Printer Friendly



The moment that Wall Street has long been dreading and financial lending institutions in the shadow banking sector have hoped for has happened. According to Tankersley (2016), the December 2016 increase in the benchmark federal funds interest rate for the first time in a year announced by Federal Reserve Chair Janet Yellen could be the first of several hikes in 2017. The effect of a monetary tightening on bank lending and activity in credit risk transfer (CRT) markets is less certain as of today because rising rates came at the end of an unconventional monetary policy stimulus that previously restored the securitized CRT capital markets, stabilized the housing markets, and made loans less costly. Reported in the same article,

Economic projections... indicate that the Fed now expects the economy to grow 1.9% in 2016 and 2.1% in 2017. The projections show that the group expects the Fed to increase rates three times in 2017, to a rate of 1.4% by year's end. Its September projections signaled only two expected hikes next year. Analysts [however] have warned that if Trump and Congress agree to slash tax rates and increase spending in areas such as infrastructure the Fed could be forced to raise rates faster than expected to counter rising prices.

Savers will cheer higher rates on their savings, and lenders' net interest margins will decompress. Some economists and legislators, however, criticize the change in the monetary stance because they want rates to remain as low as possible to boost employment as long as inflation remains below the Federal Reserve's target. For investors, the all-time highs in stock and bond valuations will likely fall, and borrowing should be less affordable. For banks and other financial institutions that borrow in capital markets, low-valued institutions should theoretically have a more difficult time attracting investors including those participating in CRT capital markets (Borio and Zhu 2012).

This study makes a contribution to the literature by using a structural factor-augmented vector autoregression (FAVAR) methodology to empirically examine the impact of monetary policy actions on different types of bank lending and CRT capital market instruments over two separate time periods (1995-2006 and 2007-2015) to account for the Great Recession. Specifically, impulse response functions help to determine the effects of a monetary tightening (measured as a 25 basis point increase in the shadow monetary policy rate) on bank mortgage, consumer credit, and business loans along with securitized asset, secondary market syndicated loan, and credit derivatives CRT instruments. Although several other studies show that the institutional environment influences the composition of banks' loan portfolios (see, e.g., Bedendo and Bruno 2012), few researchers examine how the monetary policy stance transmits its changes through the shadow banking system to bias bank lending due to liquidity in CRT capital markets. Empirical findings for this issue are important because bank lending and volumes in CRT capital markets fund spending to help drive economic growth.

The results reveal that a contractionary monetary policy stance tends to shift bank lending toward consumer-related loans fueled by liquidity in CRT capital markets. As CRT capital market liquidity changes the relationship between monetary policy actions and bank lending to the household and business sectors, monetary policy can amplify vulnerabilities in the financial system involving, in part, bank leverage and nontransparent CRT instruments. The important implication for policymakers is that monetary policy may not be sufficient to reduce the vulnerabilities inherent among closely interconnected bank loan and CRT capital markets because lending migrates away from the business sector toward households that historically have higher rates of delinquency during the past two economic downturns.

It is well known that excessive consumer debt has a negative effect on the economy. For example, student loans outstanding have quadrupled since 2004 according to the Federal Reserve Bank of New York Consumer Credit Panel at the same time that defaults have nearly doubled. Particularly after the recent recession, individuals burdened with student loan debt are less likely than their counterparts to seek out mortgages or auto loans, which also means that there are less risks taken and fewer new small businesses started that drive economic activity (Gorman 2015).

My results show strong evidence of "bank bias" from less bank-syndicated corporate loans in both periods, more single-family mortgages and credit card debt in the pre-crisis period (1995-2006), and expanded debt post-crisis (2007-2015) in all consumer lending sectors (auto, credit card, and student loans) following a monetary tightening. My results exist after controlling for overall macroeconomic performance (e.g., unemployment, gross domestic product), short-term interest rates that affect the cost of financing for banks, and other forms of nonbank finance. These findings extend Lang-field and Pagano's (2016) "bank bias" result in which banks have a tendency to overextend and misallocate credit when asset prices rise, then ration credit when asset prices fall. The authors conclude that asset price changes are an amplification mechanism that leads to higher systemic risk and lower growth in bank-based countries. My unique extension is that "bank bias" is exacerbated by contractionary monetary policy in the United States that has a financial system with a healthy share of market-based intermediation.

The empirical analysis is also motivated by comments made by Summers (2015), the former Treasury secretary, who blogs that the rate of monetary tightenings at this time is a large economic risk. Interrelated bank loan and CRT capital markets tend to reinforce the monetary policy effects found in my study. Consistent with Bertay, Gong, and Wagner's (2017) examination of securitization and growth, I find in the pre-crisis period that issuance volumes of agency mortgage-backed securities (MBS) and consumer-related asset-backed securities (ABS) increase prior to the recent recession following a monetary tightening. Therefore, increases in bank and securitized mortgages and credit card loans are evidence of monetary policy changes inducing consumer-related lending through effects on "bank bias" and investors' willingness to lend in these CRT capital markets to satisfy their investment needs.

Even in the post-crisis period with heightened awareness and more monitoring by policymakers, net liquidity measured by CRT volumes outstanding increases for consolidated agency MBS, private-label MBS, and consumer-related ABS backed by auto and student loans. Despite the increase in securitized business loans and credit derivatives outstanding following a contractionary policy rate shock, there remains no move toward bank business loans post-crisis. Apparently, "bank bias" is not structurally related to the business cycle. As CRT capital markets continue to provide liquidity to some bank loan sectors across estimation periods, the dampened effectiveness of monetary policy actions suggests further regulation of this part of the shadow banking system.

In support of expanded regulation, additional findings show that relative credit in individual credit markets measured as the share of CRT out of bank and CRT debt tends to move toward private-label securitized mortgages away from bank debt in both estimation periods. Before the recent recession in consumer credit markets, securitization is relatively more important, but in the post-crisis period, credit to the business sector flows toward securitized and insured CRT instruments away from bank business debt. The common thread among these CRT markets is that they are generally less regulated and transparent capital markets favored by investors to generate high returns and by borrowers to increase liquidity (Pagano and Volpin 2012), which makes it more difficult for effective monetary policy. Therefore, the current regulatory environment leaves contractionary monetary policy less able to influence credit because of CRT capital markets that insulate some loan markets, which may increase systemic risk. Changes in relative credit among interdependent bank loan and CRT capital markets should be a concern to policymakers given that financial assets in the U.S. shadow banking system total over $26 trillion in 2014, which still exceeds assets held in the conventional, regulated banking system well after the official end of the recent financial crisis (Financial Stability Board 2015).

An important question that has received minimal attention by researchers is whether the four-legged stool--bank loans, securitized assets, traded loans, and credit derivatives--insulates the economy from unfettered lending and liquidity in the shadow banking system when the stance of monetary policy is tight. This study supplements the pricing and balance sheet literatures by re-examining the balance sheet management of banks and its relationship to volumes in the primary and secondary CRT capital markets for household and business loans. In a changing macroeconomic environment, it is important to understand different capital market participants' lending and investing incentives in order to grasp the magnitude of systemic risk across the shadow banking system.

The rest of the paper is organized as follows. Section II emphasizes the importance of CRT to an evolving financial system. A review of the related literature is presented in Section III, which is followed by a discussion of the theoretical foundations for the hypotheses in Section IV. The sample selection process, data description, and empirical methodology are outlined in Section V. The results are presented in Section VI along with a summary and discussion in Section VII. Section VIII concludes the paper.


One of the most damaging aspects of the 2007-2009 financial crisis is the near collapse of the entire financial system brought on by failures in CRT capital markets according to Stein (2010). Investors, households, and other capital market participants have been told to brace themselves because auto loans, credit cards, student loans, and other consumer-related debt may be the next financial market to implode. For example, according to Prater (2008), few capital market participants are aware of the credit risk inherent in the $ 1 trillion credit card loan industry and that consumers today are less able to pay off their debts. She states that few investors have a detailed understanding of the terms and complexities of credit card securitization deals. Moreover, unlike large corporations, small businesses with no direct access to capital markets often rely heavily on personal financial tools such as auto, credit card, and mortgage loans that become the underlying assets in CRT capital market transactions (Wilcox 2011). A repeat collapse of any consumer or commercial debt market will make CRT capital markets less liquid and limit banks' funding opportunities.

