Exploring Systemic Risk Measurement Issues in Shadow Banks: A Case of an Emerging Economy

Abstract

The development of shadow banks, their exaggerated growth rate and the activities outside the regulatory purview gained prominence. Their activities have the ability to disrupt financial stability. India is one of the countries that has registered the highest growth rate. Therefore, the article attempts to identify the SBs that can threaten the economy by applying market-based measures, applying both traditional and modern approaches. The analysis is based on 11 years of the daily stock return of the companies listed on the National Stock Exchange. The risk emitters in the study period are identified along with the directionality of the risk that can lead to a spill over effect in an economy. Also, the various measurement approaches applied in the study are compared and found that conventional measures underestimate the risk that adds up to the leverage and can pose a greater risk in times of the systemic event. These findings have better implications as informed decisions can be taken by the investors, and the regulators can take preventive steps to curb financial instability.

JEL Classification: G230, F650, G01

Keywords

Financial crisis risk measurement shadow banks systemic risk

Introduction

The presumption that only conventional banking activities create systemic risk (SR) changed after the global financial crisis (GFC) of 2007–2008. Shadow banks (SBs) as a complementary set of institutions, presently prevalent across the global economies have also been considered serious emitters of SR. Shadow banking activities enhance further opacity in the financial intermediation process and add extra fragility to the overall financial system. They, therefore, accelerate SR, add propensity and assailability in the financial system (Fan & Pan, 2020). Globally, in recent years, the activities of SBs have increased manifold thereby bringing a higher level of SR into the economies. Emerging economies are no exception to this phenomenon. Their phenomenal growth has drawn increasing attention from researchers (Ahmad et al., 2019; Du et al., 2016; Swati et al., 2012) and is the subject of analogous research nuances of banking institutions. India being one of the growing emerging economy has also seen heightened growth of SBs during the last decade. Considering the increasing importance of India in the global economic set-up, understanding and negotiating with SR emanating from SBs are imperative. This requires understanding of various measurement tools to capture SR and to apply the most appropriate one in case of SBs. Measurement of SR of the Indian SBs is a challenge as many are not properly regulated and not listed in the stock market.

SR is one of the major concerns in the financial system that can significantly disrupt economic activity and lead to system-wide failure. With the development of financial institutions, financial market, collective behaviour of the market participant along with the regulatory arbitrage build up this risk (Smaga, 2014). It is the risk that institutional distress spreads widely negatively disrupting the economic activities resulting in the system wide failures (Adrian & Brunnermeier, 2009). It unveils the widespread imbalances caused by systemic shocks on other financial intermediaries, markets and the financial system (Arnold et al., 2012). The most common factor of SR is that an institutional failure or an economic turmoil results in a chain of negative economic events (Hendricks et al., 2017). The three main characteristics of SR is the contagion, universality and negative externally (Lin et al., 2018). The transmission of shocks between the interconnected elements of the financial system is the contagion effect which have a negative impact on the real economy as a result of the domino effect from one institution to many other (Lin et al., 2018; Smaga, 2014).

The focus on SR has widened over the time. Before the crisis, the focus was on the spill over effect and its large scale impact, whereas, after GFC along with the contagion effect, the focus is drawn to the disturbance in the activities of the financial system (Smaga, 2014). In other words, the shift from micro to macro prudential thinking that expanded the view of SR development and largely helped in implementation of macroprudential regulations (Thiemann et al., 2020). During this post-GFC phase the risk emitted from SBs has added new dimension to the regulatory framework of each major economies of the globe.

Paul McCulley coined the term SB in 2007 and described it as a group of organizations highly levered up without any access to the public backstop, but in 2011 the Financial Stability Board recognized it as a set of activities and entities engaged in credit intermediation outside the purview of the regulation (FSB, 2011)These SBs are also known as non-bank financial intermediaries (NBFIs) or non-banking financial companies (NBFCs), have an essential impact on monetary policy and financial stability (Kirchner, 2020).

Traditional banking activities are mainly confined to the originate and hold models giving rise to credit risk. To diversify the credit risk, the US Federal Government in the 1960s adopted securitization (Pozsar, 2008). Soon, the bank started adopting market-based intermediaries to transfer its risk giving rise to off-balance sheet activities (Ashcraft et al., 2013; Cetorelli et al., 2012; Luttrell et al., 2012; Pozsar, 2008; Pozsar et al., 2013). Hence, the originate to distribute model took over the originate to hold model (Fein, 2013; Pozsar, 2008; Pozsar et al., 2013). Therefore, it is worth mentioning that government and traditional banks together developed the NBFI now labelled SBs (Cetorelli et al., 2012). The interconnecting activities of bank and non-bank led to the rise of many players that typically include the money market, credit hedge, investment and exchange-trading funds; conduits or special purpose vehicles; and finance, insurance and leasing companies (Swati et al., 2012). So, the accelerated growth of non-banking activities is due to the changing facets of the financial sector.

Two theories offer a deeper understanding of non-banks: the monetary circuit theory (Graziani, 1990, 2003) and financial intermediation theory (Allen & Santomero, 1998; Solomon & Marshal, 2015). These theories explain the mobilization of the resources between the creation and use of money and resource utilization (Scholtens & van Wensveen, 2003). Initially, these theories explain the rudimentary role of traditional banking. But, the failure of these centralized models resulted in the repositioning of the circuit in a decentralized way. In the disintermediation approach, bank activities aligned with non-banks play an essential role in credit intermediation (Allen & Santomero, 1998, 2001; Cetorelli et al., 2012). The theories emphasize the significant role of financial intermediaries, in which banking activities aligned with non-banks facilitate the investor’s decision-making, economic growth and stability (Bossone, 2001; Passarella, 2014; Sawyer & Veronese Passarella, 2017). But on the other side, their faint regulations introduce excessive risk with the potential to disrupt the financial system as a whole (Liang & Reichert, 2012).

The actual mechanism of risk development follows the various steps that make the shadow banking activity more complex and opaque. Being outside the regulatory purview, the bank transforms the risk through the SBs using different mechanisms (Claessens & Ratnovski, 2014). Initially, SBs rely on the bank for credit intermediation through securitization. The loans pooled are securitized and re-securitized into various tranches (Adrian & Ashcraft, 2012) based on the ratings provided by the credit rating agencies. It aims to diversify the risk across the financial system (Claessens & Ratnovski, 2014). The securities are tranched based on the asset quality, the top tranche is safe and mostly liquid. However, the residual tranches, that is, the mezzanine and equity tranches, are exposed to high risk (Moreira & Savov, 2017). During the boom, SBs transform the riskier assets into money like instruments providing liquidity provisioning, giving rise to the asset prices followed by high growth. This surge induces an economic boom with increased investment and growth, especially in riskier sectors. But, during the phase of high market uncertainty, it brings contraction to the liquidity followed by the decline in the asset prices, and the collateral values come down (Moreira & Savov, 2017; Pozsar, 2008), resulting in financial instability. Further, they also avail cheap credit during the boom that aggravates instability during the crisis (Adrian & Shin, 2010; Helgadóttir, 2016). Comprehensively, it is worth mentioning that their activities add vulnerability resulting in system-wide failure (Kirchner, 2020).

As the major source of credit risk, which is also the primary sources of financial instability, the increase in non-performing assets (NPA) and the disintermediation of financial activities escalate the off-balance sheet activities (Altman & Saunders, 1998), which are mostly found among the SBs. With lesser regulations, it bears a higher risk than conventional banks giving a significant rise in credit risk (Erten, 2019). With an aim to increase the market share and profitability, the financing activities of the firm is expanded by compromising the credit quality (Salas & Jesus, 2002). As the short term liabilities are transformed towards long term assets, investors noticing any sense of trouble will refuse to roll over their investment leading to a run resulting in an inability to recycle the cash and the mortgages created by them that builds up liquidity risk (Acharya & Vishwanathan, 2011; Imbierowicz & Rauch, 2014; Pozsar, 2008). The sudden demand from the investors to withdraw investment and insufficient liquidity to cover the obligation builds up liquidity crunches (Diamond & Rajan, 2005). They are highly interconnected with the other components of the financial system resulting in system-wide failure.

