1
Shadow Banking and Credit to the Private
Nonfinancial Sector
Reinout Kool
De Nederlandsche Bank1
10469095 September 8, 2014 Supervisor: Lex Hoogduin
Abstract: This study shows that shadow banking increases the level of credit-to-GDP, but
that it also entails an increase in the absolute deviation of credit-to-GDP from its trend during financial crises. With data from the Financial Stability Board, the shadow banking system is approximated by aggregating money market funds, finance companies, special purpose vehicles and hedge funds. We find that the impact varies across these sub-sectors.
JEL-code: G21, G23, E510
Key Words: Shadow Banking, Credit
1 Views expressed are those of the author and do not necessarily reflect official positions of De Nederlandsche Bank. I would like to
2
1. Introduction
The Financial Stability Board (FSB, 2013) defines shadow banking as a system of credit intermediation that involves entities and activities outside the regular banking system. The shadow banking system replicates banking activity by converting short-term liabilities into long-term assets. This system offers multiple advantages, such as providing an additional source of funding and alternatives to bank deposits. However, the crisis made clear that this system also poses major threats to financial stability. The lack of prudential regulation makes shadow banks prone to financial risks, such as bank runs. The failure of shadow banks can carry systemic risks through their interconnectedness with the regular banking system.
The rapid growth of the shadow banking system can be attributed to limited regulation and supervision. Even though the shadow banking system was hit hard by the crisis, it still accounts for an important share in the financial sector of many economies. Currently, shadow banking is growing vastly in developing countries. The ten countries with the highest growth rate in 2012 were all developing countries, four of which had growth rates exceeding 20%.
The Financial Stability Board has taken the initiative to coordinate the international policy response to shadow banking. Currently, authorities are improving the collection of information on shadow banks. This does not only enable authorities to increasingly evaluate firm-specific and market risks associated with shadow banking, but it also makes it possible to analyse the effects of shadow banking on real economic activity.
The data used in this research is collected for the FSB’s annual shadow banking monitoring review, which includes data on financial institutions for 25 countries. This data is classified, which makes the use of this data in econometric research unique. Previous studies used self-constructed proxies of the shadow banking system, which are often based on securitization activity or the amount of liabilities of the shadow banking system, e.g.
3
Pozsar et al. (2010), Gallin (2013) and Sunderam (2013). It does not only takes a lot of effort to gather the data, but the gross of the data is also publicly unavailable in many countries. Therefore, such data only allows for economic evaluations, which cannot be supported by statistical analyses. This is the first study, to our knowledge, which uses cross-country data on shadow banking in an empirical study.
This study investigates how shadow banking affects credit. Figure 1 compares the level of credit-to-GDP of advanced countries with a large shadow banking system, i.e. more than 15 percent of the total financial system from 2002 to 2012, to advanced countries with a small shadow banking system. The figure shows that the level of credit-to-GDP of both groups started to diverge in the nineties. This divergence grew especially during the pre-crisis period from 2002 and stabilized during the pre-crisis. This figure signals that shadow bank rich countries tend to have higher levels of credit-to-GDP.
Figure 1. Average credit-to-GDP of countries with a small vs large shadow banking system.
This study is relevant for shadow banking policy makers and regulators. When designing regulation for financial institutions, the effect of regulation on credit intermediation should be taken into account. For example, Hyun and Rhee (2011), Puera and Keppo (2006) and Myers and Majluf (1984) show that capital requirements for banks reduce their lending activity. Next to that, this study is relevant from a monetary policy perspective. Because
4
shadow banks affect credit, the presence of shadow banking may alter the transmission of monetary policy.
The remainder of the paper is organized as follows. Section 2 describes what shadow banking entails. Section 3 discusses how the shadow banking sector can be measured. Section 4 presents the approach followed on how to investigate whether shadow banking affects credit. Section 5 shows the model. Section 6 describes the dataset and limitations. Section 7 presents the results. Section 8 elaborates on the economic impact of the results. Section 9 discusses the implications of the research and finally, Section 10 concludes.
2. What Is the Shadow Banking System?
Why does shadow banking exist? To answer this question, let’s first start with why financial intermediaries such as banks exist in the first place. Financial intermediaries exist because of the presence of information imperfections in financial markets, which leads to three sorts of transaction costs. The first are searching costs, which occurs when borrowers and lenders have to find each other. The second are screening costs, which results from lenders screening borrowers to determine the credit worthiness of borrowers to counteract adverse selection. The third are monitoring costs, which arise when lenders monitor borrowers to prevent moral hazard. Financial intermediaries exist because of economies of scale in transaction costs. They reduce searching costs by being a central point for lenders and borrowers to find each other, and in information collection.
Traditional banks perform two central roles in the economy, namely liquidity creation (e.g. Diamond and Dybvig (1983) and credit transformation (e.g. Diamond (1984)). Banks create liquidity by collecting short-term liabilities in the form of deposits and making long-term loans. Hence, banks create liquidity for depositors by holding illiquid claims against the borrower, while depositors hold liquid claims against the bank. Inherent in liquidity creation is maturity transformation, which is the funding of long-term assets with short-term liabilities. Additionally, banks transform credit risk by diversifying their portfolio; spreading their risk over numerous borrowers.
5
This intermediation role poses banks to several risks. According to Diamond and Dybvic (1983), liquidity creation exposes banks to withdrawal risk. They show that fear of withdrawal of other depositors can trigger a bank run. This can be prompted by large liquidity and maturity mismatches. Next to that, banks face credit risk, which is the risk that borrowers default on their loan, and interest rate risk, which is the risk caused by changes in interest rates.
The role of banks in credit intermediation is crucial to the economy, which results in strict regulation in the form of supervision, regulatory requirements, i.e. as capital requirements and LTV ratios, and market discipline, i.e. disclosure standards.
Shadow banks
The Financial Stability Board (FSB) defines shadow banking as a system of credit intermediation that involves entities and activities outside the regular banking system. The shadow banking system replicates banking activity, but it is spread over a chain of multiple entities.
Figure 1 shows a very simplified version of the shadow banking system. The shadow banks on the right of the chain create deposit-like products, which bring short-term liabilities in the system. The shadow banks on the left of the chain are credit suppliers, which supplies loans to borrowers. These exposures can be transferred to other shadow bank entities in the centre of the chain. These entities pool loans together and structure them in securitized assets. Hence, assets flow from left to right in the chain and funding flows from right to left. Annex 1 provides a more elaborate description of the shadow banking system, including a distinction between its sub-sectors. Also Pozsar (2008) gives an overview of the shadow banking system, in which he defines different types of shadow banks and describes the asset and funding flows within the system.
In the credit intermediation process, shadow bank entities engage in liquidity, maturity and credit risk transformation. Liquidity transformation in the shadow banking system involves
6
the pooling and structuring of illiquid loans into liquid credit products. Maturity transformation is similar to bank maturity transformation; long-term assets are funded by short-term liabilities. Credit risk enhancement is conducted by tranching cash flows of securitized products into new products with different priority claims. The credit quality of senior claims is superior to the credit quality of junior claims.