Nearly a decade after the Great Recession, the question of how CRT affects the financial system continues to need additional investigation. Calluzzo and Dong (2015) highlight the changing nature of credit risk from 2005 to 2011 in the sense that financial institutions tend to be less risky individually after the recent financial crisis from legislated monitoring and penalties (e.g., Dodd-Frank Act of 2010 and Basel III), but the financial system as a whole has become more vulnerable to systemic contagion due to integration. Although the underlying financial intermediation process that affects lending and liquidity in the overall macroeconomy is speculative, this study conjectures that the three CRT capital markets have made bank lending more biased toward consumer-related credit and some markets less influenced by monetary policy actions.

The significance of CRT capital markets to the financial system cannot be overstated. As part of the broader shadow banking or market-based financial system, U.S. market-based financial assets total over $26 trillion in 2014, exceed assets held in the conventional banking system, and represent approximately half of all (conventional and shadow bank) lending (Financial Stability Board 2015 ). (2)

Table 1 below summarizes volumes outstanding for the three different CRT instruments that are the focus of this study--securitized assets, secondary market syndicated loans, and credit derivatives--across three different credit markets for mortgages, personal consumer finance (auto, credit card, and student loans), and businesses. (3) High return-seeking investors bought these liquid securities at unprecedented levels before the recent financial crisis with the belief that the transactions fit within their diversification and risk management strategies. According to Table 1, all outstanding securitized asset markets and traded syndicated loans combined were over $10 trillion outstanding, while notional amounts of credit derivatives to insure against the credit risk of financial assets purchased by both buyers (beneficiaries) and sellers or providers (guarantors) of protection were over $13 trillion in 2006.

After the 2007-2009 financial crisis, volumes in some CRT capital markets experienced significant declines. Syndicated loans fell to $117 billion in 2009 from a height of $342 billion 2 years earlier, while securitized private-label residential mortgage assets dropped to $1.5 trillion in 2015 from $3.3 trillion in 2006. As a signal of economic recovery and less fear of defaulting financial assets, credit derivatives usage (beneficiary and guarantor) falls to $10.5 trillion in 2015 from a height of $31.2 trillion in 2007. Of the markets and types of CRT instruments reported in Table 1, securitized consumer credit and business loans along with syndicated loans trading at or above par appear to be approaching their pre-crisis levels more quickly than other CRT capital markets; mortgage REITs comprised mainly of MBS surpass their pre-crisis levels by 2011. The uneven recovery suggests something other than overall market conditions driving CRT activity, which would affect all capital markets. Alternatively, the varying recovery pattern intimates that interconnected capital markets for bank loans and CRT instruments affect monetary policy actions aimed to influence bank liquidity.

Few empirical studies consider whether monetary policymakers should explicitly account for the different CRT capital market instruments into their policymaking process. Singh, Stone, and Suda (2015) conjecture that the Great Recession created awareness of the need for central banks to anticipate how different financial sectors respond to monetary policy actions, especially given that the Federal Reserve is signaling a series of interest rate increases. This study helps to bridge this gap. Although monetary policy effects on bank lending and CRT capital market pricing have been investigated extensively, few studies evaluate how a monetary tightening will impact liquidity (issuance and volumes outstanding) among bank lending and CRT capital market sectors for mortgages, consumer credit, and corporate loans. It is important to evaluate each of these sectors in order to understand the role of CRT instruments in capital markets.

It is generally agreed that CRT capital markets play a critical role in facilitating more lending to household and business borrowers at lower fees and interest rates, but their role following a monetary contraction has been given minimal attention over the last several years due to the low interest rate environment. As an integral part of the financial system with a common purpose to unbundle and trade credit risk, policymakers, regulators, and capital market participants typically discuss two related economic benefits that allow CRT usage to facilitate lending. First, funded transactions that involve the sale of financial assets to third parties in securitization and traded syndicated loan deals increase liquidity in the banking sector and in capital markets from tradable securities that are created. Second, banks can use any of the three CRT instruments that are the focus of this study to change the credit risk profile of their held loans and free up regulatory capital of banks to achieve similar economic benefits as additional liquidity from asset sales. These economic benefits from CRT help to insulate banks from adverse liquidity shocks including those from monetary policy contractions.

In the low interest rate environment that partially brought about the Great Recession, securitized CRT activity helped to meet growing demand by investors for high-yielding assets and for subprime loans to back these transactions. As the recent financial crisis progressed, declining CRT issuance volumes and higher funding costs theoretically reduce other forms of available credit to households and businesses. In the CRT capital market for credit derivatives, contractual obligations as a guarantor against the credit risk of securitized assets brought on the collapse and near failure of American International Group (AIG) in 2008. (4)

Evidently, the Great Recession proves that efficient CRT activity can be illusory. Several economists (e.g., Cantor and Rouyer 2000) note that CRT is only beneficial to the financial system in terms of less systemic risk if credit risk is, in fact, transferred away from the banking sector to investors such as hedge funds, pension funds, insurance companies, and other financial institutions. Even if credit risk is appropriately transferred, banks can try to increase their value through higher leverage made possible through increased liquidity and regulatory capital relief. Information problems may arise in CRT capital markets when transferred credit risk creates incentives to screen and monitor borrowers less, which was the case in the run up to the subprime mortgage crisis according to Keys et al. (2010). Importantly, leverage and information costs amplify the severity of changing market conditions that affect financial asset values and possibly financial system stability (Brunnermeier and

Sannikov 2014). These benefits and shortcomings of CRT usage vary across instruments creating different paths for monetary policy to operate.


What is missing in the literature is a direct test of whether both CRT capital market transaction volumes and bank lending are affected by contractionary monetary policy shocks. If CRT capital markets enable banks to lend after a monetary policy contraction by buffering the tight monetary policy stance, it is expected that CRT volumes should also rise following an increase in the monetary policy rate. Alternatively, if an increase in the monetary policy rate lowers transactions in the primary (originating) markets of both banks and CRT capital markets, the policy action may lead to credit rationing due to an amplification effect. The sensitivity of securitized assets, traded loans, and credit derivatives to monetary policy rate shocks, however, receives minimal attention even though it is a key assumption of most literatures examining loan supply.

A limited number of papers show that banks' access to individual CRT capital markets facilitates lending under more difficult credit conditions following a monetary tightening (Drucker and Puri 2009; Loutskina 2011; Minton, Stulz, and Williamson 2009). In general, less influence of monetary policy actions on bank loan volumes and on output from less CRT activity during economic slowdowns is reported for the precrisis period (Kuttner 2000; Stanton 1998). However, cyclical downturns can negatively affect CRT issuers as borrowers in capital markets and dampen any countercyclical effects from CRT activity (Altunbas, Gambacorta, and Marques-Ibanez 2009; Gambacorta and Marques-Ibanez 2011). A frequently conjectured hypothesis is that structural changes in the financial sector alter the nature of the monetary policy transmission mechanism in general (Bernanke 2007; Sellon 2002).

Theoretical studies of financial intermediation applicable to monetary policy effects on CRT center on self-reinforcing links between liquidity and risk. Deteriorating assets can reduce profits (Diamond and Dybvig 1983), increase the liquidity risk of demand deposits (banks) and rollover risk (nonbanks) as evidenced by the recent financial crisis (He and Xiong 2012), and lead to higher financing rates and funding costs from perceived credit risk (Gorton and Met-rick 2012). Borio and Zhu (2012) suggest that these self-reinforcing links may have a material effect on monetary policy effectiveness. Therefore, the failure by policymakers to take into account the effects of monetary tightenings on liquidity in several endogenously related CRT capital markets can lead to greater systemic risk in the financial system. This gap in the literature is the focus of the present study in that volume changes in primary (issuance) and secondary (outstanding) CRT capital markets are considered along with volumes outstanding for the reference assets, which represent banks' access to liquidity. I focus directly on banks' ability to fund themselves as a supplement to the extensive pricing literature rather than pricing effects that result from investors trading CRT instruments in secondary capital markets (see, e.g., Jimenez et al. 2012; Maddaloni and Peydro 2011).