Financial institutions are considered systemically important if their failure to meet the obligations develops negative consequences in the financial system leading to a system-wide failure (de Souza et al., 2016) and SBs fall into this category. They are so large and interconnected with the other components of the financial system that their failure has a disastrous impact on the economy as a whole, as evident from the GFC 2007 and Infrastructure Leasing & Financial Services Limited (IL&FS) crisis in the Indian context. In India the regulator, that is, Reserve Bank of India has identified non-deposit taking NBFCs with asset size more than 70USD and above are systemically important.

SBs activities have a shallow margin therefore internal risk absorption capacity is very low. The low margin does not allow the firm to generate enough capital to absorb the risk internally and hence relies on external sources (Claessens & Ratnovski, 2014). Also, being outside the safety net of the deposit insurance and central funding, they cannot absorb the risk; consequently, the risk are exaggerated for the ultimate investors who do not wish to bear them (Fein, 2013). This is the case of India.

The measurement of various facets of risk, including SR, the contributing factors to these risks, the impact of these risks and many other shades of their operations are covered in the contemporary research works. Among all these facets measurement of SR has become a key topic among the researchers (Acharya et al., 2017; Betz et al., 2016; Bostandzic & Weiß, 2018; Giglio et al., 2016; Thiemann et al., 2020). It is more challenging, particularly when most of them are not listed in the financial market, making it difficult to get the market-related price data. This is the case in India.

The timely detection of systemic threats can provide early signals to the regulators to take precautionary steps (Acharya et al., 2013). Therefore, timely measurement of SR is the need of an hour to take preventive and corrective actions and curtail any vulnerability that threatens the financial system’s stability (Bisias et al., 2012). Also, the SBs’ activities lack transparency (Pozsar, 2008; Pozsar et al., 2013); quantitatively measuring and analysing the risk is challenging (Swati et al., 2012). The measurement of SR assesses the systemic importance of the financial institutions and their contribution to the system. However, there are different approaches for measuring SR that helps prevent any economic turmoil (Bisias et al., 2012; Kleinow et al., 2015).

This article tries to explore these measurement angles associated with SR. This investigation is based on both the extensive literature review as well as on the actual measurement of SR. The latter is attempted with the help of empirical data of Indian listed NBFCs. The objective is to trace the measurement incongruity and difficulty. The scheme of the rest of the article is divided into five more sections. The next section provides a brief highlight about the Indian SBs, followed by a detailed literature review. The data and methods are described in the fourth section, followed by data analysis in the fifth section. The last section offers the conclusion.

Indian Shadow Banks

In India, NBFCs perform bank-like credit intermediation activities outside the banking, which is termed shadow banks (Acharya et al., 2013; Gandhi, 2014; RBI, 2015; Sinha, 2013). They are present in India since the 1960s and gained prominence from the 1970s, but RBI started regulating them only after 1990 (RBI, 2021a). They complement traditional banking by reaching the segment not covered by banks (Acharya et al., 2013).

The Indian Companies Act, 1956 initially regulated their activities. However, complexity in their operation had raised a need for an independent regulatory authority. Hence in 1964, NBFC was included under Chapter III of the RBI Act, 1934 (RBI, 2021a). Allowing hire purchase and leasing companies to accept deposits in line of the recommendation of the James Raj Committee increased their number manifold and also attracted many investors. The fall out of this unabated growth resulted in many delinquencies and fraudulent practices, thereby attracted stringent regulation from the regulator. In line with the recommendations of the A C Sinha Committee, RBI made a comprehensive regulatory and supervisory framework in January 1998 to protect the interest of investors and depositors and safeguard their functioning (RBI, 1999). Initially, the regulatory requirement of NBFCs was focused on NBFC-D to ensure that they function on a safe line and the depositor’s interest is protected. Emphasis upon the systemically important institutions based on their asset size was accorded since 2006 with the differential regulation towards the Systemically Important Non-Deposit taking NBFC (SI-ND-NBFC) and Deposit-taking NBFC (NBFC-D) (RBI, 2007, 2015). The guidelines include the minimum capital adequacy requirement of 15%, prudential framework of stressed asset for early reporting and time bound resolution of stressed asset. The regulations also include appointment of credit risk officer with specified roles and responsibility and maintenance of liquidity coverage ratio (RBI, 2019).

Presently, this is the third-largest segment of the financial sector that accounts for 11% of total assets after commercial banks and insurance (RBI, 2020). The ND-NBFC, whose asset size is around 70 USD million or above, is systemically important (SI-ND-NBFC) and constitutes 85.7% of total NBFC assets (Table 1). Apart from NBFC-ND-SI the presence of other NBFCs are also evident in India, but they come under the regulatory purview of other regulatory agencies in India. Insurance hold the largest asset after the commercial bank followed by NBFC-ND-SI (Table 2). In India the hedge funds do not fall into SB category as they are not involved in credit intermediation activity and hence do not pose any financial stability risk (FSB, 2021).

Table 1.

Indian Shadow Banks: Highlights (INR in Billion).

	Mar’07	Mar’11	Mar’13	Mar’16	Mar’19	Sep’20
No of SI-NDNBFC	173	319	418	209	263	292
Total assets	3178.98	7303.66	11177.00	14,288	26635.88	30878.17
Total profit	74.6	156.19	222	298	350.35	381.25
Total borrowings	2119.86	5009.38	7,980	10,335	14438.23	15072.03
Loans and advances	1845.07	4581.73	7,497	10709	18975.27	20515.81
Net NPA	47.685	36.52	89.416	257	645.16	574.44
% of assets of SCB	9.18	10.17	11.20	13.50	16.04	17.00
% of total NBFC	85.32	86.20	89.42	88.90	86.32	85.70
Bank lending to NBFC	526.94	1378.27	2164	4237.35	6270.89	7062.16

Source: Reserve Bank of India.

Table 2.

Categorization of Shadow Banks.

Asset Size	No. of Companies	Asset Size ₹ (Trillion)	% of Asset to the Total Financial System	% of Asset to the GDP
Total financial asset		263		199.63
Insurance company	57	32.04	12.18	16.04
Pension funds	8	5.64	2.14	2.82
Housing finance	101	13.19	5.02	6.61
Mutual fund	44	22.29	8.48	11.17
SI-ND-NBFC¹	292	25.94	9.86	12.99
D-NBFC	64	4.85	1.84	2.43

Source: IRDAI, RBI, AIFI, IBEF.

Indian SBs’ asset size has been expanding at a rate of more than 10% (FSB, 2020). Among the assets, loans and advances form the major part. On the liability side, borrowing is the biggest funding source in which borrowing from banks constitutes the largest chunk. The faster growth of bank funding in recent years signifies the growing inter-connectedness between banks and SBs. In the past few years, it is visibly pronounced that SBs as a whole is the largest borrower of funds in the financial system with gross payable of 13.38 USD trillion and gross receivables of 1.28 USD million as of end-September 2020 (FSB, 2021), which reflects this inter-connectedness (RBI, 2021b). It is also estimated that the SBs failure can result in a huge loss of the banking system that can impact 2.26% of the bank’s Tier 1 Capital (RBI, 2021a)

As compared to other economies the share of SBs in India economy is very less. However, lack of transparency and weak regulations raise systemic concerns even they are expanding. Though they complement (Acharya et al., 2013) and supplement traditional banking, their presence incorporates fragility into the economy (Gandhi, 2014). A point of reference is the recent debacle in IL&FS crisis in 2017 which seriously impacted the top 15 SBs. This has raised the concern of developing SR in the Indian financial system from the bank as well as other financial institutions outside the banking regulations purview. At present SBs under stringent regulatory and supervisory purview but raises concern due to asset-liability mismatch, poor corporate governance and weak risk management (Sivramkrishna et al., 2019).¹

Literature Review

The emergence of the SR depends on the collective behaviour of the financial institutions. Hence, the risk in the system can develop endogenously as well as exogenously (Neveu, 2018). While former set of approaches assumes that the risk of the financial system is prone to exogenous shock and add vulnerability, the latter set believes that SR develops endogenously in the financial system (Kemp, 2017). Under these two broad categories different measurement tools are considered (Tables 3 and 4). The SR contributions measures (Table 3) try to shed light on the relative contribution of the risk to the financial system (Kleinow et al., 2015). Under this category measures like Conditional Value at Risk (CoVaR), ΔCoVar, CoRisk, Granger causality (GC) and principal component analysis (PCA) are considered (Kleinow et al., 2015). Under sensitivity measures (Table 4) marginal expected shortfall (MES), SRISK, lower tail dependence (LTD) and contingent claims analysis (CCA) (Kleinow et al., 2015).