Figure 1. Shadow banking in the credit intermediation system.
Like regular banks, the shadow banking system is exposed to liquidity and maturity mismatches and credit risk. In addition, the entities in the system are highly interconnected, so problems of specific entities are easily transmitted to the whole system. However, while banks are strictly regulated, shadow banks face little or no regulation (Plantin, 2012). The crisis shed light on the fragility of the system, resulting in a rapidly expanding body of literature on the monitoring and regulation of shadow banking.
The FSB has taken the initiative to coordinate the international policy response to shadow banking. It has set out a framework to monitor activities performed by shadow banks in its report to the G20 (FSB, 2011). Adrian et al. (2013) stress the importance of not only focussing on specific entities in isolation, but to also conduct a macro-mapping exercise. An ECB (2013) study illustrates current statistical data limitations, in which monitoring exercises mainly focus on subsectors of financial institutions. It suggested to gather statistics on shadow banking activities rather than entities, with particular regard to the securities lending and repo markets.
7
Every year, the FSB publishes a global monitoring report with data on 25 jurisdictions and the euro-area. In addition, country-specific studies have been carried out. Bakk-Simon et al. (2012) provide an overview of the euro-area. They show that shadow banking activity in the euro-area is smaller than in the US. Large differences exist across countries, reflecting differences in regulatory and legal structures. Other overviews are given by the Gravelle et al (2013) on Canada; Broos et al (2013) on the Netherlands; Li (2013) on China; Brañanova (2013) on Spain; and Farid (2013) on Malaysia.
The FSB has taken steps to develop a framework to regulate the shadow banking system. This framework consists of an entity-based regulation approach by imposing tighter regulation on specific institutions. Several policy reports have advocated countercyclical capital requirements, for example the Geneva Report (2009) and the Joint FSF-CGFS Working Group (2009) on the procyclicality of leverage. Adrian and Brunnermeier (2009) propose to base capital requirements on the measure of systemic risk of individual institutions. Others focus on an activity-based regulation approach. Kashyap et al. (2010), suggest a minimal haircut on repo transactions. Gorton and Metrick (2010) propose a strict guideline on the use of collateral for securitization and repos.
Literature on shadow banking and credit
Gallin (2013) provides an overview of the amount of short-term funding of the shadow banking system to the nonfinancial sectors of the US economy. He finds that the shadow banking system was a significant, but not a dominant provider of funding to the real economy. However, the drop of shadow banking system’s funding after 2008 was the main cause of the slowdown of credit supply during the crisis. Gallin illustrates the fragility in shadow banking funding over time and how this fragility can be transmitted to the real economy. Nevertheless, by studying funding of the shadow banking system to the nonfinancial sector, he only focusses on the direct impact of shadow banking on credit supply. In addition, he analyses the funding of the shadow banking system as a whole, but does not make any distinction between its sub-sectors.
8
Moreiro and Savov (2012) develop a macroeconomic model of financial intermediation in which households demand liquid assets. They find that shadow banks increase liquidity during good times by using the value of their assets as collateral. However, this is detrimental to financial stability, since fluctuations in uncertainty result in a flight to quality from shadow bank liabilities to safe assets. Consequently, shadow banking activity comes to a halt, prices of assets fall and investment plunges. The low level of liquidity in the economy results in the contraction of shadow banking funding. Henceforth, the collapse of shadow banking liquidity has a prolonging effect on slumps after adverse shocks.
Sunderam (2013) finds that short-term liabilities of the shadow banking system are substitutes for money and respond to money demand. High money demand before the crisis is a possible explanation for the sharp growth of shadow banking sector.
3. Measurement of the size of the shadow banking
sector
There is no consensus in the literature about what entities and activities shadow banking precisely entails. As indicated above, the FSB defines shadow banking as all financial intermediaries outside the traditional banking system that engage in credit intermediation. Pozsar et al. (2010) narrow this definition down by arguing that shadow banks engage in maturity, liquidity and credit transformation outside the traditional banking system and do not have access to central bank liquidity or public sector credit guarantees. Much simpler, Ricks (2010) only includes maturity transformation that takes place outside the traditional banking system. Table 1 presents varying definitions of shadow banking.
Different definitions of shadow banking result in different estimates of the size of the sector, ranging from $10 trillion to $24 trillion for the US (Figure 2). The measurement of shadow banking can be divided into two main categories.
9
The first approach is entity-based by comprising total assets held by shadow banking entities. An advantage of this approach is that the size of sub-sectors and the entities engaging in shadow banking activities are relatively easy to identify, which facilitates the design of appropriate regulation. However, a problem with an entity-based approach is that it is hard to perceive what activities are undertaken by these entities and how these activities develop over time.
The second approach is activity-based, focusing on shadow banking activities. An advantage is that this provides a relatively clear overview of what activities are conducted in the system. For example, securitization activity is better measured with the activity-bases approach than with the entity-based approach. However, a problem is that it is hard to gauge what entities engage in shadow banking activities.
Figure 2. Estimates of the size of the shadow banking system in the US in 2010
Within both approaches, there still exist large differences in measurement as well. The FSB (2013) takes an entity-based approach and measures the overall size by aggregating total assets of all other financial intermediaries (OFIs) outside the regular banking system.2
However, including all OFIs outside the regular banking system results in a deliberate overestimation of the actual shadow banking system. The Financial Crisis Inquiry
2 OFIs according to the FSB are money market funds, finance companies, structured finance vehicles, hedge funds, equity funds, fixed
10
Commission (FCIC, 2010) narrows down this definition and identifies finance companies, hedge funds, money market funds and special purpose vehicles as the shadow banking entities of the OFI sector. 3
3 The Financial Crisis Inquiry Commission is a US commission created to examine the causes of the financial crisis, both domestically
11
12
4. Shadow Banking and Credit Supply
We want to investigate whether shadow banking affects credit. This is done by studying its effect on the level of credit-to-GDP. In addition, we investigate whether shadow banking affects the deviation of credit-to-GDP from its trend.
Level of Credit-to-GDP
Figure 3 illustrates a three-step approach to investigate whether the shadow banking system affects the level of credit-to-GDP.
Figure 3. Approach to study the effect of shadow banking on the level of credit supply to GDP
The first step is to investigate whether shadow banking influences total credit-to-GDP. The shadow banking system is broken down into sub-sectors. The four sub-sectors used in this approach are money market funds (MMFs), finance companies (FCs), special purpose vehicles (SPVs) and hedge funds. Subsequently, it is studied through which channels these sub-sectors affect credit supply. Total credit supply is divided into bank credit and nonbank credit. With the help of the effect of sub-sectors on the two channels, their effect on total credit supply will be analysed.
In the second step, the country sample is divided into advanced and developing countries to investigate whether the effect differs between these two groups.
13
In the third step, the sample period is split into a ‘pre-crisis’ period from 2002 to 2007 and a ‘crisis’ period from 2007 to 2012.
Absolute deviation of credit-to-GDP from its trend
Next to investigating whether shadow banking affects the level of credit-to-GDP, it may also impact the deviation of credit-to-GDP, the so-called credit gap. We analyse its effect on the total deviation, i.e. both the deviation above and below its trend, to study whether shadow banking affects deviation on both sides of the trend.
