Mapping shadow banking in China: structure and dynamics

BIS Working Papers

No 701


Mapping shadow banking in China: structure and dynamics

by Torsten Ehlers, Steven Kong and Feng Zhu


Monetary and Economic Department

February 2018


JEL classification: G2

Keywords: shadow banking, wealth management products (WMPs), structured credit intermediation, investment receivables, entrusted loans, trust loans


This publication is available on the BIS website (


© Bank for International Settlements 2017. All rights reserved. Brief excerpts may be reproduced or translated provided the source is stated.

ISSN 1020-0959 (print)

ISSN 1682-7678 (online)


Mapping shadow banking in China: structure and dynamics

Torsten Ehlers, Steven Kong and Feng Zhu



We develop a stylised shadow banking map for China with the aim of providing a coherent picture of its structure and the associated financial system interlinkages. Five key characteristics emerge. One defining feature of the shadow banking system in China is the dominant role of commercial banks, true to the adage that shadow banking in China is the “shadow of the banks”. Moreover, it differs from shadow banking in the United States in that securitisation and market-based instruments play only a limited role. With a series of maps we show that the size and dynamics of shadow banking in China have been changing rapidly. This reveals a marked shift in the relative importance of different shadow banking activities. New and more complex “structured” shadow credit intermediation has emerged and quickly reached a large scale, while the bond market has become highly dependent on funding channelled through wealth management products. As a result, the structure of shadow banking in China is growing more complex.

Keywords: shadow banking, wealth management products (WMPs), investment receivables, entrusted loans, trust loans

JEL classification: G2

Torsten Ehlers, Bank for International Settlements,; Steven Kong, Hong Kong Monetary Authority,; Feng Zhu, Bank for International Settlements, Steven Kong worked on this paper when he was a staff member of the BIS. We are grateful to Hyun Song Shin, Claudio Borio, Stijn Claessens and Benjamin Cohen for valuable suggestions, and Ingo Fender, Madhu Mohanty and Christian Upper for insightful comments. We also thank the participants at the ECB Third workshop of the China Expert Network on “How strong is China’s commitment to reform?”, the KOF-ETH-UZH Seminar in international economic policy, and seminars at the People’s Bank of China, Hong Kong Institute of Monetary Research, and the Bank for International Settlements for their comments. We are indebted to Mario Barrantes for assistance with the shadow banking maps, and Jimmy Shek for help with data and graphs. The views expressed are those of the authors and do not necessarily represent the views of the Bank for International Settlements or the Hong Kong Monetary Authority.


1 Introduction

While an increasing number of studies have analysed various aspects of shadow banking in China, we attempt to provide a more comprehensive and structured view. We develop a stylised shadow banking map (Figure 1), with a particular focus on the central role of commercial banks as well as the financial system interlinkages generated by shadow banking activities. We rely on an activities-based approach and distinguish between three main stages of shadow credit intermediation in China to facilitate our analysis: the ultimate creditor stage, the intermediate stage and the ultimate borrower stage. Each stage involves distinctive financial instruments. Even though the fundamental economic drivers of shadow banking activities are similar, in China, such activities take markedly different forms from those in advanced economies such as the United States. They as well imply very different financial system interlinkages, which need to be taken into account when analysing potential financial stability risks.

We identify five key characteristics of shadow banking in China. The first distinct feature is the dominant role of commercial banks. For this reason, it is often dubbed the "shadow of the banks". Securitisation and market-based instruments play only a limited role. Second, shadow banking serves important economic functions, especially in the form of providing alternative savings instruments and intermediating credit to private firms with less privileged access to formal bank credit. Third, shadow credit intermediation generates tight linkages within the financial system, including the bond market. Fourth, it is far less complex than its US counterpart. Fifth, perceived and actual guarantees are pervasive. Although pervasive guarantees are not at all unique to shadow banking in China, the form and nature of such guarantees are.

Our series of maps for the period 2013-2016 (Figures 2-5) reveal significant changes in the size and dynamics of shadow banking activities, suggesting a rapidly evolving structure of the shadow banking system in China.2 While much of the expansion during China's Great Stimulus in 2009-2010 was due to a push for shadow credit growth to ultimate borrowers, such growth has slowed recently. Yet, shadow savings instruments (eg wealth management products and trust products) at the ultimate creditor stage have kept expanding at a rapid pace. At the same time, new and more complex "structured" shadow credit intermediation has emerged and quickly reached a large scale, with the purpose of reclassifying bank assets to ease regulatory burdens (eg non-performing loan provisions, or loan-to-deposit ratio ceilings). As a result, the shadow banking system in China has grown more complex and thereby somewhat more similar to that in the United States.

2 Related literature

We provide a structured and coherent view of shadow banking in China. Our newly developed shadow banking map delineates the structure of the shadow banking system in China, with a particular emphasis on the interconnections between banks and shadow banking entities. We view this as a useful basis for further analysis, in particular on financial stability and regulatory issues.

The activity-based approach we take allows us to distinguish between three main stages of shadow credit intermediation in China, and to provide detailed information on its size and dynamics at each stage. Most size estimates for shadow banking in China either attempt to measure credit to ultimate borrowers and neglect the other stages of shadow credit intermediation, or add up volumes across different stages which involves a significant degree of double-counting. By documenting the different sizes and dynamics at the three main stages of credit intermediation, we find that the structure of shadow banking in China has evolved rapidly in recent years.

Our work builds on a growing literature on shadow banking, sparked by the Great Financial Crisis (GFC) of 2007-09 and the role of non-bank financial intermediaries in the US financial system (Acharya et al (2010), Brunnermeier (2009)). The focus in the literature on US, and to some extent European, shadow banking entities (eg money market and investment funds, brokers and dealers, special purpose vehicles) and activities (eg securitisation, wholesale financing, securities lending) has in effect become the basis for monitoring shadow banking activities globally (Adrian and Ashcraft (2016), FSB (2016)). Nevertheless, in many emerging economies including China, market-based financial instruments or securitisation have not been as relevant a factor as in advanced economies. The financial institutions involved are different, reflecting the economies' different histories and stages of financial development. The aim of our paper is to describe the specific characteristics and elements of shadow banking in China and thereby provide a basis for monitoring and analysing it.