Table 3.

Risk Contribution Measurement Tools: Endogenous Measures.

CoVaR	Measures the stock loss of the financial systems conditional on the stock return of the individual institution	(Adrian & Brunnermeier, 2009)
ΔCoVar	The marginal contribution of a particular institution is captured to the overall systemic risk	(Adrian & Brunnermeier, 2014)
Co-risk	Analyses the tails of the default distributions for pairs of institutions, it analyses how the default risk of an institution affects the default risk of another institution	(Chan-Lau, 2010)
Granger causality	Measures the directionality of relationships or causality of price movements of securities issued by financial institutions	(Billio et al., 2012)
Principal component analysis (PCA)	It is a technique that decompose asset returns of a sample of financial institutions into linkages between those institutions	(Billio et al., 2012, León & Murcia, 2013)

Source: Compiled by authors.

Table 4.

Risk Sensitivity Measurement Tools: Exogenous Measures.

Marginal expected shortfall (MES)	The level of systemic risk is determined by measuring an institution’s losses when the financial system as a whole is doing poorly	(Acharya et al., 2010)
SRISK	By taking leverage, size and the MES of a firm the index is formed	(Brownlees & Engle, 2012)
Lower tail dependence (LTD)	It is a measure of the propensity of a single financial institution to experience joint extreme adverse effects (measured in price returns) with the market	(Weiß et al., 2014)
Contingent claims analysis (CCA)	Measures systemic solvency risk on the basis of market-implied expected losses of financial institutions by producing aggregate estimates of the joint default risk of multiple institutions as a conditional tail expectation	(Jobst, 2014)

Source: Compiled by authors.

Before the emergence of CoVaR, VaR was the most adapted measure (Adrian & Brunnermeier, 2009). This VaR, captures the risk of the financial institutions in isolation (Girardi & Tolga Ergün, 2013; Kemp, 2017). This was developed by J. P. Morgan in the late 1980s but its roots can be traced from the financial crisis of 1990s (Jorion, 2013). It is an easy statistical measure that captures the underlying volatility of assets exposed to financial risk. But it has its flaws, which were exposed during the GFC (Sinha, 2012). The measurement of risk is limited to evaluating the risk of the single institution at a given confidence level (Adrian & Brunnermeier, 2014; Bjarnadottir, 2012) ignoring the risk to the entire financial system. Therefore, the risk assessment is biased and subsequently broadens the risk taking that did not prevent economic loss during the crisis (Trabelsi & Naifar, 2017). After recognizing the several loopholes VaR, several models were developed to capture the endogenous risk build-up and their impact on the economy as a whole. These measures (Table 3) assess the systemic relevance of the individual institutions concerning their own industry, other industries and the total system (Berdin & Sottocornola, 2015). They estimate the role of the respective institutions and groups to the total SR (Berdin & Sottocornola, 2015; Karimalis & Nomikos, 2018; Kleinow et al., 2015). Most of the prominent institution-level measurement approaches are focused on the ‘contribution approach’ by including an implicit or explicit treatment of statistical dependence in determining the role of systemically important financial institutions (SIFIs) causing material distress within a defined system (Kleinow et al., 2015).

A refined version proposed by Adrian and Brunnermeier (2009) CoVaR captures the stress situation of the individual institution that can affect the system as a whole. It addresses the risk below the VaR limit focusing on the tail distribution that is more extreme than the VaR (Adrian & Brunnermeier, 2009). It captures the losses of an individual asset given the failure of the financial system and captures the particular losses exposed to it, that is, the risk spill overs from institution to institution across the whole financial network (Trabelsi & Naifar, 2017). Further, the change in CoVaR (ΔCoVaR) captures the marginal contribution of individual risk to the overall SR (Adrian & Brunnermeier, 2014). It is a tool to assess early warning indicators by capturing tail dependency during systemic events caused by both spill over and common exposure. The difference between the CoVaR conditional on the distress and CoVaR on the normal state is the ΔCoVaR.

Among the regression based measures, Granger causality is applied prominently, which explains whether the past changes in one variable explain the respective changes in other variable (Billio et al., 2012). It measures the degree of connectedness also the directionality of the relationship and provide a clear view about the receiver and causer among the financial institutions during the tranquil periods (Berdin & Sottocornola, 2015). In simple terms, it shows the number of institutions that are connected to single institutions or the return it measures the number of other institutions that causes the return of an institutions (Sedunov, 2016). Higher level of interconnectedness implies the higher risk of spillover and contagion effect during the economic disruptions. The other in this category is PCA try to explain the explicit linkages of the financial institutions and their dependency (Bisias et al., 2012).

As mentioned before, the SR can develop exogenously, the development of the risk in the financial system can adversely impact various institutions. The measures mentioned in Table 4 evaluate the single institutions’ sensitivity during the economic fragility. It can be basically termed as a ‘top-down approach’ where the development of the fragility in the financial system have an impact on the individual financial system (Thiemann et al., 2020). It measures the co-movement of a financial institution when the industry is in distress (Kleinow et al., 2015). They capture the single institutions return when the financial system is at distress.

The various measures (Tables 3 and 4) try to assess the SR based on the assumption of risk development to provide the regulators and the policymakers about practical requirements and assess the financial institutions’ systemic importance. But, all the measures described above are based on the market data. Non-availability of market data is a common phenomenon in emerging and other poor economies as most of the firms are not listed in the stock market which hinders application of these measures appropriately. Even in case of listed firms the real quality of assets of many do not adequately reflected in the market data mostly because of securitization therefore casting doubt about the reliability of all these measures. Further, undisclosed information, particularly about the governance of the firms do not get reflected, that reduces the reliability of these measures. Also, complexity of some of these measures reduces their applicability in many situations.

Along with incorporating the diversity of perspective and also the multi-dynamic aspect of the financial institutions the SR measures also need to capture different dimensions of risk like macroeconomic, interconnectedness or networking, co-dependence, etc. Considering these aspects, the measures can be classified into four groups (Table 5).

Table 5.

Categories of Systemic Risk Measurement.

Macroeconomic Measures/ Aggregate Measures	Cross/Sectional Measures	Granular and Network Measures	Forward Looking Measures
Costly asset-price boom/bust cycles	CoVaR	Default intensity model	Contingent claims analysis
Property-price indicators	Co-risk	Network analysis and systemic financial linkages	Mahalanobis distance
Equity-price indicators	Marginal and systemic expected shortfall	Simulating a credit scenario	Multivariate density estimators
Credit-gap indicators	Distressed insurance premium	Simulating a credit-and-funding-shock scenario	Simulating the housing sector
Macroprudential regulations		Granger-causality networks	Consumer credit
		Bank funding risk and shock transmission	stress test
		Mark-to-market accounting and liquidity pricing	Principal component analysis

Source: Compiled by authors.

The aggregate measures or the macro measures (Table 5) gauge the overall tension in the financial sector. These measures assume that extrinsic factors build up pressure in the economy and try to track and address the system-wide factors that might affect the large group of institutions (Kemp, 2017). They tend to concentrate on aggregate imbalances and serve as a warning signal tracker of unsustainable pressures that build up in the financial system exogenously (Bisias et al., 2012). These measures focus on the multiple factors like business fluctuations and credit cycle, unemployment, inflation and growth those add to the macroeconomic vulnerability and capture the system’s sensitivity to economic shocks (Ferrante, 2018; Kirchner, 2020). Since top-down approach, ignore the risk that develops from the individual institution, are less effective (Arnold et al., 2012), is the limitation.

However, most of the contemporary measures follow bottom-up approach. Some of them are the cross-sectional measures, assess one institution’s dependence over the other. These are simplest and based on public market data which provide a quick and inexpensive approach and reveal the firms which are more volatile and need caution (Acharya et al., 2010) and postulate the relative risk contributions of the individual firm to the financial system. They signal the emerging crisis based on a real-time basis and try to identify the co-dependence or interdependency of one institution on another (Bisias et al., 2012).