The credit gap is calculated by measuring the percentage deviation of credit-to-GDP from its trend. Secondly, total deviation is obtained by transforming the credit gap to absolute numbers.
In contrast to previous approach, we do not study through what channel, i.e. bank and nonbank credit, total credit is affected. Data for total credit consists of a long time span, generally 30 to 40 years. However, data on bank and nonbank credit is only available for ten years, resulting in imprecise trend proxies by the HP filter.
Figure 4. Approach to study the effect of shadow banking on the level of credit supply to GDP
14 Hypotheses
This section discusses the main hypotheses obtained from the two approaches above for the shadow banking system. After that, the effect of the sub-sectors on credit is discussed.
Hypothesis 1: The shadow banking system has a positive effect on credit-to-GDP
There are several reasons why the shadow banking system may have a positive effect on credit-to-GDP. The first reason is that the shadow banking system brings an additional source of funding to the economy. Carey et al. (1998) find that finance companies in the shadow banking system tend to serve risky borrowers, which banks normally neglect. Secondly, as argued by Lane (2013), a chain of shadow entities may provide cost-efficiency in credit intermediation as the system enables each individual entity to focus on their specialised expertise. Lower costs free up capital, which can be used for lending. Thirdly, shadow banks may boost competition in the financial sector, which according to Endut and Toh (2009) results in a lower lending rate. Subsequently, the fall in the interest rate causes credit demand to increase.
Hypothesis 2: The impact of the shadow banking system on credit differs across sub-sectors
As each sub-sector has its own function in the intermediation chain, the effect of shadow banking on credit supply is likely to differ across sub-sectors. How these sub-sectors are expected to affect credit supply, is discussed below.
Hypothesis 3: The sub-sectors have a different impact on bank and nonbank credit
It is expected that the impact of the sectors differs across the two channels. Some sub-sectors foster nonbank credit, but reduce bank credit. The expected effects per sub-sub-sectors on both channels are as well discussed below.
15
Hypothesis 4: The shadow banking system has a positive effect on the absolute deviation of credit-to-GDP from its trend
North (2014) argues that shadow banks have the tendency to reinforce boom and bust cycles. Unlike traditional banks, shadow banks are not constrained by regulatory capital requirements. Hyun and Rhee (2011) argue that banks can meet capital requirements by either reducing assets or increasing equity capital. They show that banks prefer to reduce lending instead of to recapitalize. Myers and Majluf (1984) suggest that the issuance of new equity is less preferred due to the pecking-order theory. This theory states that investors assume that managers have superior knowledge about the true condition of the firm. This knowledge allows managers to raise equity when they think the firm is over-valued. Consequently, investors will place lower value to the issuance of new equity. Shadow banks do not face capital requirements, which makes it easier to increase credit during economic upturns. This reinforces the boom and increases the deviation of credit-to-GDP above its trend.
During a crisis, when credit-to-GDP falls below its trend, the shadow banking system may affect the deviation in two ways. Firstly, the shadow banking system may be hit hard due to the lack of sufficient capital buffers. As a result, shadow bank lending contracts heavily, which deepens the bust of credit-to-GDP. On the other hand, the shadow banking system may be more resilient than the traditional banking system. The shadow banking system exists of many different entities, such that risk is diversified over these entities. Problems in several entities may not cause major problems in the whole system.
Hypothesis 5: The impact of shadow banks on credit-to-GDP is relatively strong for advanced countries and relatively weak for developing countries
Figure 5 illustrates the average composition of the financial sector for these two groups. The financial sector in developing countries mainly consists of banks, while the financial sector in advanced countries is more diversified. Sundarajan and Baldwin (2005) shows that diversification of the financial sector improves efficiency in the financial sector. Endut
16
and Toh (2009) find that the emergence of a more diversified financial sector in Malaysia had a downward impact on interest rates and expanded credit supply. Van Leuvensteijn et al. (2008) find that more competition results in a better transmission of monetary policy. Advanced countries will likely have a more efficient financial sector with more effective monetary channels. Therefore, shadow banking in advanced countries is likely to have a higher impact on credit-to-GDP.
Figure 5. Financial sector composition for developing and advanced countries.
Hypothesis 6: The impact of shadow banks on credit-to-GDP differs between the two periods
The share of shadow banking relative to the financial sector has fallen sharply during the crisis period. One reason is that securitization activity decreased heavily. Mortgage-securitization even almost completely came to a halt after 2007 (RBS Global Banking and Markets). Kolm (2012) argues that securitization is key to the shadow banking system, so its collapse may have changed the functioning of the system. Therefore, the effect of shadow banks may have changed since the start of the crisis.
Sub-sectorial hypotheses
This section briefly discusses the expected individual effects of the sub-sectors on credit. Table 2 presents the concomitant signs of the hypotheses.
17
Table 2. Expected signs of the effect of the sub-sectors on credit supply and the credit gap.
Bank Credit Nonbank
Credit Total Credit Credit Gap
MMFs - + +/- +
FCs - ++ + +
SPVs + ++ ++ +
Hedge funds no + +/no +
Shadow Bank +/- ++ + +
Money market funds
MMFs were created as a substitute for bank accounts to circumvent Regulation Q by offering securities that provide higher interest rates than bank deposits. In this perspective, MMFs can be viewed as competitors of banks, which implies that MMFs have an expected negative effect on bank credit supply.
Pozsar et al. (2010) explain MMFs are at the right-hand side of the shadow banking chain and bring deposit-like funding into the system. Therefore, MMFs facilitate the funding of non-bank credit.
In addition, it is expected that MMFs have a significant impact on the deviation of credit-to-GDP from its trend. Coval et al. (2009) argue that the market does not correctly take into account the fact that valuations of highly rated securities may become strongly correlated during an adverse shock. The underestimation of this correlation is an incentive for investment institutions, such as MMFs, to hold insufficient capital buffers against such events (Gennaioli et al., 2012). This can make MMFs prone to runs, resulting in asset fire sales. According to Gallin (2013), these fire sales impair the desire and ability of credit institutions to supply credit.
18
Finance companies
Carey et al. (1998) find that finance companies (FCs) tend to serve risky borrowers, which banks normally neglect. However, there is no clear distinction between safe and risky borrowers, which means that FCs and banks are to a certain extent substitutes. Sherman (1993) find that FCs pose a direct and active competition to banks. Carmichael and Pomerleano (2002) find that in Australia, FCs compete against banks in providing mortgage lending. Due to this competition effect, FCs are likely to reduce bank lending.
FCs are at the left-hand side of the shadow banking chain and are responsible for the supply of nonbank credit. Therefore, it is expected that a larger proportion of finance companies in the financial sector increases nonbank credit supply.
The overall expectation is that FCs provide the economy with an additional source of funding, which fosters total credit supply.
Carey et al. (1998) argue that FCs do not appear to be stable providers of credit as banks. FCs do not collect deposits and are, therefore, not constrained by banking regulation, i.e. capital and liquidity requirements, nor do they have access to safety nets. This makes FCs more flexible during good times, but also more prone to shocks. As a result, absolute deviation is expected to increase when an economy relies more on the funding of FCs.