While the current structure and form of shadow banking in China are unique, the fundamental economic raison d'être and drivers of shadow banking are similar. A key driver of shadow banking is the existence of some form of implicit guarantee or public backstop (Claessens and Ratnovski (2014), Kane (2014)). It enables the creation and widespread use of shadow banking products, especially those assets perceived as "safe" by investors. Dang et al (2014) note the asymmetric perception of risks among shadow banking entities, banks, and investors in China, given widely perceived credit guarantees from banks by investors.

On the supply side, the evasion of banking regulations (IMF (2014), Adrian et alet al (1995) for the United States, Wang et al (2016) for China). Borst (2013) draws parallels between shadow banking in the United States during the 1960s and 1970s and in China more recently, arguing that both were driven by the rise of shadow deposits to circumvent bank deposit rate ceilings. More generally, the demand for shadow banking is driven by the fundamental economic functions it performs (Claessens et al 2012), which in turn can be welfare improving (Gennaioli et al (2013)). One typical function it fulfils is to satisfy the demand for "safe" assets by investors and savers (Gorton et al (2012)), which is also a key characteristic of shadow banking in China (see section 3.3).

Despite the similar fundamental drivers of shadow banking activities in different jurisdictions, the distinct forms they take are important for identifying financial system interlinkages and assessing potential financial stability risks. Detailed structural analyses of shadow banking have been conducted for other economies, most notably the United States (Pozsar et al (2010)). Several papers examine specifically the structure of shadow banking systems, such as Pozsar and Singh (2011) for the United States, Bakk-Simon et al (2012) for the euro area, Acharya et al (2013) for India, and Harutyunyan et al (2015) for 26 mostly developed countries.

Research on shadow banking in China has grown in recent years, but often has a narrow focus on entrusted loans and wealth management products (WMPs). Allen et al (2015) find that entrusted loans are essentially a market reaction to credit shortages, as they increase when bank credit becomes tight. Allen et al (2016) and Chen et al (2016) conduct a transaction-level analysis of China's entrusted loans. Chen et al (2016) suggest that shadow banking is strongly linked to commercial banks' balance-sheet risks, as banks are prone to engaging in channelling risky entrusted loans. Acharya et al (2016) point out that the swift rise of shadow banking was triggered by China's stimulus plan announced in November 2008, with small and medium-sized banks significantly increasing shadow banking activities by issuing WMPs. Cai et al (2015) state that banks use WMPs for regulatory arbitrage or window­dressing, and Hachem and Song (2016) argue that tightening the loan-to-deposit ratio regulation led to a rise in bank-issued WMPs. Chen et al (2017) find that those provinces with abnormally higher bank loan growth in 2009 showed a higher degree of shadow banking activities, including entrusted loans and WMPs.

Several papers have argued that shadow credit serves Chinese private enterprises as an important source of funding, given the strong preference of the large state- owned banks for lending to large state-owned enterprises (Song et al (2011)). Ayyagari et al (2010) use firm-level data to examine firm financing patterns, including those of SMEs. They find that 80% of firm financing goes through informal channels, and firms with access to formal bank loans tend to grow faster. Hale and Long (2010), Lu et al (2015) and Tsai (2016) find that small and medium-size enterprises (SMEs) particularly rely on informal financing, as their access to formal credit is limited.

A few papers take a quantitative approach to analyse financial stability risks of shadow banking. Li et al (2014) run a stress test on the Chinese financial system and conclude that there is some risk of bankruptcies and potentially a risk of liquidity shortages. Hsu et al (2014) conclude, based on their network analysis, that trust companies present a systemic risk and banks absorb most of this risk.

3 The structure of the shadow banking system in China

To illustrate the structure of the shadow banking system in China, we map the flow of funds from ultimate creditors to ultimate borrowers, analogous to the seminal wall chart of US shadow banking by the Federal Reserve Bank of New York (Pozsar et al (2010)). In Section 5 we provide annual snapshots of the Chinese shadow banking system for 2013-2016, giving rise to a "movie" of how the system has evolved. Our stylised shadow banking maps are constructed based on an in-depth analysis of publically available data from different sources on the main shadow banking entities and shadow credit instruments.3

3.1 Definition

Most existing definitions of shadow banking are tailored to advanced economies. For this reason, we first develop a definition that can better capture the structural characteristics of shadow banking in China.

We adopt an activity-based definition, namely "all financial instruments that fulfil functions of credit intermediation typically performed by banks (such as liquidity, maturity, and credit risk transformation), but reduce the burden of or bypass banking regulation."4 Our definition is close to the one given by the People's Bank of China (PBOC).5 It specifically includes credit intermediation activities performed by banks themselves that lower or circumvent regulatory requirements.6

Definitions of shadow banking can broadly be divided into three categories: activity-based, entity-based, and a mixture of the two (IMF (2014)). As existing empirical studies typically build on entity-based definitions, our activity-based analysis is not directly comparable to those studies. The most widely used definition of shadow banking is from the Financial Stability Board's (FSB) 2012 shadow banking monitoring report: "credit intermediation involving entities and activities (fully or partially) outside the regular banking system" (FSB (2012)). The de-facto entity-based approach facilitates the collection and comparison of data across different jurisdictions for which the FSB monitors shadow banking activity, using data from national accounts as the starting point.7 Since our definition is not based on balance sheet positions but rather on a set of financial instruments, including those used or held by banks, our perspective on shadow banking in China is different from that taken in the FSB's analyses.

Our definition is consistent with other prominent and more general activity- based definitions that define shadow banking on the basis of fundamental economic characteristics (Kane (2014) ,Claessens and Ratnovski (2014)).8 Further, the financial instruments we include in our definition are generally consistent with those used in other studies on shadow banking in China, such as Elliott, Kroeber and Qiao (2015), JP Morgan (2013) and Moody's (2017a). We examine the relevant instruments in Section 4. Table A1 in Appendix I contains the list of instruments and key characteristics, and Tables 1-3 in Section 5 provide size estimates corresponding to each stage of shadow credit intermediation.

Our analysis brings to light some additional aspects which have received less attention in the literature thus far. For instance, a large share of bond issuance is funded by shadow savings instruments, creating strong linkages between the bond market and shadow banking activity. We also document new forms of "structured" shadow credit intermediation.