In contrast, Granular foundation and network measures are the refined way to exploit individual stress and failure which try to unveil the development of systemic events (Bisias et al., 2012). The interconnectedness among the financial stakeholder is the common factor among the systemic events as evident from the global turmoil and need to be considered to measure SR (Billio et al., 2012). These measures take into account this fact and help tracking the potential channels that propagate through the system, that is, the contagion effect (Bisias et al., 2012; Neveu, 2018). But, to study the financial interlinkages, it is essential to emphasize on many aspects including economic, political, legal and financial. But non-availability of real data on these aspects is a challenge.

The overpriced asset price market can lead to miscalculation; as a result, the severity of the actual loss is much higher than anticipated. Also, the above-mentioned risk measures can accurately predict the risk when distributions are known, but with the limited data exposure, measuring the weight of the tails is flawed, leading to the underestimation of the risk (Neveu, 2018). Also, at the initial stage of the credit intermediation process, the asset quality that are transformed into the various securities wrapping with insurance from the monoline insurers remain hidden (Pozsar, 2008). Therefore, the opacity and complexity develop through the different stages of securitization and hence the development of risk at each step is not well recognized, resulting in vulnerability (Neveu, 2018). Given the information asymmetry and complexity in the financial system may not be able to predict the risk appropriately and accurately (Caccioli et al., 2018) resulting in a gap between risk predicted and actual.

The forward risk measures try to assess the interdependency structure of risk not only by focusing on the linear correlation but also non-linear distress dependencies. They capture the future event of SR along with bank and financial market interdependency (Liu, 2017; Saldías, 2013; Cesare & Pico, 2018). They try to identify the potential risk before the stress can appear in the traditional financial statements (Bisias et al., 2012). The stress detection is more comprehensive rather than relying on a single indicator. They combine the univariate risk model into the multivariate model and track their direct and indirect interdependencies and linkages and provide a specific view of the market direction with relation to the tail risk dependence (Saldías, 2013). They play a significant role in SR management and are applied by international regulators. But, these measures are so intensive that they can be performed only under the regulatory mandate. Moreover, they are periodic and limited to stress scenarios (Acharya et al., 2010).

The SR can be addressed from different angles. But, evaluating the risk by concentrating on single dimension is not enough. Macro measures aggregate the risk, but the granular and forward-looking measures need more detailed information that is not readily available in the public domain, and computations of risk are complex. However, the contemporaneous measures give the flexibility to compute the risk from the available market data and provide the early warning indicators. The appropriateness of these measures depends on the context and availability of data set. Relying on one single measure may not be appropriate to comprehend the entire spectrum of SR.

Over the last few decades all these measures have been applied in different economies. On a time scale we can classify them as measures prominently used during the pre-GFC phase and during the post-GFC era (Table 6). The financial regulators in the pre-crisis arena had difficulties understanding the development of the financial modernizations and their impact on financial stability. Therefore, the VaR model and various regulatory frameworks were adopted to contain the SR (Thiemann et al., 2020). However, the risk model underestimated the risk accumulated in the system leaving the regulators and the financial institute unprepared (Borio & Drehmann, 2009; Thiemann et al., 2020). The application of the risk management framework in the pre-crisis period and its failure were validated in the post-crisis period (Thiemann et al., 2020). These methods’ estimation of the SR proved to be gloomy during the crisis (Sinha, 2012). These models failed as the actual losses were much higher than the predicted models lacking technical expertise. Moreover, they fail to capture the cross-border exposure (Borio & Drehmann, 2009).

Table 6.

Systemic Risk Measurement Techniques.

Pre-GFC	Post-GFC
VaR, network simulation, default intensity model, time varying multivariate density, distress dependence and tail risk	CoVar, marginal expected shortfall (MES), systemic expected shortfall (SES), ΔCoVar, principal component analysis, Gragner causality test, SRISK, systemic risk index, tail event driven network risk model, contingent claim analysis, lower tail dependence

Source: Compiled by authors.

The researchers and academia have acknowledged the failure of the pre-crisis risk management techniques. The measurement techniques that developed after the crisis period considered the various loopholes of the previous measurement techniques. Also, the pre-crisis measurement techniques are based on a forward measurement approach to detect the early warning indicators of the vulnerability in the economy. The post-crisis approach captures the growing interconnectedness, the spill over effect, so that preventive and corrective actions can be taken to curtail any systemic turmoil in the economy.

The growing literature on SR measurement indicates an increasing interest among the researchers and regulators to understand the linkages between the financial institutions and system-wide vulnerabilities during the common distress. Notably, it is also observed that various SR measurement techniques are present to capture the multiple channels of risk dimensions causing SR (Jobst, 2014). However, there are no SR measurement techniques that are consistently applied. Also, there is an absence of any holistic approach that captures the multiple aspects of risk.

Data and Method

Although banks are prone to contribute to SR from individual institutions’ failures, the failure of non-bank financial institutions tends to be more affected by their common exposures to asset price shocks that challenge the sector’s overall resilience. The study uses 11 years daily stock return from the year 2010–2011 to 2020–2021 of SI-ND-NBFC. There are 40 listed SI-ND-NBFC but excluded 13 companies due to the non-availability of the data. The data are collected from the CMIE database and National Stock Exchanges (NSE).

The study applies the three-step procedure:

Step 1: Estimation of SR

For this purpose we are using VaR, CoVaR and ΔCoVaR measures using the GARCH (1,1) type model following Girardi and Tolga Ergün, (2013), which applies three step model by capturing univariate volatality, capturing conditional variance with Dynamic Conditional Correlation (DCC) model of Engle (2002) and finally the estimates of CoVaR and ΔCoVaR is obtained.

The first measure VaRⁱ_q is implicitly defined as the q% quantile, that is,

Pr(Xi ≤ VaRⁱ_q ) = q%, where X_i is the daily stock return of the financial institution VaRⁱ_q is typically the negative return of the financial institution i within q% confidence interval.

CoVaR^j|i_q is the VaR of institutional j conditional on Xⁱ = VaRⁱ_q of institution i. In simple terms it is VaR of an institutuion j conditional on that other institution is on the financial distress, that is, its return being at VaR. CoVaR^j|i_q is defined as the q-quantile of the conditional probability distribution

$P (X^{j} \leq C o V a R^{j | t}_{q} | X^{i} = V a R^{i}_{q}) = q %$

1. Higher the CoVaR of an individual institution indicates the contribution to the overall SR is higher (Adrian & Brunnermeier, 2009).

$⊿ C o V a R_{q}^{^{t}} = C o V a R_{q}^{^{X i = V a R_{q}^{i}}} - C o V a R_{q}^{^{X i = V a R_{50 %}^{i}}}$

The superscripts j or i can refer to losses of individual institutions or of a portfolio of institutions.

ΔCoVaR^j|i_q = (CoVaR of institution j conditional on institution i being at its VaR level) – (CoVaR of institution j conditional on institution i being at its median level)

Step 2: Back Testing for checking the robustness of the SR estimates

This is undertaken by following the approach of Kupiec (1995) and Christoffersen (1998).

Step 3: Finding the interconnectedness of SBs

GC test is applied to find out the degree of connectedness of an institution with the other institutions by considering their weekly stock price:
$X_{t} = \sum_{j = 1}^{m} a_{j} X_{t - j} + \sum_{j = 1}^{m} b_{j} Y_{t - j} + ε_{t}$

$X_{t} = \sum_{j = 1}^{m} c_{j} X_{t - j} + \sum_{j = 1}^{m} d_{j} Y_{t - j} + η_{t}$

where X is said to ‘Granger-cause’ Y if past values of X contain information that helps predict Y above and beyond the information contained in past values of Y alone;

where ϵ_{t and} ᵑ_t are the uncorrelated noise term a_j, b_j, c_j, d_j are the coefficient of the model. Y granger cause X if b_j is non-zero and has p-value less than 5%.

Results Analysis and Discussion

The GARCH estimates (Appendix 1) indicates that the time varying conditional volatility of the returns on SBs stock. The sum of the coefficient estimates of ARCH and GARCH effect is close to one which shows the significant persistence of volatility at 5%. Taking the daily return and the volatility of the stock, the VaR is calculated on a yearly basis. The result (Table 7) shows the loss ranging below 1% over the study period, with a higher level in the recent years which could be because of COVID pandemic impact.

Table 7.
Summary of the Risk Calculated by the Different Measurement Approaches.