Special Purpose Vehicles
As explained above, special purpose vehicles (SPVs) are mainly used for the securitization of loans. During the securitization process, risky illiquid assets are transformed into liquid and perceivingly less risky assets. IMF (2010) argues that this liquidity and credit transformation decreases funding costs and allows credit suppliers (banks and shadow banks) to easily convert loans to cash. Next to that, banks were able to avoid regulatory capital requirements through the use of off balance sheet SPVs. As a result, capital that in normal circumstances was kept as a buffer was now supplied in the form of credit.
19
Henceforth, it is expected that securitization, which is captured by SPVs, fosters bank and nonbank credit supply.
As argued by many (i.e. Coval et al. (2009); Mayer (2009); Dou and Wang (2014); Hong et al. (2012)) securitization of mortgages resulted in over-expansion of credit supply, creating the housing bubble. The burst of the bubble resulted in rapid declines in prices of mortgage-related securities. The collapse in prices effectively ended new mortgage securitization. With a theoretical model, Shleifer and Vishny (2010) show that financial intermediaries transmit this volatility in securitization to the real economy through funding. Henceforth, it is expected that SPVs positively affect the absolute deviation of credit from its trend.
Hedge funds
The effect of hedge funds on credit supply is ambiguous. Since hedge funds do not engage in any form of credit intermediation, it can be expected that hedge funds do not have any direct effect on credit. On the other hand, since hedge funds are at the demand-side of the shadow banking system by buying securitized assets, it may be expected that hedge funds indirectly fuel credit. If this is the case, hedge funds increase the deviation of credit-to-GDP in economic upturns. On the other hand, hedge funds can cause lower thrusts of credit during economic downturns. Shleifer and Vishny (2011) show hedge funds contributed to fire sales. As explained above, fire sales negatively influenced the desire and ability of the credit suppliers to lend. Hence, it is expected that hedge funds increase the absolute deviation of credit-to-GDP from its trend.
20
5. Model
We investigate the hypotheses formulated in the previous section with two models. The first model is to capture the effect of shadow banking on the level of credit-to-GDP. The model focusses on its effect on the deviation of credit-to-GDP from its trend.
Model 1: Credit-to-GDP
This model builds on the framework used by Cottarelli et al. (2005), in which determinants of credit-to-GDP are studied for Central & Eastern European and Balkan countries in a random-effect panel model. We estimate the following equation:
In which:
Per capita GDP is an indicator for overall economic development in country i.
Financial openness represents the ease with which capital flows can enter and exit the country. Ağca et al. (2007) show that openness fosters competition among fund providers, resulting in increased availability of credit at lower costs. For their study, they use the Chinn-Ito index, which is an index measuring a country’s degree of capital account openness. Varela (2013) has shown that liberalization of financial flows in Hungary was followed by an expansion of credit supply.
‘Legal rights’ is an indicator to what extent collateral and bankruptcy laws protect the rights of borrowers and lenders and thus facilitate lending. The index ranges from 0 to 10, with higher scores representing laws designed for greater access to credit.
SB reflects the share of shadow banks relative to the total financial sector. Two shadow bank proxies are used. The first is the size of the OFI sector. The second is the more narrowed down proxy, which will be discussed in more detail in the next chapter.
21 Model 2: Deviation of credit-to-GDP from trend
Our second model is used to test whether the size of the shadow banking system relative to the financial sector increases the absolute deviation of credit-to-GDP from its trend. The model is formulated as follows:
In which:
Per capita GDP is an indicator for overall economic development in the country.
Agénor et al. (2013) find that the deviation tends to be larger for developing countries.
Financial openness represents the ease with which capital flows can enter and exit the country. Financial openness results in the reduction of macroeconomic volatility through an easing in borrowing constraints; see for example, McConnel and Perez-quiros (2002) and Blanchard and Simon (2001).
‘Legal rights’ is an indicator to what extent collateral and bankruptcy laws protect the rights of borrowers and lenders and thus facilitate lending. Galindo and Micco (2005) find support that better legal protections significantly reduce credit volatility.
Budget Balance is a measure of crowding-out (Cottarelli et al, 2005). In economic downturns, governments tend to run deficits. This results in higher government borrowing, which may crowd out funding to the private sector. Consequently, the deviation of credit-to-GDP to the private sector from its trend becomes larger.
Dev. Interest represents the absolute deviation of the real interest rate from its trend. Countries with interest rates deviating heavily from their trend (volatile interest rates) are expected to have larger deviations. Hasan and Sarkar (2002) show that the credit supply is highly sensitive to the lending rate. Further information on the construction of this variable is presented in the next chapter.
22
shadow bank proxies are used. The first is the size of the OFI sector. The second is the more narrowed down proxy, which will be discussed in more detail in the next chapter.
6. Data and Limitations
Data for the shadow banking sector is given by the FSB. As stated above, the FSB takes an entity-based approach by measuring the assets hold by financial institutions, which enables the identification of the sub-sectors. This dataset consists of annual data on the assets hold by banks, the central bank, insurance companies, pension funds, public financial institutions and the OFI sector. This OFI sector is divided into MMFs, FCs, SPVs, equity funds, fixed income funds and hedge funds. According to the FSB definition, this OFI sector represents the size of the shadow banking sector. From the narrow definition of the FCIC, a proxy for the size of the shadow banking sector is obtained by including MMFs, FCs, SPVs and hedge funds.
Data for credit is provided by the BIS, which constructed a data series for credit supplied to the private nonfinancial sector. The private nonfinancial sector includes nonfinancial corporations, households and non-profit institutions serving households. This credit contains loans and debt securities. Total credit supply consists of quarterly data and goes back 30 years. Data on bank credit ranges from 2001 to 2012. Nonbank credit is calculated by subtracting bank credit from total credit. This data is annualized by taking the average of the four quarters.
The time span of total credit is sufficiently long to construct a reliable estimate of the long-term trend. Firstly, the quarterly data on credit-to-GDP is annualized. Subsequently, the HP-filter with a lambda of 1600 is used to estimate the trend. Because the time span of bank and nonbank credit is insufficiently long, measurement of the long-term trend might yield incorrect proxies. Therefore, only the trend for total credit supply is calculated.
23
The deviation of credit-to-GDP from its trend is calculated by first obtaining the credit gap, which is the percentage deviation of credit-to-GDP from its trend. Subsequently, this credit gap is converted to absolute numbers, which is shown by Figure 6. We analyse its effect on the total deviation, i.e. both the deviation above and below its trend, to study whether shadow banking increases deviation on both sides of the trend.
Figure 6. Left: Percentage deviation of credit to-GDP from its trend in percentage of its trend for the US. Right: Absolute percentage deviation of credit-to-GDP from its trend in percentage of its trend for the US.