We exclude several instruments often considered to be part of shadow banking in China. In particular, we exclude bonds and investments in funds.9 In these cases of market-based intermediation, the associated risks are (to a large extent) directly passed on to investors. Therefore it is not bank-like credit intermediation, where liquid short-term and safe deposits are transformed into less liquid longer-term and risky loans. We further exclude banker's acceptances, both discounted and undiscounted, from shadow banking, as we view them merely as facilitating tools, but not as a form of shadow credit intermediation in itself. In the case of money market funds, comprehensive and sufficiently granular data was lacking. Hence, their potentially relevant activities are not included in our analysis. Nevertheless, the current structure and past dynamics of shadow banking are unlikely to be affected in a material way, as total assets of money market funds are still relatively small compared to those related to other shadow banking activities.10

Our activity-based definition comes with several caveats. Some shadow banking activities may be hidden and evade our coverage. Further, the activity-based view is relatively static and requires adaption over time. As China's financial system undergoes fast-paced and sometimes fundamental changes, new instruments and shadow credit intermediation channels can emerge and develop rapidly.

3.2 A stylised shadow banking map for China

Our shadow banking map in Figure 1 illustrates the basic structure of the shadow banking system in China. The map is stylised, in that it presents a greatly simplified structure, focussing only on the main shadow banking instruments and entities in China that currently exist, as well as the most relevant interlinkages among commercial banks and the key non-bank financial intermediaries. The orange, rose, red, and purple arrows depict the main shadow banking instruments and the resulting claims. Formal credit intermediation by banks is represented by the blue arrows. The green arrows stand for shadow credit intermediation through the bond market.

We distinguish three main stages of shadow credit intermediation to facilitate the analysis of the structure.12 The ultimate creditor stage (Figure 1, right-hand side) is the source of funding, consisting mainly of private and corporate depositors. The most relevant shadow savings instruments are represented by the orange arrows. At the intermediate stage (centre), the received funds are then intermediated and transformed into different shadow banking assets. The various forms of shadow credit intermediation generate tight interlinkages (rose and red arrows) among banks and other shadow banking entities (trust companies and securities companies including banks' wealth management arms). The ultimate borrower stage (left-hand side) is the final destination of shadow credit, represented by the purple arrows.13 Each stage involves different types of instruments and entities, and performs different economic functions (see Section 4).

Our shadow banking map points to a number of notable differences between shadow banking in China and that in the United States (see Pozsar et al (2010)). First, banks are the dominant players in the shadow banking system.14 Banks issue key instruments (ie wealth management products), they then channel the proceeds to non-bank entities such as trust companies. Banks are also the driving force behind the new and more complex forms of "structured" shadow credit intermediation. Second, shadow credit intermediation in China hardly involves securitisation or wholesale funding, which are the ultimate sources and drivers of US shadow banking. In fact, shadow banking in China has been much more akin to traditional banking: it collects "deposits" or cash from retail and corporate investors, and then transforms their savings into credit of different forms to provide funding to firms. Given the banks' central role, shadow banking in China is often dubbed the "shadow of the banks", as opposed to a capital market-based form of credit intermediation. Furthermore, it is important to point out that the role of foreign investors or foreign financial entities is negligible. Shadow banking activity in China is driven by domestic financial institutions, savers and investors. It also involves fewer types of entities.


3.3 The five key characteristics

The stylised shadow banking map illustrates five key characteristics of shadow credit intermediation in China.Commercial banks are the key players at the centre of shadow credit intermediation (Figure 1, rose and red arrows). They are the main linkage between the suppliers and borrowers of funds in both the formal and the shadow banking system. Banks issue key shadow banking instruments such as wealth management products (WMP), they channel investors' funds and provide liquidity to other shadow banking entities (eg trust companies), and they are holders of shadow banking instruments such as trust beneficiary rights (TBR) or interbank WMPs.

Besides their direct role, banks facilitate shadow credit in various ways. For instance, since direct loans between non-financial firms are legally not permissible, banks act as the trustee and middleman of the so-called entrusted loans between firms. The trustee bank collects the principal and interest and charges a handling fee. It either does not take credit risk in the process, or absorbs part of the credit risk through so-called entrusted rights (Chen et al (2016)). By distributing and intermediating a wide range of shadow banking products themselves, as well as on behalf of other entities, banks are the central player within China's shadow banking system. In this role, banks are perceived as providers of implicit guarantees to their customers in case of defaults, even though they have no such legal obligation.

More recently, banks have resorted to a combination of existing shadow credit instruments to reduce regulatory burdens. Through this "structured" form of shadow credit intermediation, bank assets are reclassified into investment receivables. Banks can thereby lower non-performing loan (NPL) provisions and alleviate loan-to- deposit (LTD) ratio constraints (see Section 4.2.3).

 Arguably, there are two main drivers. First, the low deposit rate ceiling that was only abolished in October 2015 (Wang et al (2016), Lardy (2008)). Until then, shadow instruments allowed banks to circumvent the ceiling and offer significantly higher yields. Second, competition for funds and savers' growing demand for higher- yield savings products have driven the rise of shadow savings instruments, especially since these are often perceived as safe (Dang et al (2014)). Direct access to the bond market is rather limited. Bond mutual funds and other unit trusts are still in their early stage of development. Even after the full liberalisation of bank deposit rates in October 2015, shadow savings instruments have continued to pay a noticeable premium over bank deposits.

Shadow banking provides credit to private firms which otherwise would be unavailable or too difficult to obtain. As these firms are typically more productive than their state-owned counterparts (Hsieh and Klenow (2009), Dollar and Wei (2007)), shadow credit is likely to lead to direct economic gains. Traditionally, most private firms as well as smaller state-owned enterprises (SOEs) have difficulties in accessing the formal credit market, as large state-owned banks prefer to lend to large SOEs (Hale and Long (2010), Lu et al (2015), Tsai (2016)). Banks' preferences for large SOEs reflect historical relationships and high creditworthiness due to implicit or explicit government backing, reducing the credit supply to potentially more productive private firms. Shadow credit intermediation has helped to fill this gap.

Shadow credit intermediation not only implies tight interlinkages between commercial banks and shadow banking entities (Figure 1, rose and red arrows), but also generates close ties with China's bond market (Figure 1, green arrows). A large share of the proceeds from WMPs have been invested in the bond market. This is an intended consequence of regulation on WMPs, stipulating that at least 75% of the underlying assets of WMPs must consist of so-called standardised debt instruments, including bonds, money market instruments, and bank deposits. WMPs have effectively provided a channel for retail investors to invest in bonds, as direct access to the interbank bond market is restricted to financial institutions.