Inherently, financial institutions can alleviate their information costs associated with held assets by offloading the assets' credit risk onto CRT capital markets to achieve liquidity. The extent to which investors are willing to take on credit risk and lend to financial institutions depends on, however, information costs related to the financial institutions (Bernanke 2007; Kashyap and Stein 1995). The bank lending model of Disyatat (2011) offers support to the above empirical papers that find less influence of monetary policy on loan supply from increased CRT activity due to less credit-sensitive liquidity costs to monetary policy actions. Similar to the above discussion on perceptions in Gorton and Metrick (2012), Borio and Zhu (2012) describe how the monetary policy stance influences lenders' perceptions of bank soundness, which impact investors' willingness to lend to financial institutions. The present study explores a different question that puts the focus on liquidity found in different CRT capital markets within the shadow banking system.

Monetary policy actions, perceptions, and information costs all vary substantially with macroeconomic conditions and financial institutions' balance sheets. If financial institutions poorly manage risk by taking on more leverage, high return but risky CRT instruments sought out by capital market investors only reinforce a vulnerable financial system to tighter credit conditions from contractionary monetary policy actions (Jimenez etal. 2012; Rajan 2006). Moreover, monetary policy actions that operate through financial institutions' balance sheet management reduce liquidity from term spread compression and impact lending in Adrian and

Shin (2010). Variations in short-term interest rates from monetary policy actions (through their impact on the yield curve) affect financial institutions' incentive to lend. During economic slowdowns, it is proposed that a smaller term spread reduces lending. The present study allows for different paths for monetary policy effectiveness based on changing economic conditions through CRT capital markets that are endogenously related.

In the pre-crisis period of the present study, a robust economy, low interest rate environment, little regulatory oversight, and search for yield create a situation in which investors may pay limited attention both to the liquidity in CRT instruments that is impacted by lax lending standards (Keys et al. 2010) and to signals of changing economic conditions, including those brought on by monetary policymakers. During this period, the information advantage of CRT issuers is greater, which generates liquidity for borrowers (Pagano and Volpin 2012) but subjects capital market participants and the economy to systemic risk. Faia (2010) describes how systemic risk arises in her model that combines CRT capital markets, information costs, liquidity, risk taking, and monetary policy. CRT capital markets incentivize banks to sell loans to solve their own liquidity needs and meet future investment demand (also see Maddaloni and Peydro 2011). From less monitoring, banks take on risk and become less liquid after an adverse shock, which can lead to greater financial system instability. In the post-crisis period and in response to the Great Recession, financial institutions monitor and screen more efficiently as increased regulation and investor attention create more transparent economic environments. During this time in my study, however, borrowers continue to issue in and investors continue to seek out individual CRT capital markets that are relatively less transparent to facilitate liquidity.

As lenders and borrowers respond to market conditions including the current low interest rate environment (Ball 2015), their funding and risk management objectives can create shifts within a number of credit markets. As a result, changes in relative credit of CRT capital market activity out of bank and CRT debt reflect the monetary policy effects on credit market liquidity provided by CRT capital markets. A related empirical study by Bertay, Gong, and Wagner (2017) highlights how shifts in relative spending that impacts economic growth rely on structural features of CRT capital markets. Similarly for the bank loan markets, Langfield and Pagano (2016) define movements among loans as "bank bias" in which asset price changes act as an amplification mechanism that leads to greater systemic risk and less real activity (also see Bernanke, Gertler, and Gilchrist 1999).


A key question in the financial and macroeconomics literatures is whether the U.S. shadow banking system can promote financial stability given that the credit risk of the underlying assets in CRT capital market transactions is transferred to market participants who may be unaware of the true risk and illiquidity associated with these assets. This study empirically examines whether CRT capital markets remain liquid following a monetary tightening. In maintaining liquidity, CRT capital markets improve bank intermediation for different types of consumer and commercial loans within an opaque time period characterized by limited investor attention and regulatory oversight (Kishor and Newiak 2013; Young and Bologna 2015) in the pre-crisis period (1995-2006) relative to the hyper-vigilant and transparent period that includes the recent financial crisis and recovery period (2007-2015).

The lack of transparency and investor inat-tentiveness prevalent in the pre-crisis period is driven (in part) by the complexity of CRT instruments and issuers who want a broad investor base. When investors and regulators obtain more useful information about the underlying credit and liquidity risk of different CRT instruments due to greater investor attention, information asymmetries between potential exchange parties decrease. Accounting audit stringency and detailed documentation along with intense regulatory scrutiny provide additional credible information to investors and lower potential opportunism from issuers in the CRT capital markets.

Several related theories predict the relationship between these variables and monetary policy actions. (5) There are two distinct channels reflected in Hypotheses 1 and 2 below by which monetary policy actions affect volumes in lending markets in an effort to influence the macroeconomy. Multiple financially interconnected sectors in consumer and commercial loan capital markets are comprised of banks and three CRT financing tools. The first group of theories involves the first two hypotheses described below. Banks can raise funds and generate liquidity as borrowers in CRT capital markets through transferred credit risk but face information costs that affect CRT issuance and origination volumes (Bernanke 2007; Kashyap and Stein 1995), In turn, Disyatat (2011) suggests that the effect on banks' liquidity affects banks' willingness to provide credit to bank borrowers.

The first set of the hypotheses looks at whether liquidity increases in the mortgage, consumer credit, and business loan sectors through investors' transactions in the primary and secondary CRT capital markets more in the pre-crisis period than in the post-crisis period. Since the benefits and costs of CRT vary across instruments, the relationship between monetary policy actions and liquidity in the primary and secondary CRT capital markets should vary according to the credit market.

H1A: CRT origination volumes in primary capital markets for securitized agency (private-label) mortgage assets increase (maintain levels or increase) more (less) in response to a monetary tightening in the pre-crisis period than in the post-crisis period. H1B: CRT origination volumes in primary capital markets for securitized consumer credit (mortgage REITs and syndicated loans) increase (decrease) more (less) in response to a monetary tightening in the pre-crisis period than in the post-crisis period. H1C: CRT volumes outstanding in secondary capital markets for securitized mortgages and consumer credit (securitized business loans and net credit derivatives) increase (maintain levels or increase) more (less) in response to a monetary tightening in the pre-crisis period than in the post-crisis period. HID: CRT volumes outstanding for traded syndicated loans decrease (maintain levels or increase) following a monetary tightening in the pre-crisis (post-crisis) period.

The second set of hypotheses examines banks' willingness to lend to consumers and businesses. After origination, banks can either hold loans on-balance sheet or access CRT capital markets as an issuer in the sale of loans into a securitization transaction, syndication, or as a beneficiary in the credit derivatives capital market to insure against the default risk of loans. Building on the above hypotheses, CRT activity in the shadow banking sector increases the ability of banks to extend loans.

H2: The volumes outstanding of bank mortgages, consumer credit, and business loans increase more in response to a monetary tightening in the pre-crisis period than in the post-crisis period.

Theories related to movements in originated loans and transferred credit risk explore compositional shifts in banks' balance sheets in a way that sustains "bank bias," which is an amplification mechanism through banks loan and CRT activity that impacts liquidity. In both the primary and secondary CRT capital markets, Bertay, Gong, and Wagner (2017) propose that CRT activity transfers credit risk away from consumer-related assets to drive liquidity into this consumer credit sector but at the expense of real activity from less business investment. These changes in relative credit as described in Langfield and Pagano (2016) are amplified by rising asset prices, which leads to systemic risk (see Adrian and Shin 2010).

The third set of hypotheses explores relative credit shifts between CRT and underlying/reference loans within credit markets caused by a tight monetary policy stance. Declines in the relative CRT measure imply that lending and liquidity are more scarce in CRT and thus loan markets. Rises in the relative CRT measure intimate compositional shifts in capital markets toward CRT activity and away from traditional bank loans.