Financial openness is given by the Chinn-Ito index, which measures the degree of capital account openness. This index is based on the binary dummy variables that codify the restrictions on cross-border financial transactions, which are reported in the IMF’s annual report on exchange arrangements and restrictions (2012). The IMF categorizes restrictions into four main classes.4 The index consists of annual data and ranges from -1.86 to 2.44,
with higher scores representing higher financial openness.5
The lending rate of the World Bank is used for the interest rate. This is the bank rate that meets the financing needs of the private sector. Inflation is subtracted to obtain the real lending rate. Again, a trend is calculated for the real lending rate. The deviation from its
4 These four categories are the presence of multiple exchange rates, restrictions on current account transactions, restrictions on capital
account transactions and the requirement of the surrender of export proceeds.
24
trend is measured by a subtraction rather than a percentage deviation from its trend, because at low trend values, small deviations from its trend result in large percentage deviations. For example, if the trend value is 1% and the actual value 2%, the percentage deviation from its trend is 100%. Again, the subtracted deviation is converted to absolute numbers, such that an absolute deviation of the interest rate from its trend results in an absolute deviation of credit supply from its trend. Figure 7 shows the real lending rate and its subtracted deviation from its trend.
Figure 7. The real lending rate and its deviation from the trend for the US.
A Hausman test was performed for both models. Random effects are preferred under the null hypothesis of the Hausman test and conversely, fixed effects are preferred under the alternative hypothesis. The test yields a Prob>chi2 of 0.8527 for the first model, on the level of credit-to-GDP, which indicates this model should include random effects. The test yields a Prob>chi2 of 0.0048 for the second model, on the deviation of credit-to-GDP from its trend, which specifies this model should include fixed effects.
25 Limitations
The dataset contains data for 24 countries and ranges from 2002 to 20126. This means that
the dataset only consists of a maximum of 264 observations, which goes at the cost of the statistical power of this research; it increases the likelihood of type I and type II errors. The statistical power of this research can be improved when more observations are obtained. This can be done by increasing the time span of the data, which is evidently only possible when time lapses, and by the publishment of data on financial institutions in other countries. This can be achieved quite easily, because many developed countries still do not publish data, such as Belgium, Denmark, Sweden, Finland, Austria, New Zealand etc.
The second limitation is the lack of granularity of data on financial institutions. The absence of uniform definitions of shadow banks across countries may have resulted in different measurements of the size of shadow banking sub-sectors. It is expected that this is especially true for FCs. A FC can be defined as a company that supplies small short-term loans to customers, but it can also be defined as a financial entity that extend credit to households and fund themselves with commercial papers. (A formal definition of FCs and their role in the shadow banking system is provided in Appendix 1). This range in definition will probably make data on FCs not of outstanding quality. The definitions of the other sub-sectors are more confined, such that data on these sectors is expected to be adequate.
7. Results
Table 1 in the appendix shows the descriptive statistics. There are several things worth noting. Firstly, the minimum and maximum of total, bank and nonbank credit-to-GDP are far apart, which indicates that there are large differences in the level of credit-to-GDP between countries. Secondly, the maximum level of OFIs to the total financial sector size
6 These countries are Argentina, Australia, Brazil, Canada, Chile, China, Germany, Spain, France, Hong Kong, Indonesia, India, Italy,
26
is almost twice the level of the narrowed down proxy of the shadow banking system. This confirms that the OFI proxy for shadow banking is a deliberate overestimation of the actual size of the shadow banking system. Finally, it shows that the importance of the shadow banking system and its sub-sectors differs between countries, with the minimum and maximum share of shadow banks to the total financial sector size differing considerably.
The remainder of the chapter discusses how each shadow bank sub-sector affects credit-to-GDP and the deviation from its trend. With the help of these analyses, the impact of the total shadow banking system on credit is given.
Money Market Funds
Table 2a and 2b in the appendix show the results for MMFS on, respectively, credit-to-GDP and its absolute deviation. When focussing on bank credit, MMFs in advanced countries had a negative impact on the level of bank credit-to-GDP, which confirms that MMFs and banks are competitors. Its effect was strong before the crisis, but attenuated and became only significant at ten percent during the crisis. This is an indication that MMFs were hit hard and became less competitive during the crisis. Figure 8 illustrates the assets of MMFs compared to the assets of banks and confirms the drop in competitiveness.
Figure 8. MMFs’ assets as a percentage of bank assets for the US, France and Spain. The US and France are presented, because in these countries MMFs play an important role in the financial system. Spain is presented, because Spain contains data on MMFs for this time period.
27
MMFs in advanced countries have an effect on nonbank credit-to-GDP. Again, its effect was strong and significant before the crisis and became insignificant during the crisis. This may be an indication that MMFs do not form a part of the shadow banking chain any longer.
Next to that, the effect in the pre-crisis period is unexpectedly negative. With current data on MMFs, a properly funded explanation is hard to find. A possible reason can be found in the composition of the portfolio of MMFs; MMFs do not only buy securitized assets created by shadow banks, but are, according to McCabe (2010), also large investors in highly-rated corporate bonds. This stimulates firms to attract funding via the issuance of bonds instead of obtaining loans, which decreases the demand for credit.
Hence, MMFs negatively affect total credit supply through the bank and nonbank credit channel. Therefore, MMFs’ effect on total credit-to-GDP is also negative, but it primarily applies for advanced countries in the pre-crisis period. The effect of MMFs on total credit during the crisis was insignificant, but they positively affected the deviation of credit-to-GDP from its trend. This is an indication that MMFs are very prone to adverse shocks, which can result in runs and fire sales.
Finance Companies
Table 3a and 3b in the appendix present the results for the effect of FCs on, respectively, credit-to-GDP and its absolute deviation. The effect of FCs on bank credit supply is ambiguous. Firstly, FCs have as expected a negative effect on bank credit, which indicates that FCs and banks are competitors. However, this effect is not present any more when the sample period is split in half. This might be due to the lack of granularity in the data on FCs.
FCs seem to have only a significant impact on nonbank credit during the crisis. This does not contradict the literature, which states that FCs generally serve risky borrowers. During a crisis, borrowers become more risky, such that they turn to FCs. Therefore, FCs bring an additional source of funding to the economy.
28
The positive effect of FCs on nonbank credit dominates its effect on bank credit in advanced countries during the crisis. FCs have a positive effect on total credit during the crisis. As stated by Carey et al. (1998), greater reliance on the funding of FCs results in more fluctuating credit, such that the absolute deviations of credit-to-GDP becomes larger.
Special Purpose Vehicles
Table 4a and 4b in the appendix show the results for SPVs on, respectively, credit-to-GDP and its absolute deviation. SPVs positively affected bank credit before the crisis in advanced countries, because it enabled banks to move loans, especially mortgages, to off-balance sheet SPVs to avoid capital requirements and to securitize them. The crisis resulted in regulators focussing on the elimination of regulatory arbitrage, which constrained banks to avoid capital requirements. Next to that, demand for securitized mortgage assets fell heavily, so securitization of mortgages almost completely came to a halt. As a result, the effect of SPVs on bank credit became insignificant during the crisis.