New forms of "structured" credit intermediation have flourished in recent years, in particular those related to banks' investment receivables. The reclassification of bank assets into investment receivables is generating even tighter and more complex and opaque linkages between the formal banking sector and shadow banking entities. In a first step, a bank's exposures are usually transferred to trust companies or its' asset or wealth management arms (Figure 1, rose arrows). In return, the bank receives full participation rights in the profits and losses of the underlying loans or debt securities, through TBRs and directed asset management products (DAMPs). The credit exposures are sometimes transferred among banks through interbank WMPs (Figure 1, rose arrows within joint-stock and city banks), for which the TBRs or DAMPs serve as the underlying securities.

Shadow banking in China is less complex than in the United States, as it involves fewer entity types and fewer steps of credit intermediation. Mostly, shadow credit intermediation in China is a one-step or two-step intermediation process, as it is effectively based on "plain vanilla" loans or instruments that entail a one-to-one link to the revenues from the underlying debt instruments. In contrast, a typical shadow credit intermediation process in the United States involves seven steps ("vertical slicing") and a large number of financial entities (Adrian and Ashcraft (2016)).

Nevertheless, the tight linkages between shadow savings instruments and bond market, as well as the new forms of structured shadow credit intermediation, signal that shadow banking in China is growing more complex.While shadow credit intermediation operates without access to central bank liquidity and does not benefit from deposit insurance protection, shadow banking instruments in China are backed by implicit and explicit guarantees (Figure 1, grey arrows). In both China and the United States, the expectation of an eventual government rescue of systemically important financial institutions that get into trouble provides implicit guarantees to investors and creditors. In the United States, high credit ratings for market-based financial instruments gave a false sense of safety ahead of the GFC. In addition, banks and other financial institutions explicitly offered conditional liquidity provision arrangements to special purpose entities issuing such instruments.

Nevertheless, the precedents set by past bail-outs and the perceived priority placed by the authorities on maintaining financial market and social stability contribute to perceptions of implicit bank guarantees. If a state-owned bank is involved, customers may view the products as being ultimately backed by a government guarantee. In a competitive environment, banks have an incentive to encourage such assumptions. In addition, credit guarantee companies provide explicit guarantees for a wide range of shadow banking activities in China. While these companies help smaller and newly founded enterprises access the credit market, they also create opaque financial interdependencies. Credit guarantee companies are often lightly regulated by provincial and local authorities. In some instances, defaults
of trust products or WMPs have entailed secondary defaults of credit guarantee companies (Caixin (2012)).

4 The main shadow banking instruments

Each of the three main stages of shadow credit intermediation in China features a distinct set of instruments. At the ultimate creditor stage, shadow banking is directly and primarily funded by depositors, who purchase bank-issued WMPs or trust products. Consequently, shadow credit intermediation does not require collateralisation or securitisation at this stage, in contrast to US shadow banking system.15 For this reason, shadow banking in China is better characterised as a bank­like credit intermediation process, with deposits being collected through higher-yield WMPs or trust products. The lack of collateralisation or securitisation on the funding side is one key reason for why the structure of the shadow banking system in China remains so different from that in the United States.

4.1 The ultimate creditor stage​​​​​​​

Apart from entrusted loans and P2P loans (see Section 4.3), shadow bank funding in general relies on two main instruments: wealth management products (WMPs) and trust products.

4.1.1 Wealth Management Products​​​​​​​

Bank-issued WMPs serve as alternative savings instruments, which promise higher returns than traditional bank deposits, but are still regarded as safe. Banks are the dominant issuers of WMPs and offer two main types of WMPs.16 First, the principal- or return-guaranteed WMPs (PRG-WMPs) that entail full bank guarantees either on the principal or on the return. They are recorded on banks' balance sheets along with the underlying investment which they finance. PRG-WMPs are akin to negotiable certificates of deposit (NCD) and subject to normal banking regulations. We therefore do not consider them to be part of shadow banking.17 Second, banks offer non­principal guaranteed WMPs with no explicit bank guarantees, which are not recorded on banks' balance sheets. The underlying investment of the non-principal guaranteed WMPs is usually held by a channelling company, eg banks' investment or wealth management arms, but the issuing bank normally retains full control over the investment. Effectively, banks act as asset managers, charging fees to investors, without being subject to regulatory restrictions, except for those governing the admissible design of WMPs. By the end of 2016, the share of non-guaranteed WMPs in total outstanding WMPs reached almost 80% (Graph 1, left-hand panel).


Most WMPs are closed-ended, meaning that they have a definite life cycle and investors have to subscribe during specific periods. They accounted for around 57% of outstanding balances in mid-2016. In the case of open-ended WMPs, investors can subscribe and redeem at any time (Credit Suisse (2014)). Open-ended WMPs have risen to 43% in mid-2016 from 35% at the end of 2014. Importantly, only a small share of outstanding WMPs (6% in mid-2016) are open-ended instruments where regular information on the net asset value (NAV) of the underlying assets is available. In all other cases, investors typically do not have timely information on the value of the underlying assets (non-NAV WMPs). In this sense, WMPs are a relatively opaque savings instruments compared to bank deposits or debt securities.

The "standardised assets", which are essentially cash-like assets (eg bank deposits and money market instruments) or tradable securities (eg tradable bonds), now make up the lion's share of the underlying WMP investments (Graph 2, right- hand panel). The nature of the underlying investments of bank-issued WMPs suggests a potentially large maturity mismatch. At end-2015, more than 27% of outstanding close-ended WMPs had maturities of less than three months, 59% less than six months, and over 90% less than 12 months. The underlying debt and equity investments as well as non-standardised assets, however, typically have significantly longer maturities.

4.1.2 Trust products

Trust products are issued by trust companies which perform credit intermediation functions like commercial banks, but are not subject to the same regulations. Traditionally, they have played an important role in China as lenders to the less- favoured and riskier borrowers with limited access to bank credit - in particular new and smaller private firms. This was helped by the relatively looser regulations in comparison to banking regulations. But since 2007, a series of regulatory measures have significantly altered the framework under which trust companies could operate (Zhu and Conrad (2014)). Regulators aimed to transform trust companies into professional third-party wealth managers, and the measures resulted in an increase in the share of tradable securities in trust investments (Graph 3, left-hand panel). Nevertheless, trust loans remain the single most important type of underlying assets held by trust companies.