Year VaR
CoVaR
ΔCoVaR

Below 1% Below 1% Between 1% and 5% Between 5% and 10% Above 10% Below 1% Between 1% and 5% Between 5% and 10% Above 10%

2011 27 2 21 4 0 4 22 1 0

2012 27 2 25 0 0 4 22 1 0

2013 27 4 23 0 0 5 22 0 0

2014 27 4 27 0 0 4 23 0 0

2015 27 0 27 0 0 2 25 0 0

2016 27 0 26 1 0 0 26 1 0

2017 27 0 26 1 0 0 26 1 0

2018 27 3 23 1 0 0 26 1 0

2019 27 0 26 1 0 1 26 0 0

2020 27 0 6 7 14 0 7 5 15

2021 27 2 22 3 0 0 22 5 0

Source: Computed by Authors.

Since VaR underestimates the SR, the CoVaR and ΔCoVaR are estimated. Taking the example of 2010–2011, the estimated loss of all 27 SBs as per VaR is below 1%, however, as per the CoVaR as well as ΔCoVaR the loss is much higher, indicating underestimation of actual distress by VaR. Both the later measures give almost same result. Over the years the number of SBs falling under each range of risk is very close and most of the SBs fall within 1%–5% of risk category. However, in the first year of COVID (2019–2020) there was a surge in the number of SBs in the above 10% risk category, which was due to sudden fall in the stock prices. Figure 1 portrays the comparative risk of each SB measured through VaR and CoVaR. The underestimation of risk under VaR indicates its failure to capture the tail-end events. This leads to resorting to high leverage by the SBs that put them under severe stress in crisis. Among the SBs IFCI Ltd, Williamson Magor Ltd, Tourism Finance Corporation of India, Pilani Investment & Inds. Corpn. Ltd. and Muthoot Finance are more vulnerable (Appendix 1). They are the major risk emitters and can prove to be more vulnerable in times of crisis.

CoVaR Backtesting

For checking the robustness of CoVaR model, its validity is estimated under the unconditional coverage, and conditional coverage following the Gaussian distribution model at 5% confidence level. The average test statics of LR_uc and LR_ind show the validity of our model. Under the unconditional property the Gaussian assumption is rejected at 5% level signifying that the model underestimates the risk. However, the model satisfies the assumptions under the conditional coverage property by considering the time varying dependency (Appendix 1).

Granger Causality Test

The various measures of SR capture the individual SBs’ risk and their contribution to the overall system leaving aside the aspect of their interconnectedness that is captured through this test. This is employed using the weekly stock return data to see whether the previous changes of one SB can explain the present changes of any SBs. Any significant relationships shows contagion effect. Therefore, this test helps in deciphering the channels of any further systemic crisis. The t and t + 2 weekly stock returns are applied to estimate the pairwise GC between the financial institutions (Appendix 2). The distribution of correlation of return between the financial institutions and the financial system indicates the interconnectedness (Figure 2). The summary result (Table 8) also depicts the interconnectedness of SBs. The connectivity level evaluates the possible connections over the total number of SBs from a single institution.

Figure 1.
Source: Computed by Authors.

Figure 2.
Source: Computed by Author.

Table 8.
Summary of Interconnected Banks.

Shadow Banks 2011–2012 2014–2015
2017–2018
2020–2021

No of Interconnected Banks Shadow Banks No of Interconnected Banks Shadow Banks No of Interconnected Banks Shadow Banks No of Interconnected Banks

PTC 22 Capri Global 7 Paisallo 14 Williamson 9

Williamson 11 PNB Gilts 8 Williamson 6 IIFL 8

PNB Gilts 9 Muthoot 6 Aneri 5 IFCI 7

Bharat Financial Inclusion 3 Vardhaman 6 Industrial investment 4 Vardhaman 6

Capital trust 2 IFCI 4 Cholamandalam 2 REC 2

Source: Computed by Authors.

Under different measures different SBs figure as highest risk emitters and lowest risk emitters (Table 9). But considering the interconnectedness, which is captured through GC test Williamson Magor Ltd continues to be highly interconnected. It implies that any erratic incident in Williamson Magor Ltd may transmit a larger degree of shocks to all other entities. This is also the major risk receiver (Appendix 2). This implies that it can get the major setback due to any uneven development in the other major risk transmitter.

Summarizing the result of the various measurement approaches applied in the study (Table 9), the companies with the highest risk predicted by CoVaR and ΔCoVaR are correlating with one another. The CoVaR measures consider the tail risk unlike VaR; therefore, the risk measurement is more appropriate and quantifies the impact on the system as a whole due to the development of stress in an economy. The CoVaR and ΔCoVaR measure not only disclose a risk an institution might face during the jeopardized scenario but also the contribution to the overall system. However, they fail to measure the interconnectedness, which is rather captured through the GC. The SBs those who appear as high-risk emitters under the CoVaR measures therefore do not appear under GC measure, suggesting that they are rather the risk receivers. GC that measures the directionality of the risk also pointed out that Williamson Magor is one of the major risk emitters, in line with the findings of the other measures, which can spill over the risk in case of any crisis in the economy.

Table 9.
Summary of Highest and Lowest Risk Emitting Shadow Banks (SBs) as per the Variegated Risk Measures.

CoVaR
ΔCoVaR
Granger Causality

2011–2012 2020–2021 2011–2012 2020–2021 2011–2012 2020–2021

SBs with highest risk Bharat Financial Inclusion Cholamandalam Pilani Cholamandalam/ Muthoot PTC Williamson Magor Ltd

Muthoot Finance Muthoot Finance Bharat Financial Inclusion Pilani Williamson Magor Ltd IIFL

SBs with lowest risk Capital India Finance Indian Industrial Trust Ltd Muthoot IFCI IDFC Bajaj Holdings

Capital India Trust IDFC Capital India Finance IIFL Bajaj Holdings IDFC

Source: Computed by authors.

Conclusion

The academic literature indicated that conventional banks have led to the development of SBs. Over the years, they registered an exaggerated growth rate complementing and supplementing the traditional banks. But, being outside the regulatory purview and lack of transparency in their functioning has added stress to the financial system. Indian economy is one of the emerging economies registered with the highest growth in SB activities, and their activities can add to the system’s fragility. They play a viable alternative source of financial stability, especially in times of crisis. The findings of our study which are based on the application of multiple risk approaches, enable us to assess the risk and their spill over effect in detail.

The empirical results highlighted the risk can be captured in a better way through the modern measurement approaches, namely CoVaR and ΔCoVaR those identify the most vulnerable SBs. The Granger causality test identifies their interrelatedness, that may result in contagion effect during the turmoil in an economy. Overall, the article identifies that Williamson Magor Ltd’s marginal contribution to the overall system is high since 2018–2019 as it is facing major debt challenges. The findings are correlating with the test of directionality, which reveals its high interconnectedness. SBs like this are not only connected to other systemically important SBs but also with many others. Therefore, identification of similar SBs on the basis of their interconnectedness help regulator in preventing the future turmoil in the economy. Also enable the investors to make proper decisions concerning their investment.

The result also reveals the exaggerated risk development and uncertainty after the outbreak of COVID-19, which is evident in SBs. The risk generated by the individual SBs is much higher, and their marginal contributions and impact on other institutions are very high comparing the pre-pandemic scenario. The IL&FS² crisis has impacted the other SBs, and presently, the outbreak of crisis has developed a vulnerability in the economy. Focusing on the SBs reveals that the risk of individual institutions and their marginal contributions to other institutions and economies is very high. As the economy is already going through a major setback, any further failure can negatively impact the economy.

However, the results obtained do not reflect the various factors that build up stress in the SBs. Also, RBI identified other SBs as systemically important, but they are not listed in the stock exchanges. Therefore, identification of the vulnerable SBs that pose a threat to the financial system is important. The evidences gathered are based on market-based measurement techniques, so there is a need to explore other techniques that appropriately measure the risk considering the other factors. There is immense scope of further research to arrive at a comprehensive picture of risk emitting SBs in emerging economies and their directionality of the risk.