H3A: Volumes outstanding for relative securitized private-label mortgages (traded syndicated loans) increase (decrease) more in response to a monetary tightening in the pre-crisis period than in the post-crisis period.

H3B: Volumes outstanding for relative securitized business loans, net credit derivatives, and securitized agency mortgages (consumer credit) increase (decrease) following a monetary tightening when moving from the pre-crisis period to the post-crisis period.


The FAVAR model estimation uses 345 data series (including the monetary policy rate) over a sample period from January 1995 to December 2015. Two main sites provide most of the series: the Federal Reserve Economics Database (FRED) and IHS Global Insight. Of all the series, 288 are used to estimate the macroeconomy, and 56 series are used to construct indicators to measure activity in CRT capital markets. Section A1.1, Appendix S1, details the construction of the key indicator series.

B. Empirical Methodology

This study's empirical framework to analyze monetary policy effects on bank loans and CRT instruments is a FAVAR model first developed by Bernanke, Boivin, and Eliasz (2005). A FAVAR model incorporates information in a large number of data series summarized by a few common factors into a conventional vector autoregression (VAR) model of monetary policy analysis. The model enables us to evaluate the effect of policy rate changes on any series in the large data set. Since the data set provides comprehensive information on macroeconomic activity, the model mitigates potential omitted variable bias sometimes found in a conventional VAR setup.

A factor model decomposes the driving forces of a large data set of economic variables X, into a small number of common factors Z, that are linear combinations of the original variables. A factor model is represented as

(1) [X.sub.t], = [LAMBDA][Z.sub.t], + [e.sub.t], t = 1,...,T,

where X, is a N X 1 vector of observed variables, A is a N x K factor loading matrix relating the factors to the observed variables, [Z.sub.t] is a K x 1 vector of factors common to all variables, and [e.sub.t] is a N X 1 vector of idiosyncratic disturbances.

Our application of the FAVAR model applies some structure to the large set of economic variables by grouping variables with similar economic concepts following Belviso and Milani (2006). Section A 1.2 further develops the structural FAVAR model, discusses the identification of the monetary policy shocks and two-step estimation procedure, and explores the statistical properties of the estimated factors.


Although the intent of a monetary tightening is to restrict the supply of funds to households and businesses and discourage spending, it is unknown whether CRT capital market activity moderates the relationship between an increase in the shadow policy rate and macroeconomic activity through the banking lending sector. Using impulse response functions, the empirical analysis simultaneously examines the impact of monetary policy actions on different parts of the shadow banking system for different types of household and business credit over the pre-crisis (1995-2006) and post-crisis (2007-2015) time periods in a comprehensive manner. The 25 basis point increase in the shadow policy rate elicits responses in the dynamic system of variables that measure the behavior of liquidity over time and in percentage magnitude for issuance and volumes outstanding in the bank lending and CRT sectors along with variables that proxy for overall macroeconomic activity.

Relying on impulse response functions in addition to economic and finance theories, the results are consistent with a tight monetary policy stance affecting bank lending and liquidity through securitized mortgage and consumer credit debt channels as well as syndicated and securitized business loan channels across estimation periods. Additional findings reveal changes across estimation periods of the relative importance of CRT activity in credit markets following contractionary policy rate shocks through conventional and unconventional tools.

A, Bank Loans

Recall that in the pre-crisis period (1995-2006), it is assumed that the information environment is opaque in Figures S4 and S5, Appendix S1. (6) Volumes in Figure S4 for bank-originated corporate loans that are syndicated but not yet traded (i.e., primary market syndicated loans) appear to decline but show no statistically significant response to an unexpected increase in the monetary policy rate over the pre-crisis period until 3 years after the shock. Investment grade traded loans show no significant decline out 4 years following a monetary tightening. Likewise, in Figure S5, bank extended credit for auto loans, student loans, and business loans (total consumer credit) over the same period appear to rise (fall), but the responses are not statistically significant. Multi-family and commercial mortgages do show a brief and shallow rise for 2 months following an increase in the policy rate reaching a maximum increase of 0.19 and 0.10%, respectively. This group of results suggests that a tight monetary policy stance is not a strong signal to banks that induce management to decrease lending in these areas.

The finding that banks choose not to increase lending in the corporate loan and commercial mortgage sectors (beyond the first 2 months) is surprising given that this sector did not experience a relatively large number of defaults and frequently had variable interest rates. These results are consistent with "bank bias" as presented by Langfield and Pagano (2016) during an opaque period of banks' tendency to amplify the effects of monetary policy actions by underextending credit to the business sector and reallocating credit to the residential mortgage and credit card loan sectors. For these two loan sectors, a contractionary policy rate shock results in a total maximum increase of 4.1 (1.5)% for bank single-family mortgages (credit card loans) in Figure S5.

In the post-crisis period with greater monitoring by bank regulators in a more transparent environment, it appears that banks continue to decrease loans to businesses following a contractionary interest rate shock. In Figure S6, all types of primary market syndicated corporate loans decrease appreciably following a monetary tightening, which is in contrast to the statistically insignificant behavior found in the pre-crisis period. Leveraged syndicated loans experience the largest maximum drop of 1.05% with significance following the initial tightening for half a year. For underlying loans in this CRT capital market, contractionary policy actions are effective in the post-crisis period.

Similar to the pre-crisis period, a monetary tightening is not a strong signal to banks for management to cut back on lending from 2007 to 2015. Figure S7 shows that to varying degrees, bank auto, student, and credit card loans all tend to rise following a monetary contraction. The difference between the pre- and post-crisis period for mortgage debt is that the brief and immediate increase in bank multi-family mortgages is not significant, and commercial mortgages display a maximum increase of 0.28% with significance for 4 months following the initial policy rate shock. Overall, the findings are consistent with "bank bias" as presented by Langfield and Pagano (2016) in the more transparent period. I conjecture that this "bank bias" is driven by near zero interest rates that persist in both time periods. Similar to the pre-crisis period, banks continue to amplify the monetary policy effects with less credit extended to the business sector while extending more credit to residential mortgage and other consumer credit markets.

B. Securitized Assets (Primary and Secondary Markets)

In general, banks' incentives to securitize their assets should increase unless the spread on loans rises because loan demand also increases reflecting a strong macroeconomy that is invariant to rising short-term interest rates. Low interest rates and a perceived healthy economy both found in the pre-crisis period usually increases banks' willingness to lend. To analyze this issue, the next part of the study explores how CRT issuance volumes respond to contractionary monetary policy used to restrict lending to all sectors of the economy. Are the above findings for the behavior of bank loans to households the result of CRT investors who prefer residential mortgages and credit cards and believe that issuance volumes credibly signal their derived value and demand for structured finance products?

Issuance volumes in the pre-crisis period for securitized consumer-related assets do, in fact, increase in Figure S4. Following a monetary tightening, securitized agency mortgages (single-family and multifamily) display a sustained and significant increase out 1 year reaching a maximum of 1.07%. It appears from the behavior of total securitized mortgages, those securitized by private-label issuers tend to fall during the first year following a monetary contraction, delaying the statistically significant increase in total securitized mortgages from month 13 to month 30. For this securitized asset sector, the maximum increase of 0.32% is reached at month 21. In comparison, issuance volumes for auto and credit card loans both display a significant rise for over 1 year following the initial contractionary policy rate shock. However, response volumes for auto loans are nearly five times the maximum of credit card loans at 1.70 and 0.36%, respectively. (Responses for student loan issuance volumes remain somewhat flat and are insignificant.) The differences in the responses of the issuance volumes reveals that, prior to the recession, residential mortgage lending remains relatively constant and stable following a signal by the Federal Reserve to slow credit volumes in the overall economy.