Trust companies issue single- or collective-investor trust products, as well as property trust products. Each type is directly linked to the type of shadow credit intermediation that is performed. Single-investor trust products (SITP) are related to the channelling business by banks and other large-scale investors, eg securities firms, investment funds, pension funds and insurance companies. They are specifically

tailored to the needs of a single investor and typically have either one or a small number of underlying investment assets. SITPs are also used when banks channel proceeds from WMPs to trust companies. More generally, they facilitate investment in specific assets determined by a large investor, who cannot or does not want to hold them directly on its balance sheet. The share of SITPs in all trust products fell to 50% at end-2016, from close to 70% at end-2012 (Graph 3, right-hand panel).

Collective-investor trust products (CITPs) bundle funds from various investors and have a larger number of underlying assets, though typically the underlying investments are less diversified than in the case of WMPs. Financial institutions are large investors in CITPs. In addition, CITPs are sold to wealthy individuals and retail investors. Their share rose steadily from around 25% in 2012 to 36% at end-2016.

Non-pecuniary property trusts manage non-monetary assets such as physical or other illiquid assets, mostly on behalf of a single client. In many cases, property trusts are used to achieve bankruptcy isolation, rather than for investment management purposes. They have grown over time, but their share remains small.

4.2 The intermediate stage and financial system interlinkages

The intermediation of shadow credit in China is where banks, as the dominant players, become most closely linked with other financial institutions. Funds are intermediated from banks to shadow banking entities ("channelling business"), and bank balance sheet positions are reclassified through a "restructuring" of different shadow banking instruments ("structured" shadow credit intermediation). An important recent phenomenon is the growth of linkages between shadow savings instruments and capital markets - in particular the bond market.

4.2.1 Channelling business​​​​​​​

The channelling business is driven by banks directing a substantial portion of funds raised through bank-issued WMPs towards other shadow banking entities and instruments. The main examples are the channelling of funds to trust companies (the so-called "bank-trust cooperation") or to securities brokerages (the so-called "bank­security brokerage cooperation"). This not only creates strong ties between banks and other shadow intermediaries, but also significantly boosts the volume of shadow credit, as banks effectively redirect the raised funds into shadow banking instruments.

At the end of 2016, bank-trust cooperation provided 23.5% of funding for trust products. Assuming that single-investor trusts are the main recipient of funds channelled by banks, the dependence on funding channelled through bank-issued WMPs is even higher. (Graph 4, left-hand panel). While the amount of funds channelled from bank-issued WMPs into trust companies and trust products has declined steadily starting in 2010, the dependence of trust companies on bank-trust cooperation has actually risen from 2013 onwards.

4.2.2 Shadow funding of capital markets​​​​​​​

The large share of WMP investments into the bond market is associated with recent regulatory changes, which have strongly encouraged investments in tradable debt assets. Another factor has been the rapid development of China's bond market, opening the door to a wider range and much larger volume of investable debt securities being offered on the primary and secondary markets. The regulations are designed to increase the resilience of the financial system, by steering more funds from ultimate creditors into the formal financial sector and by diversifying firms' funding sources. Yet, as bank-issued WMPs have become a major source of funding for bonds, the bond market remains dependent on the ability of banks to continue to issue and roll-over large volumes of WMPs. The growing dependence of bond market funding on bank-issued WMPs therefore significantly reduces the potential diversification benefits of increased bond financing.

4.2.3 Structured shadow credit intermediation​​​​​​​

Shadow banking in China is growing more complex, although its complexity remains relatively limited compared to that of the US shadow banking system. An important driver is the recent rise of structured shadow credit intermediation, in particular the reclassification of banks' on-balance-sheet assets into so-called investment receivables. Not all investment receivables are necessarily related to shadow banking, as they may also correspond to regular investments in bonds or funds. But the sudden and rapid growth of investment receivables (Graph 5, left-hand panel) may signal a possible change in the nature of shadow banking in China. This is illustrative of how pockets of risks can quickly emerge as a by-product of the rapid change in China's financial landscape.

The use of investment receivables allows banks to reclassify on-balance-sheet loans and debt securities in order to avoid or reduce regulatory burdens - especially the required provisions for NPLs or the LTD ratio.20 As NPLs have risen steadily in recent years, necessary provisions for new and old loans have become an increasingly binding regulatory constraint. In addition, as the growth of bank deposits slowed, a reclassification of existing loans would provide space for banks to concede additional loans under stringent LTD ratio requirements. But, the reclassification also creates additional linkages between banks and trust companies, as well as securities firms. Furthermore, reclassifying assets into investment receivables allows banks to obscure the eventual credit risks associated with the underlying loans or debt securities.

In its simplest form, an investment receivable results when a bank transfers an on-balance-sheet loan or debt security to a trust company or an asset management firm (Figure 1, rose arrows).21 In return, banks receive TBRs or DAMPs, and thereby become entitled to the full participation rights in the gains and losses of the transferred assets. Since direct loan transfers to trust companies are prohibited, the trust company instead issues a trust product with the loan as the underlying asset. The trust company can then issue a TBR on the trust product back to the bank. But the trust product can also be used by a securities firm as an underlying asset for a DAMP. In either case, the resulting asset shows up on the bank's balance sheet as a TBR or DAMP, which are broadly classified as investment receivables.

In a more complicated form of bank-to-bank TBRs transfer (Lu et al (2015)), Bank A is unable to extend a loan given the regulatory constraints it faces, and therefore asks Bank B to lend on its behalf. Bank B then records the investment receivable on its balance sheet, and it issues an interbank WMP to Bank A (Figure 1, rose arrows within joint-stock and city commercial banks; see also Table A1). The interbank WMP allows Bank A to fully participate in the profits and losses of the underlying assets held by Bank B. Since the interbank WMP does not count as a loan, it therefore has little impact on Bank A's NPL provisions or its loan-to-deposit ratio. Thus far, interbank WMPs have been largely issued and held by joint-stock and city commercial banks (Moody's (2017b)).

While investment receivables have grown rapidly in recent years, growth has levelled off as regulators have weighed in. Investment receivables held by joint-stock banks and smaller city commercial banks, however, still grow at a considerable rate (Graph 5, right-hand panel). To better proxy the structured shadow credit intermediation associated with the growth in investment receivables, we identify those that are associated with TBRs and asset management products from the financial reports of the major banks. This suggests that the reclassification of bank loans and debt securities holdings into investment receivable is much more prominent among joint-stock banks and city commercial banks. These banks have typically been more exposed to loans to private sector borrowers and smaller SOEs, which could have left them more exposed to rising corporate NPLs. State-owned banks, on the other hand, lend primarily to large SOEs which often receive explicit government guarantees or are perceived to have implicit guarantees from different levels of government. Therefore, their loan books and debt securities portfolios have been less affected by NPLs.