The effect of SPVs on nonbank credit also mainly applies for advanced countries. The results show that SPVs positively affected nonbank credit and that this effect remained constant over time. Nonbank credit is mainly fuelled by the securitization of nonbank loans; such as financing for cars, consumer durables and home improvements etc, see Carey et al. (1998). This securitization activity did not fall as heavily as the securitization of mortgages. Hence, this result is an indication that SPVs remain an important part of the shadow banking system during the crisis.
SPVs had a positive effect on total credit, but its effect decreased during the crisis, which signals that securitization activity, and especially securitization of mortgages, fell during the crisis. The collapse of securitization resulted in larger absolute deviations of credit-to-GDP from its trend.
29 Hedge Funds
Table 5a and 5b in the appendix present the results for the effect of hedge funds on, respectively, credit-to-GDP and its absolute deviation. As expected, hedge funds had no real impact on bank credit. On the other hand, hedge funds had a positive impact on nonbank credit before the crisis, which indicates that hedge funds were significant demanders of shadow bank created assets. In order to investigate whether the causality runs from hedge funds to nonbank credit or vice versa; shadow bank products were highly profitable and led to higher growth of hedge funds, a granger causality test was performed on seven advanced countries, including Canada, Spain, Honk Kong, Italy, Japan, Korea and the US7. Its results are presented in table 7a in appendix 2 and show that hedge funds
granger cause nonbank credit in the US, Japan and Italy. The causality runs the opposite direction in Canada, while there is no clear causality present in Spain, Hong Kong and Korea. Hence, although the regression suggests hedge funds boosted nonbank credit, granger causality tests show this only applies for the US, Japan and Italy.
Hedge funds had a positive effect on the deviation of credit-to-GDP from its trend during the crisis in advanced countries, which can be caused by the contribution of hedge funds to fire sales when the crisis hit. As explained above, fire sales impair the desire and ability of credit suppliers to lend.
Shadow Banking System
Table 6a and 6b in the appendix show the results for the shadow banking system on, respectively, credit-to-GDP and its absolute deviation. We find that the shadow banking system as a whole had a positive effect on credit-to-GDP from 2002 to 2012. The effect is larger when a narrowed down proxy for shadow banking is used instead of the OFI sector as a proxy. In addition, its effect on nonbank credit is more significant and constant over time. This indicates that the narrow shadow banking proxy is a better measure of the actual shadow banking system.
30
The results show that the shadow banking system affects total credit through the nonbank credit channel. The significant positive influence of the shadow banking system on nonbank credit remains constant over time. The influence of the shadow banking system on bank credit is insignificant. However, each sub-sector does have a different impact on bank credit; MMFs and FCs hamper bank credit supply, while SPVs increase bank credit supply and hedge funds do not have an influence.
The shadow banking system resulted in a larger absolute deviation of credit-to-GDP from its trend during the crisis. Hence, countries with a large shadow banking system tend to have larger deviations when adverse shocks hit.
8. Economic Impact
Overall, we see that shadow banking significantly impacts credit. The results show us that when the size of the shadow banking system relative to the size of the total financial sector increases by one percentage-point, credit-to-GDP in advanced countries increases by 1.21 percentage points. This means that a financial sector with a shadow banking system consisting of 10% of its size results in an increase of credit-to-GDP of 12.1%, which is a considerable amount.
Furthermore, we see that shadow banking causes the absolute deviation of credit-to-GDP to increase during financial crises. An increase in the shadow banking sector relative to the total financial sector by one percentage point results in deviation to increase by 1.068 percentage points. This means that when the share of the shadow banking sector to the total financial sector is 10%, absolute deviation will increase by 10.68%, which is also substantial.
Below, we discuss whether these results are applicable to the real world. We do so by reviewing how shadow banking affects credit in Australia, the Netherlands and the US.
31
These countries have large shadow banking systems, but the role of shadow banking in their financial systems differs between each other.
Australia
Figure 9 shows the development of credit-to-GDP and of the size of the shadow banking system for Australia. Australia experienced a large growth of total credit of 46 percentage points from 2002 to 2008. In that same period, the size of the shadow banking sector has been steadily declining. Only from 2007 to 2008, shadow banking grew sharply with 1.5 percentage points. The effect of this growth can be seen on bank and nonbank credit. The banking sector and the shadow banking sector appear to be strong competitors of each other, since bank credit and shadow banking move in exact opposite ways. The increase in nonbank credit due to the rise in shadow banking seems to crowd-out bank credit. Nevertheless, total credit is still influenced by shadow banking; the slowdown of total credit during the crisis can be mainly attributed to the collapse of shadow banking. From 2008 to 2011, the size of the shadow banking sector fell with 3.1 percentage points. As a result, credit fell with 9.5 percentage points.
Furthermore, nonbank credit appears to be more sensitive to the movement of shadow banking than our results indicate. The actual increase in nonbank credit from 2007 to 2008 of 11.9 percentage points is notably higher than our estimated increase of 1.21 percentage points.8 The subsequent decrease in shadow banking of 3 percentage points results in a fall
of nonbank credit of 23.2 percentage points. Therefore, a one percentage points change in the size of the shadow banking sector seems to results in an approximate change of 7.7 percentage points in nonbank credit.9
8 The results give a beta of 0.787 for nonbank credit. Therefore, we estimate that a change of 1.5 percentage points in shadow banking
to the financial sector results in a change of nonbank credit of 1.21 percentage points.
9 From 2007 to 2008, shadow banking grew with 1.54 percentage points and nonbank credit with 11.9 percentage points. A one
percentage point change in shadow banking results than in a 11.9/1.5=7.7 percentage points change in nonbank credit. Similarly, from 2008 to 2011, we get -23.2/-3=7.73 percentage points.
32
Figure 9. Development of credit-to-GDP and the size of the shadow banking system for Australia. The narrow shadow banking proxy is used in this figure.
The Netherlands
Figure 10 depicts the development of credit-to-GDP and of the size of the shadow banking system for the Netherlands. As indicated by the results, shadow banking has no visible impact on bank credit. The effect of shadow banking on nonbank credit, however, differs from the results, since nonbank credit seems to react to changes in the size of shadow banking with a lag of one year. In addition, its reaction appears to have become more sensitive over time, which can be seen from figure 11. This is an indication that shadow banking has become more important in the financial system over time.
Through the nonbank credit channel, shadow banking has a strong impact on total credit in the Netherlands. With the large increase in nonbank credit, we see that shadow banking was the main cause of the credit boom from 2008 to 2009. The one percentage point change in shadow banking caused total credit to grow with ten percentage points. Similarly, the collapse of shadow banking was the main cause of the stagnation of credit during the financial crisis.
Figure 10. Development of credit-to-GDP and the size of the shadow banking system for the Netherlands. The narrow shadow banking proxy is used in this figure.
33
Figure 11. The percentage point change in the size of the shadow banking sector relative to the total financial sector size and the percentage point change in nonbank credit-to-GDP. Nonbank credit has a lag of one year, which means that, for example, the ten percentage points change from 2007 to 2008 in this figure actually took place from 2008 to 2009. The narrow shadow banking proxy is used in this figure.