The combined findings appear to show that monetary policy effects cause shifts in bank lending from business investment toward (single-family and multi-family) residential mortgages, many of which eventually defaulted, and consumer credit card loans, which also experienced high default rates after the onset of the financial crisis. The results are consistent with a 25 basis point increase in the monetary policy rate acting as an amplification mechanism through the CRT securitization process, which increases systemic risk in the shadow banking system from apparent shifts of extended credit to households and away from businesses. This conclusion is justified by the fact that, prior to the recession, inflated home asset prices and understated credit quality ultimately lead to a severe economic downturn. Essentially, banks extend additional credit to the household sector during the upswing of the economy and lower credit that facilitates business investment, which is more supportive evidence of "bank bias" as described by Langfield and Pagano (2016).

Banks only continue to increase the issuance volumes for securitized mortgages in Figure S6 for the post-crisis period. However, the positive impact on securitized agency single-family mortgages is less than what is found in the pre-period. After the crisis, issuance volumes rise to a maximum of only 0.24%, and the increase is statistically significant for 8 months after a monetary tightening. The combined findings reveal that a contractionary monetary policy stance continues to affect banks through a shift in CRT now from business and non-mortgage consumer credit toward residential mortgages. Thus, it is informative to know that the conventional and unconventional monetary tools used in the post-crisis period do not eliminate "bank bias" completely.

Another aspect of the analysis shown in Figures S5 and S7 looks at how a monetary tightening affects volumes outstanding for CRT capital markets representing the net effect of liquidity. Bedendo and Bruno (2012) find that U.S. commercial banks' incentives to engage in CRT, as well as the impact of CRT on lending practices and risk, vary over the business cycle. During the 2007-2009 financial crisis, recourse to CRT capital markets becomes much more expensive or even unfeasible in segments of the shadow banking system in which asymmetric information issues are perceived to be severe because of uncertainty about the fair value of bank assets. The authors propose that if CRT helps mitigate the underinvestment problem in the business sector, then CRT issuance volumes should have a stronger impact on corporate loan growth during the crisis (see Stanton 1998). They conceptualize this idea in a context of a distressed financial sector during a recession. Banks may use the funding from CRT capital markets to either reconstitute liquidity in order to make additional business loans or reduce leverage by selling loans at a greater rate than originations. As such, the effect of a monetary tightening on net CRT volumes outstanding is considered below.

In Figure S5,1 find that securitized asset volumes outstanding increase following a 25 basis point increase in the monetary policy rate for all loan markets but those for business loans, (agency and private-label) multi-family mortgages, and private-label commercial mortgages. Thus, prior to the recession, consumer-related securitization instruments appear to buffer the Federal Reserve's intent to restrict net liquidity and credit to households. This is an interesting result given that the types of consumer debt vary substantially by maturity, collateral, interest rate structure, and investor base. Credit cards are short-term, variable rate unsecured loans, while auto loans are typically intermediate-term, fixed rate loans that are collateralized. Student loans have an average life of 10 years, can be fixed or variable rate, and are unsecured or backed with government protection against default. Finally, mortgages typically have the same life span as student loans, can be fixed or variable rate, are collateralized, and offer government guarantees.

A look at the responses for the post-crisis period in Figure S7 shows the positive net effect on liquidity following a monetary tightening among securitized agency mortgages only appears for those assets that are consolidated (i.e., held on- and off-balance sheet). In essence, the cumulative effects of higher rates and a reduction of MBS holdings by the Federal Reserve, which comprise the path toward monetary policy normalization, do not slow down CRT activity for securitized agency mortgages. For other non-agency CRT assets, it is interesting to find that securitized business loans, private-label mortgages, auto loans, and student loans all respond positively following a monetary tightening. From these results combined with the increase in leveraged secondary market syndicated loans discussed below, it appears that there is a deliberate move toward seemingly higher-yielding CRT instruments in the post-crisis period or to those markets with less regulation.

C. Traded Syndicated Loans and Credit Derivatives

Similar to securitized assets, traded syndicated loans are sold but the underlying collateral are loans to large corporate borrowers rather than consumer-related assets. Therefore, a changing interest rate and regulatory environment may affect traded loans differently. I find that all categories of secondary market syndicated loans in Figure S5 appear to fall in response to a contractionary monetary policy shock in the pre-crisis period. The decline in par traded loans is slightly longer in duration and larger in magnitude than that for distressed traded loans by about 0.02%. Reductions in volumes outstanding of traded loans is beneficial given that Dahiya, Puri, and Saunders (2003) find that over half of the firms whose loans are sold file for bankruptcy within 3 years of the initial sale of one of their loans. Moreover, Gorton and Pennacchi (1995) find that loan sales reduce banks' incentives to screen loan applicants and to monitor borrowers during the life of the loan because they do not have exposure to default risk after the loan sale.

The incentive for less screening and monitoring from loan sales is concerning because a small group of investors creates a traded security that is collateralized by numerous loans originated from different banks. The process makes it difficult for the security investor to trace the performance of a traded loan to the originator's screening abilities or link the security's performance with the monitoring activities of a new loan buyer. Essentially, banks are less attentive to hard information about corporate borrower's credit risk. Bord and Santos (2012) find that originating banks charge significantly higher interest rates on syndicated corporate loans that are eventually securitized when compared to those that are not. Apparently from the responses in Figure S7, investors seek highly leveraged traded loans whose risk levels present opportunities for higher yields.

In 2008 however, investment banks' failures such as that of Lehman Brothers coupled with AIG's credit default swap crisis steered the subprime mortgage conversation toward credit derivatives that transfer the risk of default on debt from financial institutions to a third party. Credit derivatives are supposed to provide a type of insurance against poorly underwritten or managed loans that can plague loan sales. Unlike loan sales or securitized instruments, credit derivatives indirectly provide liquidity to financial institutions through capital relief as discussed above and, therefore, should be impacted differently by a change in the monetary policy stance.

Although credit derivatives have value from a risk management perspective for an individual bank, in the event of an unexpected and adverse shock, they have the potential to destabilize the entire shadow banking system. Recall that AIG Financial Products division's participation and involvement in the credit default swap market contributed toward the recent recession by adding to system-wide instability (Eisenbeis 2009). (7) In fact, the division's inability to compensate banks for loan losses as a guarantor on credit derivatives contracts in addition to its deteriorating publicly traded REIT portfolio holdings made AIG subject to the same systemic risks to which banks were exposed. As a result, any negative shock that adversely affects credit derivatives volumes has negative externalities for the entire banking system.

The interconnected relationship between the banking industry and the credit derivatives market motivates the analysis of the monetary policy effects on volumes outstanding of credit derivatives, which should be an important indicator of liquidity in this market. Given that a rise in short-term interest rates affects spreads, the credit quality of the underlying asset, and the borrower's ability to repay debts, a monetary tightening should have a negative impact on credit derivatives usage unless capital market participants have limited attention in the opaque, pre-recession environment.

The results in Figure S5 support the limited attention hypothesis in the pre-crisis period. Net protection (the notional amount bought minus the amount sold) from credit derivatives and the notional amount reported as a beneficiary (buyer of protection) both remain relatively flat and statistically insignificant in response to a 25 basis point increase in the monetary policy rate. There does, however, appear to be a marginally significant increase in the total (cumulative) response of corporate bonds out 2 years following a monetary tightening that remains significant for months 30-34. Again, participants in the bond markets also appear to not place much emphasis on the monetary tightening signal by the Federal Reserve in that the net effects of liquidity from volumes outstanding increase slightly after a rise in short-term rates.

For the post-crisis period, Figure S7 does not show a decline of corporate bonds after a 25 basis point increase in the shadow policy rate but displays relatively maintained levels of corporate bonds for about 1 year after a 25 basis point increase in the shadow policy rate although the response is statistically insignificant. Banks, however, do seek protection on a net basis from findings of a positive and sustained total response of the net protection credit derivatives measure following a contractionary policy rate shock.