4.3 The ultimate borrower stage

Trust loans and entrusted loans account for the bulk of shadow credit to ultimate borrowers (Graph 6, left-hand panel).22 Both are components of total social financing (TSF) - a measure of aggregate financing to the real economy published by the People's Bank of China. 23 In addition, new internet-based forms of credit intermediation, such as P2P lending, are fast-growing but still relatively small components of China's shadow credit.24 Another relevant component is informal credit provided by small private lenders, which is hard to quantify and therefore is not part of our shadow banking map.


Shadow credit to ultimate borrowers provides funding for firms and individuals with limited access to formal bank credit, which is dominated by state-owned banks that favour lending to large SOEs. Trust loans, akin to bank loans but extended by trust companies, have traditionally served the important function of supplying credit to new and smaller firms, as well as higher-risk projects. Entrusted loans are loans between two companies with a bank serving as a middle-man. Their growth is both driven by mostly larger corporates looking to earn a better return on their extensive savings and the intermediation of credit from large SOEs to their subsidiaries or associates that have a more restricted level of access to formal financing.

In recent years new forms of internet-based shadow credit intermediation have emerged. Peer-to-peer (P2P) lending has grown fast, as borrowing costs have fallen (Graph 6, right-hand panel). Yet, the overall size of P2P credit is still rather small compared to bank credit. Some P2P platforms have specialised in consumer credit; others such as Ant Financial of Alibaba and Tencent have set up P2P platforms dedicated to loans to small and medium enterprises. Thus far, these platforms have been operating without full banking licenses (CGFS and FSB (2017)). They are, however, subject to binding borrowing limits for individuals (a maximum of CNY 200,000) and institutions (a maximum of CNY 5 million). Regulations have been strengthened recently, outlining the basic requirements and core obligations of online lenders, including the requirement to appoint a commercial bank as fund custodian (Caixin (2017)).

In contrast to other studies, we do not include bank acceptances (BAs) in shadow banking. BAs are essentially letters of credit issued by banks and therefore can be seen as a form of formal bank credit. Normally, a bank issues a BA promising the future payment by Firm A to Firm B, for which it provides a full guarantee. At the maturity date, the bank pays the promised amount to Firm B, even if Firm A is in default. BAs can play a facilitating role in the shadow banking system, as they provide a bank-guaranteed and hence trusted means of payment that can be used for a wide range of transactions. In this way, BAs can serve as a link between the formal and the shadow banking sectors. However, we do not regard BAs as shadow credit intermediation in itself. As regulations on the usage of BAs have tightened, the amount of outstanding BAs has dropped substantially in recent years (Graph 6, left- hand panel).

5 Size and dynamics

Based on our stylised shadow banking map (Figure 1), Figures 2-5 display the size and dynamics of shadow banking activities from 2013 to 2016. The size of the arrows is determined by the outstanding stocks of the corresponding shadow credit instruments at the end of each year. Quickly paging through Figures 2-5 reveals that growth in shadow credit intermediation at the ultimate creditor stage and the intermediate stage has outpaced growth in the ultimate borrower stage by a sizeable margin. This points to a rapidly changing structure of shadow banking in China.

We construct size estimates for each main stage of shadow credit intermediation in China in Tables 1-3. As each stage is associated with a distinct economic function, the different size estimates give a clearer picture of the role and importance of shadow banking in financial intermediation in China. The estimates for the three different stages illustrate and circumvent the problem of double-counting.

Measures of the size and dynamics of shadow banking in China depend on the specific perspective that is taken. For instance, taking the ultimate creditor view yields a much larger size estimate than taking the ultimate borrower perspective, as a large portion of the proceeds from WMPs is channelled into the bond market. Taking a holistic view by summing up the volumes of shadow credit intermediation over the entire credit intermediation process naturally results in an even higher estimate. But doing so introduces a high degree of double-counting, if the measure of interest is shadow credit to ultimate borrowers. A sizeable share of the underlying investment of bank-issued WMPs, for instance, appears on trust companies' balance sheets. Adding up the two may capture the total volume of activity by shadow banking intermediaries, but it substantially overstates the amount of shadow credit that eventually flows to ultimate borrowers.

Recent estimates of the relative size of shadow credit in China usually focus on the credit provided to ultimate borrowers and broadly range between 15% and 70% of GDP. The wide range is due both to differences in which items are considered as shadow credit and to double-counting. From the ultimate borrower perspective the size of shadow credit in China is not particularly high, as even the highest estimates are still smaller, or similar, to those for advanced economies. The FSB's 2015 estimates for the total size of shadow banking were 147% of GDP for the United Kingdom, 82% for the United States, and 60% for Japan (FSB (2016)).

The ultimate borrower perspective, however, leaves out important elements of shadow credit intermediation in China. These include WMPs and trust products at the ultimate creditor stage, or the rapidly rising structured shadow credit intermediation driven by joint-stock and city commercial banks. Those elements are highly relevant for understanding the structure and dynamics of shadow banking in China and particularly important for regulators and analysts conducting research on financial stability issues.

5.2 Size and dynamics at the ultimate creditor stage​​​​​​​

At the ultimate creditor stage, we estimate the volume of outstanding shadow savings instruments at end-2016 at around 71% of GDP, or 46% of total bank deposits (Table 1). Shadow savings instruments are therefore a very important alternative to bank deposits; and their importance has been growing rapidly in recent years. The outstanding amount of non-guaranteed WMPs rose by 33% in 2016 and at an annual rate of 63% between 2013 and 2015. Outstanding trust products were of similar size as WMPs at end-2016 and have also expanded at a fast pace.

WMPs in the more complex form of structured shadow credit intermediation, which typically does not lead to direct credit to ultimate borrowers.

5.3 Size and dynamics at the intermediate stage​​​​​​​

Shadow banking instruments at the intermediate stage have grown at an even higher pace in recent years (Table 2).26 Growing bank-trust cooperation, structured credit intermediation, and WMP investments in the bond market imply heightened and increasing financial system interlinkages. The rapid rise in structured credit intermediation between end-2014 and end-2016 signals the growing importance of regulatory avoidance as a driver of shadow banking activities. Even though regulators have managed to contain the overall growth in investment receivables associate with structured shadow credit intermediation in 2016, joint-stock and city commercial banks appear to continue to engage in such intermediation on a large scale.