Year	VaR	CoVaR	ΔCoVaR
2011	27	2	21	4	0	4	22	1	0
2012	27	2	25	0	0	4	22	1	0
2013	27	4	23	0	0	5	22	0	0
2014	27	4	27	0	0	4	23	0	0
2015	27	0	27	0	0	2	25	0	0
2016	27	0	26	1	0	0	26	1	0
2017	27	0	26	1	0	0	26	1	0
2018	27	3	23	1	0	0	26	1	0
2019	27	0	26	1	0	1	26	0	0
2020	27	0	6	7	14	0	7	5	15
2021	27	2	22	3	0	0	22	5	0

Shadow Banks	2011–2012	2014–2015	2017–2018	2020–2021
PTC	22	Capri Global	7	Paisallo	14	Williamson	9
Williamson	11	PNB Gilts	8	Williamson	6	IIFL	8
PNB Gilts	9	Muthoot	6	Aneri	5	IFCI	7
Bharat Financial Inclusion	3	Vardhaman	6	Industrial investment	4	Vardhaman	6
Capital trust	2	IFCI	4	Cholamandalam	2	REC	2

	CoVaR	ΔCoVaR	Granger Causality
SBs with highest risk	Bharat Financial Inclusion	Cholamandalam	Pilani	Cholamandalam/ Muthoot	PTC	Williamson Magor Ltd
Muthoot Finance	Muthoot Finance	Bharat Financial Inclusion	Pilani	Williamson Magor Ltd	IIFL
SBs with lowest risk	Capital India Finance	Indian Industrial Trust Ltd	Muthoot	IFCI	IDFC	Bajaj Holdings
Capital India Trust	IDFC	Capital India Finance	IIFL	Bajaj Holdings	IDFC

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Appendix

Appendix 2.
Network of Shadow Banks (Granger Causality Measure).

2011–2012
2020–2021

Risk Emitter Risk Receiver F Risk Emitter Risk Receiver F

Williamson Magor Tourism_Finance 12.17** Williamson IDFC 9.0305**

Cholamandalam 6.8631** Tourism_Finance 7.6286**

Bharat_Financial 5.1244** Bajaj_Holdings 6.021**

REC 4.9461** Paissalo 5.0518**

Magma 4.6449** REC 4.3064*

Nahar 0.51852* Power_Finance 4.2343*

PNB_Gilts 4.5946* Cholamandalam 3.7237*

Power_Finance 4.2318* Industrial_invt 3.6289*

IIFL 3.9246* Pilani 3.221*

Aneri 3.5381* Vardhaman Magma 11.541**

Bajaj_Holdings 3.5103* IIFL 6.1909**

Vardhaman Tourism_Finance 8.6087** Bajaj_Holdings 6.0184**

Manappuram 3.1099* PTC 5.5945**

Bajaj Holdings Ugro 3.693* Industrial_invt 3.4629*

Bharat Financial Inclusion ALL 4.8424** IFCI Vardhaman 7.3402**

Williamson 9.9527** Bajaj_Holdings 5.3825**

Ugro 3.3136* Magma 3.547*

Capital India Trust Williamson 15.32** Cholamandalam 3.46*

PTC Capital India Trust 30** Nahar 3.4429*

Cholamandalam 22.8** IIFL 3.2608*

PNB_Gilts 22.2** IIFL Paissalo 6.6756**

Pilani 22.3* Williamson 5.6282**

Tourism_Finance 23.9** Capri_Global 4.947**

Paissalo 19.00** PNB_Gilts 4.9243**

Bharat_Financial 31.3** IDFC 4.7622**

Bajaj_Holdings 13** Vardhaman 3.381*

Nahar 12.3** Cholamandalam 3.3387*

Power_Finance 11** Tata_Investment 3.2322*

REC 12** Capital_India_F~e 3.2091*

Aneri 10** Williamson 3.1222*

Industrial_invt 9.43** Capital India Finance Vardhaman 10.725**

Muthoot 9.5093** Capital_Trust_Ltd IIFL 15.366**

Capital India Finance 7.8018** Magma IIFL 5.8538**

IDFC 8.312** REC Vardhaman 3.7938*

Tata_Investment 7.7445** REC 9.2174**

Manappuram 5.5135** Tata_Investment Muthoot 6.2163*

Capri_Global 4.9338** Nahar 6.209*

SREI 4.5659** Industrial_invt 6.0828*

Magma 3.3449**

Paissalo Williamson 4.1617*

Capital_Trust_Ltd Williamson 5.6226

IIFL
Nahar 0.89249*

PNB_Gilts 3.5617*

Capital_Trust_Ltd 3.2211*

Source: Computed by authors.

Note: **1% significance and *5% significance level.

2011–2012	2020–2021
Williamson Magor	Tourism_Finance	12.17**	Williamson	IDFC	9.0305**
Cholamandalam	6.8631**	Tourism_Finance	7.6286**
Bharat_Financial	5.1244**	Bajaj_Holdings	6.021**
REC	4.9461**	Paissalo	5.0518**
Magma	4.6449**	REC	4.3064*
Nahar	0.51852*	Power_Finance	4.2343*
PNB_Gilts	4.5946*	Cholamandalam	3.7237*
Power_Finance	4.2318*	Industrial_invt	3.6289*
IIFL	3.9246*	Pilani	3.221*
Aneri	3.5381*	Vardhaman	Magma	11.541**
Bajaj_Holdings	3.5103*	IIFL	6.1909**
Vardhaman	Tourism_Finance	8.6087**	Bajaj_Holdings	6.0184**
Manappuram	3.1099*	PTC	5.5945**
Bajaj Holdings	Ugro	3.693*	Industrial_invt	3.4629*
Bharat Financial Inclusion	ALL	4.8424**	IFCI	Vardhaman	7.3402**
Williamson	9.9527**		Bajaj_Holdings	5.3825**
Ugro	3.3136*		Magma	3.547*
Capital India Trust	Williamson	15.32**		Cholamandalam	3.46*
PTC	Capital India Trust	30**		Nahar	3.4429*
	Cholamandalam	22.8**		IIFL	3.2608*
	PNB_Gilts	22.2**	IIFL	Paissalo	6.6756**
	Pilani	22.3*		Williamson	5.6282**
	Tourism_Finance	23.9**		Capri_Global	4.947**
	Paissalo	19.00**		PNB_Gilts	4.9243**
	Bharat_Financial	31.3**		IDFC	4.7622**
	Bajaj_Holdings	13**		Vardhaman	3.381*
	Nahar	12.3**		Cholamandalam	3.3387*
	Power_Finance	11**		Tata_Investment	3.2322*
	REC	12**		Capital_India_F~e	3.2091*
	Aneri	10**		Williamson	3.1222*
	Industrial_invt	9.43**	Capital India Finance	Vardhaman	10.725**
	Muthoot	9.5093**	Capital_Trust_Ltd	IIFL	15.366**
	Capital India Finance	7.8018**	Magma	IIFL	5.8538**
	IDFC	8.312**	REC	Vardhaman	3.7938*
	Tata_Investment	7.7445**		REC	9.2174**
	Manappuram	5.5135**	Tata_Investment	Muthoot	6.2163*
	Capri_Global	4.9338**		Nahar	6.209*
	SREI	4.5659**		Industrial_invt	6.0828*
	Magma	3.3449**
Paissalo	Williamson	4.1617*
Capital_Trust_Ltd	Williamson	5.6226
IIFL	Nahar	0.89249*
PNB_Gilts	3.5617*
Capital_Trust_Ltd	3.2211*

Notes

References

1.
Acharya V. , Brownlees C. , Engle R. , Pedersen L. , & Philippon T. (2010). How to measure and regulate systemic risk. Business, 1–18.

2.
Acharya V. V. , Khandwala H. , & Sabri Öncü, T. (2013). The growth of a shadow banking system in emerging markets: Evidence from India. Journal of International Money and Finance, 39(1), 207–230.

3.
Acharya V. V. , Pedersen L. H. , Philippon T. , & Richardson M. (2017). Measuring systemic risk. Review of Financial Studies, 30(1), 2–47.

4.
Acharya V. V. , & Vishwanathan S. (2011). Leverage, moral hazard, and liquidity. The Journal of Finance, 66(1), 99–137.

5.
Adrian T. , & Ashcraft A. B. (2012). Shadow banking: A review of the literature. Federal Reserve Bank of New York Staff report. 10.2139/ssrn.2175144

6.
Adrian T. , & Brunnermeier M. K. (2009). CoVaR.

7.
Adrian T. , & Brunnermeier M. K. (2014). CoVaR.

8.
Adrian T. , & Shin H. S. (2010). The changing nature of financial intermediation and the financial crisis of 2007–2009. Annual Review of Economics, 2(1), 603–618.

