of less than five years.


The ongoing work of the joint IMF-World Bank Financial Sector Assessment Program will provide additional detailed recommendations.

Investing in Soft Infrastructure