5.4 Size and dynamics at the ultimate borrower stage

We distinguish between a narrow and broader measure of shadow credit to ultimate borrowers (Table 3).27 At around 32% of GDP, the narrow measure is in the lower range of the comparable size estimates from other studies, even though we have included in it the estimates of informal credit from a survey by the PBOC from 2011. Notwithstanding the missing times series on informal lending, our narrow measure of direct shadow credit to ultimate borrowers shows much lower growth in recent years in comparison with the other two stages of shadow credit intermediation.

5.5 The changing face of shadow banking in China​​​​​​​

The distinct dynamics at the three stages of shadow credit intermediation, from ultimate creditors via the intermediate stage to ultimate borrowers, point to major changes in China's shadow banking system.

Following the GFC, and especially during the period of strong stimulus policies in China implemented during 2009-2010, shadow banking has played an important role in channelling funds to the underserved and potentially riskier sectors of the real economy. But direct shadow credit to ultimate borrowers has slowed considerably in recent years, whereas the volume of shadow funding as well as structured shadow credit intermediation has grown at a fast pace. The shift in the relative importance of the different stages of shadow credit intermediation is clearly visible in Figures 2-5.

The structure of the shadow banking system in China, however, is changing. The growth of shadow savings instruments, in particular bank-issued WMPs and trust products, suggests an increasing demand for shadow banking in China to provide alternative higher-yielding savings instruments. At the same time, the increased use of structured products combining existing shadow banking instruments, which are based on the collateralisation of bank assets, points to an increasing degree of sophistication and incipient changes in the shadow banking structure. If NPL ratios rise and NPL provisions become a more binding regulatory constraint, structures designed to convert loans into other types of assets that obscure the underlying credit risks and result in lower provision requirements are unlikely to fade away.

Consequently, shadow banking in China is becoming more complex. It increasingly relies on more sophisticated structures of existing shadow banking instruments, which are based on the collateralisation of bank assets. In this respect, shadow banking in China is becoming somewhat more similar to that in the United States. Nevertheless, the level of complexity still remains substantially lower. Furthermore, the growing de facto dependence of the bond market on funding from bank-issued WMPs has generated tight financial interlinkages, and presents an additional transmission channel for financial shocks between depositors, banks and the bond market.

6 Conclusions

Understanding the structure of China's shadow banking system is crucial for analysing China's financial system. We provide a stylised map of shadow banking in China, which highlights the main forms of shadow banking and the resulting financial system interlinkages. Shadow banking in China takes a markedly different form compared to that in the United States. A key characteristic is that commercial banks are the dominant players in China's shadow banking system. The system is effectively a "shadow of the banks", while securitisation and market-based instruments still play only a limited role.

We show that the structure of shadow banking in China is evolving. Its size and dynamics have changed rapidly in recent years. The main area of growth has shifted from shadow credit provision to private firms with less privileged access to formal bank credit, towards offering alternative savings instruments (eg WMPs and trust products). Similarly, at the intermediate stage, new and more complex "structured" shadow credit intermediation has emerged and quickly has reached a large scale. This is driven by banks trying to alleviate regulatory burdens (eg NPL provisions or LTD ratio ceilings) through a reclassification of existing bank assets into investment receivables. Tight and growing financial sector linkages further raise the potential for the transmission of financial shocks among savers, banks and the bond market. In addition, new forms of internet-based credit intermediation, such as P2P lending, have been expanding at an extraordinary pace. As a result, shadow banking in China is growing more complex and thereby becoming slightly more similar to the US shadow banking.


Acharya, V. V., T. Cooley, M. Richardson, and I. Walter (2010): "Manufacturing tail risk: A perspective on the financial crisis of 2007-2009", Foundations and Trends in Finance, Vol. 4, No. 4, pp. 247-325.

Acharya, V. V., Khandwala, H., and Oncu, T. S. (2013): "The growth of a shadow banking system in emerging markets: Evidence from India", Journal of International Money and Finance, No. 39, pp. 207-230.

Acharya, V., Qian, J., and Yang, Z. (2016): "In the shadow of banks: Wealth management products and issuing banks' risk in China", Mimeo.

Adrian, T., and Ashcraft, A. B. (2016): "Shadow banking: a review of the literature", in "Banking Crises", Palgrave Macmillan UK, pp. 282-315.

Allen, F., Qian, Y., Tu, G., and Yu, F. (2016): "Entrusted Loans: A Close Look at China's Shadow Banking System", Mimeo.

Ashcraft, A., T. Adrian, N. Cetorelli (2014): "Shadow Bank Monitoring", in "The Oxford Handbook of Banking, Second Edition (2 ed.)", Oxford University Press.

Ayyagari, M., Demirgug-Kunt, A., and Maksimovic, V. (2010): "Formal versus informal finance: Evidence from China", The Review of Financial Studies, Vol. 23, No. 8, pp. 3048-3097.

Bakk-Simon, K., Borgioli, S., Giron, C., Hempell, H. S., Maddaloni, A., Recine, F., and Rosati, S. (2012): "Shadow banking in the Euro area: an overview", ECB Occasional Papers No. 133.

Berger, A. N., A. K. Kashyap, J. M. Scalise, M. Gertler, B. M. Friedman (1995): "The transformation of the US banking industry: What a long, strange trip it's been", Brookings Papers on Economic Activity, Vol. 1995, No. 2, pp. 55-218.

Bernstein Research (2014): "Chinese Banks - Initiating coverage; Deleveraging, De- risking and (finally) Diverging?", September.

Borst, N. (2013): "Shadow Deposits as a Source of Financial Instability: Lessons from the American Experience for China", Peterson Institute for International Economics, Policy Brief No PB 13-14.

Brunnermeier, M. K. (2009): "Deciphering the liquidity and credit crunch 2007-2008", The Journal of Economic Perspectives, Vol. 23, No. 1, pp. 77-100.

Cai, J., Garcfa-Herrero, A., and Xia, L. (2015): "China's shadow banking sector: Arbitrage, window-dressing and wealth management products", BBVA Research Working Paper No. 15/30.

Caixin (2012): "Fool's Gold Behind Beijing Loan Guarantees", press article published online on 20 March 2012.