9.
Ahmad W. , Pathak B. , & Bhanumurthy N. R. (2019). Understanding systemic symptoms of non-banking financial companies. Economic and Political Weekly, 54(13), 59–67.

10.
Allen F. , & Santomero A. M. (1998). On the theory of financial intermediation. Journal of Banking and Finance, 21(2), 1461–1485.

11.
Allen F. , & Santomero A. M. (2001). What do financial intermediaries do? Journal of Banking and Finance, 25(2), 271–294.

12.
Altman E. I. , & Saunders A. (1998). Credit risk measurement: Developments over the last 20 years. Journal of Banking and Finance, 21(11–12), 1721–1742.

13.
Arnold B. , Borio C. , Ellis L. , & Moshirian F. (2012). Systemic risk, macroprudential policy frameworks, monitoring financial systems and the evolution of capital adequacy. Journal of Banking and Finance, 36(12), 3125–3132.

14.
Ashcraft A. B. , Adrian T. , Ashcraft A. B. , & Cetorelli N. (2013). Shadow bank monitoring (Report No. 638). FRBNY Staff Reports. https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr638.pdf

15.
Berdin E. , & Sottocornola M. (2015). Insurance activities and systemic risk (Working Paper No. 121). SAFE Working Paper. https://ssrn.com/abstract=2701821

16.
Betz F. , Hautsch N. , Peltonen T. A. , & Schienle M. (2016). Systemic risk spillovers in the European banking and sovereign network. Journal of Financial Stability, 25(4), 206–224.

17.
Billio M. , Getmansky M. , Lo A. W. , & Pelizzon L. (2012). Econometric measures of connectedness and systemic risk in the finance and insurance sectors. Journal of Financial Economics, 104(3), 535–559.

18.
Bisias D. , Flood M. , Lo A. W. , Valavanis S. , Flood M. , Lo A. W. , & Valavanis S. (2012). A survey of systemic risk analytics. https://alo.mit.edu/wp-content/uploads/2015/07/A-Survey-of-Systemic-Risk-Analytics.pdf

19.
Bjarnadottir F. (2012). Implementation of CoVaR, a measure for systemic risk. https://www.math.kth.se/matstat/seminarier/reports/M-exjobb12/120807a.pdf

20.
Borio C. , & Drehmann M. (2009). Assessing the risk of banking crises. BIS Quaterly Review, 29–46. https://www.bis.org/publ/qtrpdf/r_qt0903e.htm

21.
Bossone B. (2001). Circuit theory of banking and finance. Journal of Banking and Finance, 25(5), 857–890.

22.
Bostandzic D. , & Weiß G. N. F. (2018). Why do some banks contribute more to global systemic risk? Journal of Financial Intermediation, 35, 17–40. 10.1016/j.jfi.2018.03.003

23.
Brownlees C. T. , & Engle R. (2012). Volatility, correlation and tails for systemic risk measurement (Working paper). New York University.

24.
Caccioli F. , Barucca P. , & Kobayashi T. (2018). Network models of financial systemic risk: A review. Journal of Computational Social Science, 1(1), 81–114.

25.
Cesare A. D. C. , & Pico R. A. (2018). A survey of systemic risk indicators (Occasional Paper). Bank of Italy.

26.
Cetorelli N. , Mandel B. H. , & Mollineaux L. (2012). The evolution of banks and financial intermediation: Framing the analysis. FRBNY Economic Policy Review, 18(2), 1–93.

27.
Chan-Lau J. A. (2010). Regulatory capital charges for too-connected-to-fail institutions: A practical proposal. Financial Markets, Institutions & Instruments, 19(5), 355–379.

28.
Christoffersen P. F. (1998). Economics department of the University of Pennsylvania evaluating interval forecasts. International Economic Review, 39(4), 841–862.

29.
Claessens S. , & Ratnovski L. (2014). What is shadow banking? (IMF Working Papers WP/14/25). https://www.imf.org/external/pubs/ft/wp/2014/wp1425.pdf

30.
de Souza S. R. S. , Silva T. C. , Tabak B. M. , & Guerra S. M. (2016). Evaluating systemic risk using bank default probabilities in financial networks. Journal of Economic Dynamics and Control, 66, 54–75. 10.1016/j.jedc.2016.03.003

31.
Diamond D. W. , & Rajan R. G. (2005). Liquidity shortages and banking crises. The Journal of Finance, 60(2), 615–647.

32.
Du J. , Li C. , & Wang Y. (2016). A comparative study of shadow banking activities of non-financial firms in transition economies. China Economic Review, 46(S), S35–S49. 10.1016/j.chieco.2016.09.001

33.
Engle R. F. (2002). Dynamic conditional correlation: A simple class of multivariate generalized autoregressive conditional heteroscedasticity models. Journal of Business & Economic Statistics, 20(3), 339–350.

34.
Erten I. (2019). The shadow disintermediation of risk sensitive capital [Working Paper]. London School of Economics.

35.
Fan H. , & Pan H. (2020). The effect of shadow banking on the systemic risk in a dynamic complex interbank network system. Complexity, 2020(3), 1–10.

36.
Fein M. L. (2013). The shadow banking charade. SSRN Electronic Journal. https://ssrn.com/abstract=2218812

37.
Ferrante F. (2018). A model of endogenous loan quality and the collapse of the shadow banking system. American Economic Journal: Macroeconomics, 10(4), 152–201.

38.
FSB. (2011). Shadow banking: Strengthening oversight and regulation. https://www.fsb.org/wp-content/uploads/r_111027a.pdf

39.
FSB. (2020). Global monitoring report on non-bank financial intermediation. https://www.fsb.org/2020/12/global-monitoring-report-on-non-bank-financial-intermediation-2020/

40.
FSB. (2021). Global monitoring report on non-bank financial intermediation. https://www.fsb.org/2021/12/global-monitoring-report-on-non-bank-financial-intermediation-2021/

41.
Gandhi R. (2014, August, 25–32). Danger posed by shadow banking systems to the Global Financial System-The Indian case [Paper presentation]. Address at the Indian Council for Research on International Economic Relations Conference on Governance and Development: Views from G20 Countries. Mumbai (Vol. 21).

42.
Giglio S. , Kelly B. , & Pruitt S. (2016). Systemic risk and the macroeconomy: An empirical evaluation. Journal of Financial Economics, 119(3), 457–471.

43.
Girardi G. , & Tolga Ergün A. (2013). Systemic risk measurement: Multivariate GARCH estimation of CoVaR. Journal of Banking and Finance, 37(8), 3169–3180.

44.
Graziani A. (1990). The theory of the monetary circuit. Économies et Sociétés, 24(6), 7–36.

45.
Graziani A. (2003). The monetary theory of production. Cambridge University Press

46.
Helgadóttir O. (2016). Banking upside down: The implicit politics of shadow banking expertise. Review of International Political Economy, 23(6), 915–940.

47.
Hendricks D. , Kambhu J. , & Mosser P. (2017). Systemic risk and the financial system. Palgrave Macmillan (pp. 65–80).

48.
Imbierowicz B. , & Rauch C. (2014). The relationship between liquidity risk and credit risk in banks. Journal of Banking and Finance, 40(1), 242–256.

49.
Jobst A. A. (2014). Systemic risk in the insurance sector: A review of current assessment approaches. The Geneva Papers on Risks and Insurance, 39(3), 440–470. https://econpapers.repec.org/article/palgpprii/v_3a39_3ay_3a2014_3ai_3a3_3ap_3a440-470.htm

50.
Jorion P. (2013). Value at risk: The new benchmark for managing financial risk. Journal of Chemical Information and Modeling, 53(9).

51.
Karimalis E. N. , & Nomikos N. K. (2018). Measuring systemic risk in the European banking sector: A copula CoVaR approach. European Journal of Finance, 24(11), 944–975.

52.
Kemp M. H. D. (2017). Measuring systemic risk.

53.
Kirchner P. (2020). On shadow banking and financial frictions in DSGE modeling (Joint Discussion Paper Series in Economics, No 19-2020). https://www.econstor.eu/handle/10419/216659

54.
Kleinow J. , Nell T. , & Kleinow J. (2015). Determinants of systemically important banks: The case of Europe. Journal of Financial Economic Policy, 7(4), 446–476.