Combined with the response of corporate loans in the post-crisis period, there does not appear to be a shift between bank loans and bond financing for corporations, although the insignificant responses for both do tend to move in opposite directions in the post-crisis period for the first year following a negative shock. Yet maintained funding levels from bonds following a monetary tightening may give some support for existing evidence of the substitutability between loans and bonds during tighter overall credit conditions (see, e.g., Becker and Ivashina 2014). In the context of "bank bias," businesses substitute away from loans with bonds because negative policy rate shocks force banks to ration loans as lower asset values force banks to deleverage. Increased credit derivatives usage during this period of heightened awareness eases bank capital requirements and pressure on leverage. It is left up to the analysis below on relative CRT movements to conclusively suggest no substitutability between loans and bonds.

D. Relative CRT

Bertay, Gong, and Wagner (2017) find evidence of "bank bias" that is driven by the securitized CRT capital market. In order to measure the interaction between CRT capital markets and bank lending for each type of loan market, I define the relative CRT asset ratio as the volumes outstanding of the CRT instrument divided by the sum of the volumes outstanding of the CRT instrument plus the associated bank loans for the related market. For the relative credit derivatives ratio, bank loans are replaced by nonfinancial corporate bonds. Figures 1 and 2 below show responses of the constructed CRT asset ratios to a monetary tightening for the pre- and post-crisis periods, respectively.

The relative CRT ratio helps to determine moves toward CRT across capital markets following a contractionary monetary policy stance and therefore "bank bias." Within credit markets, the relative measure also helps to determine the importance of the respective CRT capital market and gives critical insight into the responses from policy rate changes across markets. In general, banks' efforts to borrow liquidity on capital markets is easier the more highly valued the banks are (Borio and Zhu 2012). The use of CRT capital market instruments under tighter credit conditions is partially explained by the fact that corporate debt has been found to be less cyclical than bank debt for firms (Langfield and Pagano 2016, and the references therein). Yet the important consideration in CRT transactions is investors' willingness to lend in particular credit markets that appears to drive bank lending and CRT activity in different sectors. Theoretically, a large group of underlying assets that can be easily sold and have readily traceable cash flows should encourage relative movements toward CRT capital market instruments away from bank debt within credit markets (Bertay, Gong, and Wagner 2017; Loutskina 2011). However, looking across estimation periods and after the Great Recession, the heightened scrutiny and regulatory changes in some markets such as mortgages and credit cards push borrowers active in CRT capital markets toward credit markets that are less regulated to satisfy investors' return objectives. Supportive findings for this line of thought are discussed below.

Both agency and private-label MBS and bank loans are subject to "bank bias" if the ratio for each type of CRT instrument increases after a monetary tightening relative to the ratios for other securitized consumer credit and business loans, traded syndicated loans, and credit derivatives. The findings in Figure 1 are not consistent with all CRT capital markets inducing "bank bias." The ratios for securitized total agency (private-label) mortgages actually decrease (increase) in the pre-crisis period. The maximum total decline of 0.29% from months 7 to 9 for securitized total agency mortgages indicates that the behavior in agency mortgages is not fully driven by volume increases in the securitized capital market before 2007. Apparently, prior to the recession, banks increase government guaranteed mortgages to both keep on their balance sheet and sell through their special purpose entities. In contrast, it appears that the CRT capital market drives securitized private-label mortgages as evidenced by the 0.04 maximum total increase from months 1 to 5. This positive increase is consistent with banks keeping their highest quality mortgages on the balance sheet and selling their lowest quality mortgages to willing investors in the private-label market, which end up to be poorly performing investments. CRT capital markets for relative consumer credit also tend to rise following a series of contractionary shocks in the pre-crisis period, and more liquidity for banks is directed toward more profitable consumer credit sectors.

This relative credit effect of a monetary tightening is not seen in the business sector, which does not experience substantial losses. Relative securitized business loans tend to fall in response to a 25 basis point increase in the monetary policy rate, but the finding is statistically insignificant. Moreover, relative traded syndicated loans and credit derivatives fall after a contractionary policy rate shock. Thus, it appears that investors in the securitized asset and syndicated loan markets are less likely to fund bank business loans than consumer loans prior to the recent recession, which ex post represents "bank bias."

After 2006, macroeconomic conditions dramatically change. The collapse of the real estate and other financial markets brought with it intensified regulation, heightened awareness by investors, and greater transparency of adversely affected capital markets. Based on this description of the 2007-2015 estimation period, it should be the case that relative securitized agency mortgages out of all mortgage debt, which measures the importance of the agency MBS capital market, should continue to fall in response to a monetary tightening. Yet Figures 1 and 2 show that, in fact, agency securitization remains just as important as mortgage debt in the 2007-2015 time period. The behavior here is justified when considering the massive amount of liquidity pumped into capital markets by the Federal Reserve through its lending facilities and asset purchases of agency debt and MBS. New mandates in the Dodd-Frank legislation to increase transparency and provide all investors with more granular information to determine CRT instrument values appear not to spill over to the private-label ABS capital market in that the move toward this market is essentially the same as before the recent financial crisis (0.020 pre-crisis vs. 0.026 post-crisis unaccumulated maximum percentage increase). For the (aggregate) securitized consumer credit capital market, the push for transparency by regulators and through accounting rule changes is even more pronounced. (8) Here, the relative importance of consumer ABS displayed in Figure 1 prior to the recent financial crisis is not present after 2006 in Figure 2.

Also in the post-crisis period, there appears to be growing importance of the business loan ABS capital market relative to bank business loans in this ratio's increase for about 6 months following a monetary tightening. Investors in this market are perhaps more aware of macroeconomic conditions that can drive issuance volumes in CRT capital markets, but banks accessing this market for their liquidity needs are still able to benefit from market opaqueness. Across securitized CRT markets, corporate loans are far more heterogeneous than agency mortgages, which have to meet common loan terms, and even than private-label mortgages. Underwriting standards and loan terms may vary significantly along with available historical information particularly for small business loans. Thus, differences and low information tends to drive investors toward the business loan ABS capital market in the post-crisis period when these investors seek high yields from a relatively low regulated CRT capital market.

Interestingly, it is these same opaque features of bank-syndicated corporate loans that continue to favor the importance of the primary market following a monetary contraction across estimation periods in Figures 1 and 2. Since a majority of traded loans are highly leveraged or considered of low quality and purchased by nonbanks, syndicate participants tend to keep in their portfolio the safer, less leveraged syndicated loans (Drucker and Puri 2009; Yago and McCarthy 2004).


In the previous section, the liquidity provided by CRT capital markets changes the relationship between monetary policy actions and bank lending to the household and business sectors. To varying degrees, the sensitivity of CRT capital market instruments to the monetary policy stance acts to reinforce the relationship. The interrelated loan and CRT capital markets are influenced by a tight monetary policy stance such that bank credit shifts among loan sectors creating uneven liquidity and lending activity throughout the financial system. More specifically, underlying bank loan markets show evidence of "bank bias" before and after the recent recession in that banks increase lending to consumers, while cutting back on lending to businesses. Securitized and traded instruments in CRT capital markets are the tools to facilitate bank leverage and loans to the consumer-related sectors. Therefore, following a monetary contraction, CRT capital markets continue to provide liquidity to some bank loan sectors throughout both estimation periods.

Relative credit tends to move toward private-label securitized mortgages away from bank mortgage debt in both estimation periods and toward securitized consumer credit before the recent recession. Although the movement toward private-label MBS continues in the post-crisis period, credit flows post-recession toward securitized business loans and credit derivatives away from corporate debt. Evidence of a decline in relative traded syndicated loans in both estimation periods empirically supports "bank bias." The noted shifts in relative credit post-crisis give insight into why CRT activity helps to insulate some loan markets from monetary policy transactions under vastly different economic conditions in that CRT capital market participants appear to transact in markets that are relatively less transparent and regulated.