CGFS and FSB (2017): "FinTech credit: Market structure, business models and financial stability implications", Report prepared by a Working Group established by the Committee on the Global Financial System and the Financial Stability Board, May.

Chan, E., Chui, M. K., Packer, F., and Remolona, E. M. (2012): "Local currency bond markets and the Asian Bond Fund 2 initiative", BIS Papers No. 63, pp. 35 - 60.

Chen, Z., He, Z., & Liu, C. (2017): "The financing of local government in China: Stimulus loan wanes and shadow banking waxes", Mimeo.

Chen, K., Ren, J., and Zha, T. (2016): "What we learn from China's rising shadow banking: exploring the nexus of monetary tightening and banks' role in entrusted lending", NBER Working Paper No. 21890.

Claessens, S., and Ratnovski, L. (2014): "What is shadow banking?", IMF Working Paper No. 14/25.

Credit Suisse (2014): "Significant improvement of WMP visibility; positive for China banks", China Banks Sector, August.

Dang, T. V., Wang, H., and Yao, A. (2014): "Chinese shadow banking: Bank-centric misperceptions", Hong Kong Institute for Monetary Research Working Paper, No. 22/2014.

De Fiore, F., and Uhlig, H. (2015): "Corporate debt structure and the financial crisis", Journal of Money, credit and Banking, Vol. 47, No. 8, pp. 1571-1598.

Dollar, P. D., and Wei, S. J. (2007): "Das (Wasted) Kapital: Firm Ownership and Investment Efficiency in China", IMF Working Paper No. 07/9.

Elliott, D., Kroeber, A., & Qiao, Y. (2015): "Shadow banking in China: A primer", The Brookings Institution, Economic Studies at Brookings, Vol. 13.

Engle, R. F., Moshirian, F., and Wong, C. (2015): "Global systemic risk: What's driving the shadow banking system?", Institute of Global Finance Working Paper No. 1.

Financial Stability Board (2012): "Global Shadow Banking Monitoring Report 2012".

Financial Stability Board (2016): "Global Shadow Banking Monitoring Report 2015".

Gennaioli, N., A. Shleifer, R. W. Vishny (2013): "A model of shadow banking", The Journal of Finance, Vol. 68, No. 4, pp. 1331-1363.

Gilbert, R. A. (1986): "Requiem for regulation Q: What it did and why it passed away", Federal Reserve Bank of St. Louis Review, February, pp. 22-37.

Gorton, G., S. Lewellen, and A. Metrick (2012): "The safe-asset share", The American Economic Review, Vol. 102, No. 3, pp. 101-106.

Hachem, K. C., Z. M. Song (2016): "Lqiudity regulation and unintended financial transformation in China", NBER Working Papers, No. 21880.

Hale, G., & Long, C. (2010): "What are the Sources of Financing of the Chinese Firms?", Hong Kong Institute for Monetary Research Working Paper No. 19/2010.

Harutyunyan, A., Massara, M. A., Ugazio, G., Amidzic, G., and Walton, R. (2015): "Shedding Light on Shadow Banking", IMF Working Paper No 15/1.

Hsieh, C. T., and Klenow, P. J. (2009): "Misallocation and manufacturing TFP in China and India", The Quarterly Journal of Economics, Vol. 124, No. 4, pp. 1403-1448.

Hsu, S., Li, J., and Xue, Y. (2014): "Shadow banking and systemic risk in China", Political Economy Research Institute Working Paper No. 349.

IMF (2014): "Global Financial Stability Report - Risk Taking, Liquidity, and Shadow Banking", October.

JP Morgan (2013): "Shadow Banking in China", Economic Research Note, May.

Kane, E. J. (2014): "Shadowy banking: Theft by safety net", Yale Journal on Regulation, Vol. 31, No. 3, pp. 773-807.

Li, C. (2016): "The changing face of shadow banking in China", Federal Reserve Bank of San Francisco, Asia Focus, December.

Li , T. (2014): "Shadow banking in China: expanding scale, evolving structure", Journal of Financial Economic Policy, Vol. 6, No. 3, pp. 198 - 211.

Li, J., and Hsu, S. (2012): "Shadow banking in China", MPRA Paper No. 39441.

Li, J., Hsu, S., and Qin, Y. (2014): "Shadow banking in China: Institutional risks", China Economic Review, No. 31, pp. 119-129.

Lu, Y., Guo, H., Kao, E. H., and Fung, H. G. (2015): "Shadow banking and firm financing in China", International Review of Economics and Finance, No. 36, pp. 40-53.

McLoughlin, K. and Meredith, J. (2017): "The rise of Chinese money market funds", Reserve Bank of Australia Bulletin, March.

Moody's (2017a), "Quarterly China shadow banking monitor", May.

Moody's (2017b), "Measures to curb shadow banking benefit banks but also bring adjustment risks", July.

Moody's (2017c), "China's new guidelines on asset management products are credit positive", November.

Perry, E., and Weltewitz, F. (2015): "Wealth management products in China", RBA Bulletin, pp. 59-67, June.

Pozsar, Z., T. Adrian, A. Ashcraft, and H. Boesky (2010): "Shadow Banking", Federal Reserve Bank of New York, Staff Reports No. 458.

Pozsar, Z., and Singh, M. (2011): "The nonbank-bank nexus and the shadow banking system", IMF Working paper No. 11/289.

Rajan, R. (2005): "Has Financial Development Made the World Riskier?", in "Proceedings of the 2005 Jackson Hole Economic Policy Symposium", Federal Reserve Bank of Kansas City, pp. 313-369.

Shin, H. (2009): "Securitisation and Financial Stability", The Economic Journal, Vol. 119, No. 536, pp. 309-332.

Song, Z., Storesletten, K., and Zilibotti, F. (2011): "Growing like China", The American Economic Review, No. 101, Vol. 1, pp. 196-233.

Standard Chartered (2016): "China - Taking stock of wealth management products", Global Research, June.

Tsai, K. S. (2016): "When Shadow Banking Can Be Productive: Financing Small and Medium Enterprises in China", The Journal of Development Studies, pp. 1-24.

Wang, H., Wang, H., Wang, L., and Zhou, H. (2016): "Shadow Banking: China's Dual­Track Interest Rate Liberalization", Mimeo.

Zhu, N., & Conrad, J. (2014): "The People's Republic of China: Knowledge Work on Shadow Banking-Trust Funds and Wealth Management Products", Consultant's Report for Asian Development Bank.