55.
Kupiec P. H. (1995). Techniques for verifying the accuracy of risk measurement models. Journal of Derivatives, 3(2), 73–84.

56.
León C. , & Murcia A. (2013). Systemic importance index for financial institutions: A principal component analysis approach. ODEON No. 7, 127–165.

57.
Liang H. Y. , & Reichert A. K. (2012). The impact of banks and non-bank financial institutions on economic growth. Service Industries Journal, 32(5), 699–717.

58.
Lin E. M. H. , Sun E. W. , & Yu M. (2018). Systemic risk, financial markets, and performance of financial institutions. Annals of Finance, 262(2), 579–603.

59.
Liu X. (2017). Measuring systemic risk with regime switching in tails. Economic Modelling, 67(8), 55–72.

60.
Luttrell D. , Rosenblum H. , & Thies J. (2012). Understanding the risks inherent in shadow banking: A primer and practical lessons learned (Staff Papers No. 18). Dallas Fed, 1–47. https://www.dallasfed.org/~/media/documents/research/staff/staff1203.pdf

61.
Moreira A. , & Savov A. (2017). The macroeconomics of shadow banking. Journal of Finance, 72(6), 2381–2432. 10.1111/jofi.12540

62.
Neveu A. R. (2018). A survey of network-based analysis and systemic risk measurement. Journal of Economic Interaction and Coordination, 13(2), 241–281.

63.
Passarella M. V. (2014). Financialization and the monetary circuit: A macro-accounting approach. Review of Political Economy, 26(1), 128–148. 10.1080/09538259.2013.874195

64.
Pozsar Z. (2008). The rise and fall of the shadow banking system. Regional Financial Review, 44, 1–13. https://www.economy.com/sbs

65.
Pozsar Z. , Adrian T. , Ashcraft A. , & Boesky H. (2013). Shadow banking. FRBNY Economic Policy Review, 1–16. https://www.newyorkfed.org/medialibrary/media/research/epr/2013/0713adri.pdf

66.
RBI. (1999). Report on trend and progress of banking in India.

67.
RBI. (2007). Report on trend and progress of banking in India.

68.
RBI. (2015). Report on trend and progress of banking in India.

69.
RBI. (2019). Report on trend and progress in India 2018–19.

70.
RBI. (2020). Report on trend and progress of banking in India.

71.
RBI. (2021a). Discussion paper on the revised regulatory framework for NBFCs—A scale-based approach.

72.
RBI. (2021b). Financial stability report (Issue 22).

73.
Salas V. , & Jesus S. (2002). Credit risk in two institutional regimes: Spanish commercial and savings banks. Journal of Financial Services Research, 22(3), 203–224.

74.
Saldías M. (2013). Systemic risk analysis using forward-looking distance-to-default series. Journal of Financial Stability, 9(4), 498–517.

75.
Sawyer M. , & Veronese Passarella M. (2017). The monetary circuit in the age of financialisation: A stock-flow consistent model with a twofold banking sector. Metroeconomica: International Review of Economics, 68(2), 321–353.

76.
Scholtens B. , & van Wensveen D. (2003). The theory of financial intermediation: An essay on what it does (not) explain. SUERF Studies, 1, 7–53. https://www.suerf.org/docx/s_903ce9225fca3e988c2af215d4e544d3_143_suerf.pdf

77.
Sedunov J. (2016). What is the systemic risk exposure of financial institutions? Journal of Financial Stability, 24(3), 71–87.

78.
Sinha A. (2012, August 23). Reserve Bank of India, speech on perspective on risk and governance at the risk & governance summit organised. Indian School of Business, Hyderabad and Deloitte, Mumbai.

79.
Sinha A. (2013). Regulation of shadow banking—Issues and challenges. RBI Bulletin, 67(3), 23–36.

80.
Sivramkrishna S. , Gune S. , Kandalam K. , & Moharir A. (2019). Shadow banking in India: Nature, trends, concerns and policy interventions. Review of Economic and Business Studies, 12(2), 29–46.

81.
Smaga P. (2014). The concept of systemic risk. The London School of Economics and Political Science, 5, 1–29.

82.
Solomon I. D. , & Marshal I. (2015). The intermediation functions of finance companies and economic growth: Issues, theory and empirical evidence from Nigeria. Journal of Finance and Accounting, 3(3), 47–56.

83.
Swati G. , Del M. I. G. , & Ötker R. İ. (2012). Chasing the shadows: How significant is shadow banking in emerging markets? The World Bank—Economic Premise, 88, 1–7. https://econpapers.repec.org/article/wbkprmecp/ep88.htm

84.
Thiemann M. , Melches C. R. , & Ibrocevic E. (2020). Measuring and mitigating systemic risks: How the forging of new alliances between central bank and academic economists legitimize the transnational macroprudential agenda. Review of International Political Economy, 28(6), 1433–1458.

85.
Trabelsi N. , & Naifar N. (2017). Are Islamic stock indexes exposed to systemic risk? Multivariate GARCH estimation of CoVaR. Research in International Business and Finance, 42(4), 727–744.

86.
Weiß G. N. F. , Bostandzic D. , & Neumann S. (2014). What factors drive systemic risk during international financial crises? Journal of Banking and Finance, 41(4), 78–96.

2011–2012			2020–2021
Risk Emitter	Risk Receiver	F	Risk Emitter	Risk Receiver	F
Williamson Magor	Tourism_Finance	12.17**	Williamson	IDFC	9.0305**
	Cholamandalam	6.8631**		Tourism_Finance	7.6286**
	Bharat_Financial	5.1244**		Bajaj_Holdings	6.021**
	REC	4.9461**		Paissalo	5.0518**
	Magma	4.6449**		REC	4.3064*
	Nahar	0.51852*		Power_Finance	4.2343*
	PNB_Gilts	4.5946*		Cholamandalam	3.7237*
	Power_Finance	4.2318*		Industrial_invt	3.6289*
	IIFL	3.9246*		Pilani	3.221*
	Aneri	3.5381*	Vardhaman	Magma	11.541**
	Bajaj_Holdings	3.5103*		IIFL	6.1909**
Vardhaman	Tourism_Finance	8.6087**		Bajaj_Holdings	6.0184**
Vardhaman	Manappuram	3.1099*		PTC	5.5945**
Bajaj Holdings	Ugro	3.693*		Industrial_invt	3.4629*
Bharat Financial Inclusion	ALL	4.8424**	IFCI	Vardhaman	7.3402**
	Williamson	9.9527**		Bajaj_Holdings	5.3825**
	Ugro	3.3136*		Magma	3.547*
Capital India Trust	Williamson	15.32**		Cholamandalam	3.46*
PTC	Capital India Trust	30**		Nahar	3.4429*
	Cholamandalam	22.8**		IIFL	3.2608*
	PNB_Gilts	22.2**	IIFL	Paissalo	6.6756**
	Pilani	22.3*		Williamson	5.6282**
	Tourism_Finance	23.9**		Capri_Global	4.947**
	Paissalo	19.00**		PNB_Gilts	4.9243**
	Bharat_Financial	31.3**		IDFC	4.7622**
	Bajaj_Holdings	13**		Vardhaman	3.381*
	Nahar	12.3**		Cholamandalam	3.3387*
	Power_Finance	11**		Tata_Investment	3.2322*
	REC	12**		Capital_India_F~e	3.2091*
	Aneri	10**		Williamson	3.1222*
	Industrial_invt	9.43**	Capital India Finance	Vardhaman	10.725**
	Muthoot	9.5093**	Capital_Trust_Ltd	IIFL	15.366**
	Capital India Finance	7.8018**	Magma	IIFL	5.8538**
	IDFC	8.312**	REC	Vardhaman	3.7938*
	Tata_Investment	7.7445**		REC	9.2174**
	Manappuram	5.5135**	Tata_Investment	Muthoot	6.2163*
	Capri_Global	4.9338**		Nahar	6.209*
	SREI	4.5659**		Industrial_invt	6.0828*
	Magma	3.3449**
Paissalo	Williamson	4.1617*
Capital_Trust_Ltd	Williamson	5.6226
IIFL	Nahar	0.89249*
	PNB_Gilts	3.5617*
	Capital_Trust_Ltd	3.2211*