US
Figure 12 shows the development of credit-to-GDP and the size of the shadow banking system for the US. Total credit in the US experienced a rapid growth of 33 percentage points from 2003 to 2008. The shadow banking sector grew with 6 percentage points in this period. According to our estimations, this would mean that total credit was estimated to grow with only 6.7 percentage points. This means that, in this period, either the effect of shadow banking on total credit in the US is stronger than our results suggest or that other factors played a role in the strong growth of credit, such as low interest rates and over-optimism in financial markets.
The fall of credit during the crisis can, however, be explained by the collapse of the shadow banking system. From 2008 to 2012, shadow banking fell with 13 percentage points. According to our estimations, credit should have fallen with 15.3 percentage points, which is close to the actual drop of 16.4 percentage points.10 Total credit mainly fell due to the
drop of nonbank credit. Our estimation of the fall of nonbank credit of 10.2 percentage points is also close to the actual drop of 10.8 percentage points.11 These results are similar
10 The results give a beta of 1.21 for total credit. Hence, we estimate that a decrease of 13 percentage points of the size of the shadow
banking sector relative to the total financial sector results in decrease of total credit of 15.3 percentage points.
11 The results give a beta of 0.787 for nonbank credit. Therefore, we estimate that a change of 13 percentage points in shadow banking
34
to the findings of Gallin (2013), who argues that the contraction in shadow banking in the crisis was the entire reason for the fall in credit in the US.
Figure 12. Development of credit-to-GDP and the size of the shadow banking system for the US. The narrow shadow banking proxy is used in this figure.
Overall economic impact
From these three cases, we see that the role of shadow banking in the financial system can differ substantially between countries. The shadow banking system is a competitor of the traditional banking system in Australia, it is strongly intertwined with the banking system in the US and it has no visible involvement with the banking system in the Netherlands. As a result, its effect on total credit can also differ heavily between countries. Hence, the estimation that a one percentage point change in the size of the shadow banking sector results in a 1.21 percentage points change of total credit-to-GDP is not necessarily applicable to all countries. Although this estimation turns out to be close to the actual value in the US, it is probably an underestimation of the true effect of shadow banking on credit in Australia.
These three cases also make clear that shadow banking was presumably not the only cause for the strong growth in credit before the crisis. On the other hand, the slowdown of credit during the crisis can be attributed to the collapse of the shadow banking system. This shows that the fragility of the shadow banking system to financial shocks can have a large impact on the economy.
35
9. Implications
The literature speaks about the shadow banking system as the way it functioned before the crisis. The findings of this paper shows that the effect of shadow banks on credit changed significantly over time. For example, pre-crisis results show that MMFs and hedge funds significantly affected credit, but this vanished during the crisis. This indicates that these sub-sectors may not be part of the shadow banking credit intermediation chain anymore. A lesson for supervisors is to take this continuously evolving characteristic into account when monitoring the shadow banking system.
In addition, this study is relevant for the regulators of shadow banks. When designing regulation for financial institutions, the effect of the regulation on credit should be taken into account. For example, many banks have expressed criticism in that increased capital ratios hamper their ability to lend. This study can contribute to the evaluation of how regulation of shadow banks affects lending.
This research also has implications for monetary policy makers. Figure 13 shows the transmission mechanism of monetary policy. Shadow banks amplify the effect of monetary policy in two ways. Firstly, because the shadow banking system has an increasing effect on credit, the presence of shadow banks in the financial system amplifies the effect of monetary policy. Secondly, shadow banks may also influence the bank and market interest rate. Mier-y-Teran (2012) finds that increasing banking competition enhances the pass-through of monetary policy; the money market interest rate converges to the official interest rate set by policy makers. Our results show that shadow banks, such as finance companies, bring competition to the market, which therefore may result in the improvement of the transmission of monetary policy. Further research lies in to what extent inflation, price developments in Figure 13, is actually influenced by the presence of the shadow banking system.
36
Figure 13. Influence of shadow banks on the transmission mechanism of monetary policy. The green arrow in the box of the money market interest rate represents the improvement of the pass-through of monetary policy in the money market rate due to competition of shadow banks. The green arrow in the box of credit represents the increasing impact of shadow banks on credit and the amplification effect of shadow banks on monetary policy.
Source: ECB, Transmission Mechanism of Monetary Policy
10. Conclusion
In this paper, we describe how shadow banking affects credit to the private sector. This is done by investigating its impact on the level of credit-to-GDP and on the absolute deviation of credit-to-GDP from its trend for the period of 2002 to 2012. This research uses two proxies of shadow banking: a proxy of the OFI-sector as a measurement of the shadow banking sector and a proxy consisting of the aggregation of money market funds, finance
37
companies, special purpose vehicles and hedge funds. We believe that the latter is a more expedient proxy for the actual size of the shadow banking system.
We find that the shadow banking system has a significant positive effect on the level of credit-to-GDP. As expected, the narrowed down proxy yields higher and more significant results, which indicates this proxy is more accurate for the measurement of the size of the shadow banking system. Furthermore, the shadow banking system mainly affects credit through the nonbank credit channel. The break-up of the system into the four sub-sectors MMFs, FCs, SPVs and hedge funds shows that each sub-sector has their own impact on bank and nonbank credit.
In addition, we find that the shadow banking system increases the absolute deviation of credit-to-GDP from its trend during financial crises. Hence, this study shows that shadow banking results in a higher level of credit-to-GDP, but that it also entails a higher absolute deviations of credit-to-GDP from its trend during financial crises.
Nevertheless, when we apply these results to the real world, we see that the role of shadow banking in the financial system differs between countries. In some countries, shadow banking is strongly interwoven with the traditional banking system, whereas in others, the shadow banking and the banking system are competitors. As a result, its effect on credit can also differ substantially between countries. This means that the significant results are not necessarily applicable to all country.
References
Adrian. Arshcraft. Cetorelli. (2013) “Shadow Bank Monitoring”. Federal Reserve Bank of New York Staff Reports. No. 638
Adrian. Brunnermeier. (2008) “CoVaR”. Federal Reserve Bank of New York Staff Reports 348.
38
Ağca. Gianni De Nicolò. Detragiache. (2007) “Financial Reforms, Financial Openness, and Corporate Borrowing: International Evidence”.
Agénor. Alper. Pereira da Silva. (2013). “Capital Regulation, Monetary Policy, and Financial Stability”. International Journal of Central Banking, Vol. 9, No. 3, pp. 193-238. Bakk-Simon. Borgioli. Giron. Hempell. Maddaloni. Recine. Rosati. (2012) “Shadow Banking in the Euro Area – An Overview”. European Central Bank, Occasional Paper Series No. 133.
Blanchard. Simon. (2001) “The Long and Large Decline in U.S. Output Volatility”, Brooking Papers on Economic Activity, Brookings Institution, pp.135-174.
Brañanova. (2013) “Shadow Banking in Spain”. Bank for International Settlements. Broos. Carlier. Kakes. Klaaijsen. (2012) “Shadow Banking: An Exploratory Study for the Netherlands”. DNB Occasional Studies, Vol. 10/No.5
Brunnermeier. (2009) “The Fundamental Principles of Financial Regulation”. Princeton University, Geneva Report 11
Carey. Post. Steven. (1998) “Does Corporate Lending by Banks and Finance Companies Differ? Evidence on Specialization in Private Debt Contracting” The Journal of Finance, Volume 53, Issue 3.