The results provide investors and participants in the shadow banking system with a better understanding of how a monetary tightening creates movements in investments in underlying assets of both bank balance sheets and CRT capital markets for different lending sectors of the macro-economy. Unfettered lending combined with, in general, highly leveraged financial institutions creates excessive losses during economic downturns. The resulting systemic risk forces taxpayers to subsidize these institutions when they are near financial collapse as evidenced by the federal bailouts of AIG, General Motors,

Chrysler, and American Express to name a few. (9)

More recently, the Federal Reserve Bank of New York Consumer Panel Survey reports that consumer debt is at an all-time high surpassing levels in 2008 before the subprime mortgage crisis spread to other consumer-related credit markets. A large part of this restored growth is due to student and auto loans. Among the current top 20 auto lenders, there is a low presence of regulated banks in which attractive yields to investors have pushed some securitizers in CRT capital markets into the subprime submarket where delinquencies and defaults tend to be higher. In fact, delinquent payments have reached levels not seen since 1996 according to several industry news accounts. Although the size of this market relative to other CRT capital markets is small, market participants should worry that these levels of stress among the underlying assets will only grow under tighter overall credit conditions.

The important policy implication is that the potential for financial stability vulnerabilities from a renewed interest in CRT instruments should be a concern of U.S. policymakers. Langfield and Pagano's 2016 recommendation to limit the aggregate growth of large banking institutions in predominantly bank-based financial systems is needed for the U.S. financial system but in combination with an intensified look at transactions in CRT capital markets given the possible instability created from "bank bias" and a healthy share of market-based intermediation. The different behaviors of CRT instruments and bank loans implied by regulatory changes among relative CRT capital markets after the recent recession offer a path for policymakers to reduce vulnerabilities. Moreover, how these changes in the lending intermediation process and monetary policy actions affect available loans to households and businesses is a unique contribution to the literature.


The great expansion of financial markets in recent decades on both the demand and the supply sides has renewed interest concerning the effectiveness of monetary policy actions to influence household and business spending and macroeconomic activity. A crucial element of this discussion is the liquidity generated in CRT capital markets that affects bank lending.

This study's comprehensive approach to directly test monetary policy effects on both bank lending and CRT capital market transaction volumes contributes to the literature on banks' loan portfolio management across different credit markets. If one believes that CRT activity can add to a systemically fragile financial system through instruments that facilitate increased leverage, greater liquidity, and more severe information problems, then the potential for economic vulnerability is disquieting because equivocal rule-making and timid enforcement by regulators may impair the Federal Reserve's ability to prevent future crises. The Great Recession may well repeat itself with the result that adequate protections against systemic risk will not be agreed upon and implemented in the future by policymakers.


Adrian, T., and H. S. Shin. "Financial Intermediaries and Monetary Economics," in Handbook of Monetary Economics, Vol. 3, Chapter 12, edited by B. M. Friedman and M Woodford. New York: Elsevier B.V, 2010. 601-50.

Altunbas, Y., L. Gambacorta, and D. Marques-Ibanez. "Securitisation and the Bank Lending Channel." European Economic Review, 53(8), 2009, 996-1009.

Ball, L. "Ben Bernanke and the Zero Bound." Contemporary Economic Policy, 34(1), 2015, 7-20.

Becker, B., and V. Ivashina. "Cyclically of Credit Supply: Firm Level Evidence." Journal of Monetary Economics, 62, 2014, 76-93.

Bedendo, M., and B. Bruno. "Credit Risk Transfer in U.S. Commercial Banks: What Changed during the 2007-2009 Crisis?" Journal of Banking & Finance, 36(12),2012,3260-70.

Belviso, F, and F. Milani. "Structural Factor-Augmented VARs (SFAVARs) and the Effects of Monetary Policy." Topics in Macroeconomics, 6(3), 2006, Article 2.

Bernanke. B. S. "The Financial Accelerator and the Credit Channel." Speech delivered at The Credit Channel of Monetary Policy in the Twenty-first Century Conference, Federal Reserve Bank of Atlanta, Atlanta. Georgia. June 15,2007.

Bernanke, B. S., J. Boivin, and P. Eliasz. "Measuring the Effects of Monetary Policy: A Factor-Augmented Vector Autoregressive (FAVAR) Approach." Quarterly Journal of Economics, 120(1), 2005, 387-422.

Bernanke, B. S.. M. Gertler, and S. Gilchrist. "The Financial Accelerator in a Quantitative Business Cycle Framework." in Handbook of Macroeconomics, Vol. 1, Part 6, Chapter 21, edited by J. B. Taylor and M. Woodford. Amsterdam: Elsevier North-Holland, 1999, 1341-93.

Bertay, A. C. D. Gong, and W. Wagner. "Securitization and Economic Activity: The Credit Composition Channel." Journal of Financial Stability, 28, 2017, 225-39.

Bord, V. M., and J. Santos. "The Rise of the Originate-to-Distribute Model and the Role of Banks in Financial Intermediation." Federal Reserve Bank of New York Economic Policy Review, 18(2), 2012, 21-34.

Borio, C, and H. Zhu. "Capital Regulation, Risk-Taking and Monetary Policy: A Missing Link in the Transmission Mechanism?" Journal of Financial Stability, 8(4), 2012, 236-51.

Brunnermeier, M. K., and Y. Sannikov. "A Macroeconomic Model with a Financial Sector." American Economic-Review, 104(2), 2014, 379-421.

Calluzzo, P., and G. N. Dong. "Has the Financial System Become Safer after the Crisis? The Changing Nature of Financial Institution Risk." Journal of Banking & Finance, 53, 2015, 233-48.

Cantor, R., and S. Rouyer. "Another Perspective on Credit Risk Transfer and Asset Securitization." Journal of Risk Finance, 1(2), 2000, 37-47.

Dahiya, S., M. Puri, and A. Saunders. "Bank Borrowers and Loan Sales: New Evidence on the Uniqueness of Bank Loans." Journal of Business, 76(4), 2003, 563-82.

Diamond, D. W., and P. H. Dybvig. "Bank Runs, Deposit Insurance, and Liquidity." Journal of Political Economy, 91(3), 1983,401-19.

Disyatat. P. "The Bank Lending Channel Revisited." Journal of Money, Credit and Banking, 43(4), 2011, 711 -34.

Drucker, S., and M. Puri. "On Loan Sales. Loan Contracting, and Lending Relationships." Review of Financial Studies, 22(7), 2009, 2835-72.

Eisenbeis, R. A. "Multinational Banking and Regulatory Challenges: Lessons from the US Experience with AIG." CESifo DICE Report, 7(2), 2009, 13-8.

Faia, E. "Credit Risk Transfers and the Macroeconomy." Working Paper Series No. 1256, European Central Bank. 2010.

Financial Stability Board. "Global Shadow Banking Monitoring Report 2015." 2015.

Gambacorta, L., and D. Marques-Ibanez. "The Bank Lending Channel: Lessons from the Crisis." Economic Policy, 26(66), 2011, 135-82.

Gorman, R. "How Student-Loan Debt Is Dragging Down the Economy." Business Insider, 2015.

Gorton, G., and A. Metrick. "Securitized Banking and the Run on the Repo." Journal of Financial Economics, 104(3), 2012,425-51.

Gorton, G. B., and G. G. Pennacchi. "Bank and Loan Sales: Marketing Nonmarketable Assets." Journal of Monetary Economics, 35(3), 1995, 389-411.

He, Z., and W. Xiong. "Dynamic Debt Runs." Review of Financial Studies, 25(6), 2012, 1799-843.

Jimenez, G., S. Ongena, J.-L. Peydro, and J. Saurina. "Credit Supply and Monetary Policy: Identifying the Bank Balance-Sheet Channel with Loan Applications." American Economic Review, 102(5), 2012, 2301-26.

Kashyap, A. K.. and J. C. Stein. "The Impact of Monetary Policy on Bank Balance Sheets." Carnegie-Rochester Conference Series on Public Policy, 42, 1995, 151-95.

Keys, B. J., T. Mukherjee, A. Seru. and V. Vig. "Did Securitization Lead to Lax Screening? Evidence from Subprime Loans." Quarterly Journal of Economics, 125(1), 2010. 307-62.

Kishor, N. K., and M. Newiak. "The Instability in the Monetary Policy Reaction Function and the Estimation of Monetary Policy Shocks." Contemporary Economic-Policy, 32(2), 2013, 390-402.