Carmicheal. Pomerleano. (2002). “The Development and Regulation of Non-Bank Financial Institutions”. The World Bank.
Cottarelli. Dell’Ariccia. Vladkova-Hollar (2005). “Early Birds, Late Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central and Eastern Europe and in the Balkans”. Journal of Banking & Finance, Volume 29, Issue 1, pp. 83-104.
Coval. Jurek. Stafford. (2009) “The economics of structured finance”. Journal of Economic Perspectives 23, pp. 3–26.
Chinn. Ito. (2008) ”A New Measure of Financial Openness”. Journal of Comparative Policy Analysis, Volume 10, Issue 3, pp. 309-322
Deloitte (2012). “The Deloitte Shadow Banking Index – Shedding Light on Banking’s Shadows”.
39
Diamond. (1984). “Financial Intermediation and Delegated Monitoring”. The Review of Economic Studies, Volume 51, Issue 3, pp. 394-414.
Diamond. Dybvig. (1983). “Bank Runs, Deposit Insurance, and Liquidity”. The Journal of Political Economy, Vol. 91, No. 3, pp. 401-419.
Dou. Wang. (2014) “Asset Securitization and Bubbles: An Illustration of Subprime Mortgage Default Crisis”. Advances in Economics and Business 2, pp. 112-119 Endut. Toh. (2009) “Household Debt in Malaysia”. Bank of International Settlements. BIS paper No. 46.
European Central Bank. “Transmission Mechanism of Monetary Policy”. https://www.ecb.europa.eu/mopo
European Central Bank. (2013) “Enhancing the Monitoring of Shadow Banking”. http://www.ecb.europa.eu/pub
Farid. (2013) “Monitoring Shadow Banking and its Challenges: the Malaysian Experience”.
Financial Stability Board. (2011) “Shadow Banking: Strengthening Oversight and Regulation”.
Financial Stability Board. (2013) “Global Shadow Banking Monitoring Report 2013”. Financial Crisis Inquiry Commission. (2010) “Shadow Banking and the Financial Crisis,” FSF-CGFS Working Group. (2009) “The Role of Valuation and Leverage in
Procyclicality”.
Gennaioli. Shleifer. Vishny. (2012) “A Model of Shadow Banking”. Journal of Finance 58, no 4, pp. 1331-1363.
Galindo. Micco. (2005) “Creditors Protection and Credit Volatility”. IDB Working Paper No. 439.
Gallin. (2013) “Shadow Banking and the Funding of the Nonfinancial Sector”. Federal Reserve Board.
40
Economic Activity.
Gravelle. Grieder. Lavoie. (2013) “Monitoring and Assessing Risks in Canada’s Shadow Banking Sector”. Bank of Canada, Financial System Review.
Geneva Report 11. (2009) “The Fundamental Principles of Financial Regulation.
Gorton, Metrick (2010) “Regulation the Shadow Banking System”. Brookings Papers on Economic Activity.
Hassan. Sarkar. (2002) “Banks’ option to lend, interest rate sensitivity, and credit availability” Review of Derivatives Research 10-2002, Volume 5, Issue 3, pp 213-250 Hong. Chang. Kaustav. (2012) “The Impact of Mortgage Securitization on the Housing Bubble and Subprime Mortgage Crisis: A Self-organization Perspective”. International Research Journal of Applied Finance, Vol. 3, Issue 6, pp. 830
Hyun. Rhee. (2011) “Bank Capital Regulation and Credit Supply”. Journal of Banking & Finance, Volume 35, Issue 2, pp. 323-330.
IMF. (2010) “Securitization: Lessons Learned and the Road Ahead”
IMF. (2012). “Annual Report on Exchange Arrangements and Exchange Restrictions” Kashyap. Stein. Hanson. (2010) "An Analysis of the Impact of .Substantially
Heightened.Capital Requirements on Large Financial Institutions"
Kolm. (2012) “Securitization, Shadow Banking and Bank Regulation”. University of Vienna
Lane (2013) “Shedding Light on the Shadow Banking System.”
Li. Qin. (2013) “Shadow Banking in China: Dimensions, Risks, Prospects”. National Natural Science Foundation of China, Project No. 71173246.
Van Leuvensteijn, Sorensen, Bikker, van Rixtel. (2008) “Impact of Bank Competition on the Interest Rate Pass-Through in the Euro Area”. ECB Working Paper Series No 885. Mayer. (2009) “Housing, Subprime Mortgages, and Securitization: How did we go wrong and what can we learn so this doesn’t happen again?” Columbia Business School & NBER
41
Mayer. Pence. Sherlund. (2009) “The rise in mortgage defaults”. Journal of Economic Perspectives 23, pp. 27–50.
McCabe. Patrick. (2010) “The cross section of money market fund risks and financial crises”. Divisions of Research & Statistics and Monetary Affairs
McConnell. Perez-Quinos. (2000) “Output Fluctuations in the United States: What Has Changed Since the Early 1980s?” American Economic Review, vol. 90, pp. 1464-76. Mier-y-Teran. (2012) “Bank Competition and the Transmission of Monetary Policy”. UCLA
Moreira. Savov (2013) “The Macroeconomics of Shadow Banking”. New York University
Myers. Majluf. (1984) “Corporate Financing and Investment Decisions when Firms Have Information That investors Do Not Have” Journal of Financial Economics, Issue13, pp. 187–221
North. (2014) “Dissecting Shadow Banking”. Digital Finance Analytics.
Peura. Keppo. (2006) “Optimal bank capital with costly recapitalization” Journal of Business, 79 (4), pp. 2163–2201
Plantin (2012). “Shadow Banking and Bank Capital Regulation”. Toulouse School of economics and CEPR
Pozsar. (2008) “The Rise and Fall of the Shadow Banking System”. Moody’s Regional Financial Review
Pozsar. Adrian. Ashcraft. Boesky. (2010) “Shadow Banking”. Federal Reserve Bank of New York. Staff Report No. 458.
Ricks. (2010) “Regulating Money Creation after the Crisis,” Harvard Law School. Royal Bank of Scotland (RBS), Global Banking and Markets. mib.rbs.com
Sherman. (1993) “Survey of Asset-Basses and Other Competitive Lending Activities”. The Secured Lerider, pp.18-34.
42
Shleifer. Vishny. (2011) “Fire Sales in finance and Macroeconomics”. Journal of Economic Perspectives, Volume 25, pp. 29-48.
Shleifer. Vishny. (2010) “Unstable Banking’. Journal of Financial Economics 97, pp. 306–318
Sundararajan. Baldwin. (2005). “Regulation of financial conglomerates: Is unification of financial sector supervision the answer”. Regulation of Financial Intermediaries in Emerging Markets, pp.118-162
Sunderam. (2013) “Money Creation and the Shadow Banking System”. Harvard Business School.
Varela. (2013) “Financial Liberalization, Competition and Productivity”. Job Market Paper.