Kuttner, K. "Securitization and Monetary Policy." Mimeo, Federal Reserve Bank of New York, 2000.

Langlield. S., and M. Pagano. "Bank Bias in Europe: Effects on Systemic Risk and Growth." Economic Policy, 31(85), 2016,51-106.

Loutskina, E. "The Role of Securitization in Bank Liquidity and Funding Management." Journal of Financial Economics. 100(3), 2011, 663-84.

Maddaloni, A., and J.-L. Peydro. "Bank Risk-Taking, Securitization. Supervision, and Low Interest Rates: Evidence from the Euro-Area and the U.S. Lending Standards." Review of Financial Studies, 24(6), 2011, 2121-65.

Minton, B. A., R. Stulz, and R. Williamson. "How Much Do Banks Use Credit Derivatives to Hedge Risk?" Journal of Financial Services Research. 35(1), 2009, 1-31.

Pagano, M.. and P. Volpin. "Securitization, Transparency, and Liquidity." Review of Financial Studies, 25(8), 2012, 2417-53.

Prater, C. "Will Credit Card Securities Be the Next Financial Mess?", 2008.

Rajan, R. G. "Has Finance Made the World Riskier?" European Financial Management, 12(4), 2006, 499-533.

Sellon, G. H., Jr. "The Changing U.S. Financial System: Some Implications for the Monetary Transmission Mechanism." Federal Reserve Bank of Kansas City Economic Review, 87(1), 2002, 5-35.

Singh, A., S. Stone, and J. Suda. "Monetary Policy and the Financial Sector." Economics Letters, 132, 2015, 82-86.

Stanton, S. W. "The Underinvestment Problem and Patterns in Bank Lending." Journal of Financial Intermediation, 7(3), 1998.293-326.

Stein, J. C. "Securitization, Shadow Banking & Financial Fragility." Daedalus, 139(4), 2010, 41-51.

Summers, L. "Why the Fed Must Stand Still on Rates." Larry Summers, 2015.

Tankersley, J. "Federal Reserve Raises Interest Rates for Second Time in a Decade." Washington Post, 2016. a6b 15b47.

Wilcox, J. A. "Securitization and Small Business." Federal Reserve Bank of San Francisco Economic Letter No. 2011-22,2011.

Yago, G.. and D. McCarthy. "The U.S. Leverage Loan Market: A Primer." Milken Institute Research Report, 2004.

Young, A. T, and J. Bologna. "Crises and Government: Some Empirical Evidence." Contemporary Economic Policy, 34(2), 2015, 234-49.


Additional supporting information may be found online in the Supporting Information section at the end of the article.

Appendix S1. Extended data and empirical methodology


(*) Thanks to Jocelyn Evans along with participants at the 2015 IBEFA/WEAI Summer Meeting and 2016 Academy of Economics and Finance Annual Meeting for their helpful comments and suggestions. All errors and omissions are my sole responsibility.

Robertson: Assistant Professor, Department of Economics, Carl H. Lindner College of Business, University of Cincinnati, Cincinnati, OH 45221. Phone 1 513 556 2614, E-mail


ABS: Asset-Backed Securities

AIG: American International Group

CRT: Credit Risk Transfer

FAVAR: Factor-Augmented Vector Autoregression

FRED: Federal Reserve Economics Database

MBS: Mortgage-Backed Securities

OTS: Office of Thrift Supervision

REIT: Real Estate Investment Trust

SIFMA: Securities Industry and Financial Markets Association

VAR: Vector Autoregression


(2.) Market-based residential mortgages and consumer credit card loans also represent approximately half of all (conventional and shadow bank) lending due to their long history of securitized CRT capital market transactions.

(3.) Several empirical studies combine mortgage real estate investment trusts (REITs) with securitized private-label mortgages as a majority of mortgage REITs invest in MBS. Since data on both series are available for issuance and volumes outstanding examined in this study, 1 separate the two data series.

(4.) AIG also experienced severe losses on securities lending. Rather than investing received cash for lent out securities into safe, low-yielding assets, AIG invested funds into risky, subprime securitized mortgages. (For further discussion, see

(5.) A graphical representation of the tested hypotheses is presented in Figure S3 of Appendix S1, Supporting information.

(6.) Due to space limitations, figures for the impulse response functions in Sections VI.A to VI.C are presented in Section A3 of Appendix S1. (Also see footnote 5.) Figures in Section VI.D are presented here.

(7.) As described by Eisenbeis (2009). AIG became a Savings & Loan Holding Company in 1999 subject to federal oversight by the Office of Thrift Supervision (OTS). Therefore, at the time of the 2007-2009 financial crisis, AIG and its Financial Products subsidiary were under the consolidated supervision of the OTS despite the fact that capital market activities by the Financial Products division were technically unregulated.

(8.) Basel III's changes to capital charges finalized in September 2010 have an implementation phase from January 2013 to 2019. Also, in January 2010, the Financial Accounting Standards Board changed off-balance-sheet treatment for some securitization transactions (including credit cards) with the intent to reduce credit risk and improve transparency for investors. FAS 166 and 167 require issuers to consolidate the issuing entity if they have the power to direct activities that impact economic performance, are obligated to absorb credit losses, or receive benefits.


doi: 10.1111/coep.12264
Credit Risk Transfer Assets (Amounts Outstanding, Billions of USD)

Instrument                         1995     1999      2001     2006
Securitized assets
 Agency mortgage                 1,570.3  2.292.2   2,830.13 3,837.9
 Agency mortgage (consolidated)  1,570.3  2,292.2   2,830.13 3,837.9
 Private-label mortgage            316.5    882.6   1,085.7  3,284.8
 Consumer credit                   188.5    364.3     484.4    687.2
 Business loan                      88.0    198.8     278.5    447.4
Mortgage REITs                       6.8     39.2      46.2    283.0
Traded syndicated loans:
 Par                                25.6     70.2      75.8    198.7
 Distressed                          8.2      8.9      41.7     39.9
Credit derivatives
 Beneficiary and guarantor             -    341.7     510.1  9,264.1
 Beneficiary                           -    189.2     262.8  4,670.7

Instrument                          2007      2009      2011     2015
 Securitized assets
 Agency mortgage                  4,459.9   5,372.2   1,310.2  1,773.1
 Agency mortgage (consolidated)   5,801.0   6,636.4   6,947.7  7,304.5
 Private-label mortgage           3,571.6   2,705.2   2,096.1  1,486.0
 Consumer credit                    735.6     675.1     521.3    536.5
 Business loan                      549.6     399.0     243.3    289.2
Mortgage REITs                      255.3     167.7     331.1    495.2
Traded syndicated loans:
 Par                                310.2      81.5      73.7    132.6
 Distressed                          31.8      35.3       6.9     10.1
Credit derivatives
 Beneficiary and guarantor       13,153.0  20,715.8  14,759.2  6,986.5
 Beneficiary                      9,685.0  10,493.4   7,461.7  3,545.7

Notes: Securitized asset data come from the Federal Reserve Flow of
Funds and Securities Industry and Financial Markets Association
(SIFMA). Consolidated agency mortgages result from accounting rule
changes in the first quarter of 2010 that require issuers to
consolidate certain securitized assets back on the balance sheet of
the issuers' parent financial institution. Securitized consumer credit
sums vehicle, credit card, and student loans; securitized business
loans include franchise and Small Business Administration loans.
Mortgage real estate investment trust (REIT) data are from the Federal
Reserve Flow of Funds and report total assets (cash, securities,
loans, and other financial assets) of which the majority are
mortgage-backed securities. Thomson Reuters Loan Pricing Corporation
provides traded syndicated corporate loan volumes. Par loans
(historically) trade at or above par value, and distressed loans trade
at a discount from potential risk of less than full payment. Credit
derivatives data come from the Federal Financial Institutions
Examination Council, Report of Condition and Income (Call Reports).
Beneficiaries transfer the credit risk of an asset to guarantors.
COPYRIGHT 2019 Blackwell Publishers Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Robertson, Mari L.
Publication:Contemporary Economic Policy
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2019

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters