Risk Taking and Low Longer-term Interest Rates: Evidence from the U.S. Syndicated Loan Market

Finance and Economics Discussion Series Divisions of Research & Statistics and Monetary Affairs Federal Reserve Board, Washington, D.C. Risk Taking and Low Longer-term Interest Rates: Evidence from the U.S. Syndicated Loan Market Sirio Aramonte, Seung Jung Lee, and Viktors Stebunovs 2015-068 Please cite this paper as: Aramonte, Sirio, Seung Jung Lee, and Viktors Stebunovs (2015). “Risk Taking and Low Longer-term Interest Rates: Evidence from the U.S. Syndicated Loan Market,” Finance and Economics Discussion Series 2015-068. Washington: Board of Governors of the Federal Reserve System, http://dx.doi.org/10.17016/FEDS.2015.068. NOTE: Staff working papers in the Finance and Economics Discussion Series (FEDS) are preliminary materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. References in publications to the Finance and Economics Discussion Series (other than acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers.

Risk Taking and Low Longer-term Interest Rates: Evidence from the U.S. Syndicated Loan Market Sirio Aramonte, Seung Jung Lee, and Viktors Stebunovs∗ July 2015 Abstract We use supervisory data to investigate risk taking in the U.S. syndicated loan market at a time when longer-term interest rates are exceptionally low, and we study the ex-ante credit risk of loans acquired by different types of lenders, including banks and shadow banks. We find that insurance companies, pension funds, and, in particular, structured-finance vehicles take higher credit risk when investors expect interest rates to remain low. Banks originate riskier loans thattheytendtodivestshortlyafterorigination, thusappearingtoaccommodateotherlenders’ investment choices. These results are consistent with a “search for yield” by certain types of shadow banks and, to the extent that Federal Reserve policies affected longer-term rates, the resultsarealsoconsistentwiththepresenceofarisk-takingchannelofmonetarypolicy. Finally, we find that longer-term interest rates have only a modest effect on loan spreads. JEL classification: E43, E44, E52, E58, G11, G20. Keywords: Syndicated loans; Shared National Credit Program; Shadow banking; Zero lower bound; Search for yield; Risk-taking channel of monetary policy. ∗Board of Governors of the Federal Reserve System, 20th and C Streets NW, Washington, DC 20551. Contact information: sirio.aramonte@frb.gov, seung.j.lee@frb.gov, viktors.stebunovs@frb.gov (contact author). We thank Robert Cote for valuable comments and guidance with the Shared National Credit data. We thank William Bassett, Mark Carey, Stijn Claessens, Francisco Covas, Joa˜o Santos, Gretchen Weinbach, and Egon Zakrajˇsek, and other seminar participants at the Federal Reserve Board, the System Committee Meeting on Financial Structure and Regulation, the Basel Committee on Bank Supervision Research Task Force meeting “Systemically important financial institutions: a research agenda”, the 2014 Western Economic Association International Meeting, the 2014 European Finance Association Conference, the 2014 Northern Finance Association Conference, the 14th FDIC Annual Bank Research Conference, the 2014 Financial Management Association Annual Meeting, the 2014 Southern FinanceAssociationAnnualMeeting,theFederalReserveSystem“Day-Ahead”ConferenceonFinancialMarketsand Institutions, and the 2015 Financial Intermediation Research Society Conference. We are grateful to our discussants Alyssa Anderson, Burcu Duygan-Bump, Thomas Gilbert, Juanita Gonzalez-Uribe, Rainer Jankowitsch, Greg Nini, Lars Norden, Mitchell Petersen, Gregory Sutton, and David Vera for helpful comments and suggestions. We thank AmandaNg,GregCohen,andChristopherCorderoforexcellentresearchassistance. Thispaperreflectstheviewsof theauthors,andshouldnotbeinterpretedasreflectingtheviewsoftheFederalReserveSystemorothermembersof its staff. 1

1. Introduction In this paper, we study risk taking in the U.S. syndicated loan market in the aftermath of the 2009 financial crisis. We ask whether risk taking changes as longer-term interest rates decline, whether risk-taking patterns vary across different lender types, and whether the same risk-taking patterns can be found in the primary and secondary markets. Our questions are related to the literature on “search for yield” and to the possible existence of a risk-taking channel of unconventional monetary policy. While, as discussed below, increased risk taking can raise financial stability concerns, accommodative monetary policy can help “prompt a return to the productive risk taking that is essential to robust growth.”1 In this regard, syndicated loans are a suitable asset class to study because they provide a large amount of credit to the productive sector. Weanalyzerecentrisk-takingtrendsinthe$900billionmarketforU.S.syndicatedtermloans using confidential supervisory data available at a quarterly frequency since the end of 2009. The Shared National Credits Program (SNC) covers syndicated loans amounting to at least $20 million and in which three or more federally supervised banks participate as lenders. The database reports all lenders and their syndicate shares, even if they are not supervised banks. Given that nonbank lenders play a significant role in syndicated term loans (Ivashina and Sun, 2011), we can analyze the risk-taking behavior of a rich cross section of intermediaries with distinct business models and subject to different regulatory environments. We find that a number of nonbank financial institutions—like insurance companies, pension funds, and, in particular, collateralized loan/debt obligations (CLOs/CDOs)—increase the credit risk of their syndicated-loan investments when longer-term interest rates are low. CLOs and CDOs 1 The quote is from Chairman Ben S. Bernanke’s speech “Long-Term Interest Rates” at the Annual Monetary/MacroeconomicsConference: ThePastandFutureofMonetaryPolicy,sponsoredbyFederalReserve Bank of San Francisco, San Francisco, California, March 1, 2013. 2

are structured-finance vehicles that purchase a pool of fixed-income assets like loans or bonds and issue notes of different seniority backed by these assets. Banks originate riskier loans that they tend todivestafterorigination, apparentlyaccommodatingotherlenders’investmentchoices. Giventhat banks have a competitive advantage in screening and monitoring borrowers (Gorton and Pennacchi, 1995), they are well-suited to investing in higher-risk loans in times of economic uncertainty, when interest rates are likely to be low. However, Maddaloni and Peydr´o (2011) find a weaker relation between short-term rates and lending standards when supervisory standards are stronger, raising the possibility that the intense regulatory activity following the 2008 financial crisis may have counterbalanced any incentives that banks may have had to hold riskier assets. Our conclusions are robust to a number of business-cycle controls, to different proxies for interest rate expectations, to using Treasury rates already orthogonalized relative to the control variables, and to specifications that reduce the influence of a potentially omitted economy-wide credit risk factor. The results are consistent with a search for yield by nonbank intermediaries and with the existence of a risk-taking channel of monetary policy during a period when the Federal Reserve engaged in unconventional monetary policy initiatives to put downward pressure on longer-run interest rates, with policies like forward guidance and large scale asset purchases programs (LSAPs) (D’Amico, English, Lopez-Salido, and Nelson, 2012; Krishnamurthy and Vissing-Jorgensen, 2013). Studies such as Maddaloni and Peydr´o (2011); Paligorova and Santos (2013); Dell’Ariccia, Laeven,andSuarez(2014); Ioannidou,Ongena,andPeydr´o(2015); andAltunbas,Gambacorta,and Marques-Ibanez(forthcoming)findevidenceofarisk-takingchannelthatassociatesaccommodative monetary policy, measured by short-term rates, to the origination of riskier loans by banks. The effect is stronger in the case of smaller banks that are not part of a large corporate group with deep internal capital markets (Buch, Eickmeier, and Prieto, 2014; Jimenez, Ongena, Peydr´o, and 3

Saurina, 2014; and Campello, 2002).2 Our analysis leverages the SNC data to contribute to the literature in several ways. First, we can track activity in the secondary as well as the primary syndicated loan markets, which is important because the effect of low interest rates on risk taking in the primary market may be dampened by the attempt by certain intermediaries to cater to existing lending relationships (see, for instance, Degryse and Ongena, 2007).3 Second, we are able to measure the ex-ante credit risk of each borrower using the default probability that the banks coordinating the syndicates use to determine regulatory capital. Regulations require banks to use default probabilities that provide a long-run assessment of a loan’s credit risk, which assuages concerns about the endogeneity of U.S. interest rates and default risk, because a long-run default probability should be less sensitive to contemporaneous interest rate shocks.4 Third, our analysis is novel because, while other researchers have studied particular types of intermediaries, we compare the behavior of a diverse set of lenders who all operate in the syndicated loan market but face different incentives when adapting to an environment of persistently low interest rates. Generally, various types of lenders have an incentive to rebalance their portfolios investing in riskier assets when returns on safer assets decline. Indeed, an objective of nontraditional monetary policy was to promote a return to productive risk taking. Certain types of lenders may have 2 In addition, Chodorow-Reich (2014) finds that money market funds and some defined-benefits pension funds engaged in a search for yield between 2009 and 2011. Di Maggio and Kacperczyk (2015) also show that money marketfunds,especiallythosenotaffiliatedwithotherlargefinancialintermediaries,tookmoreriskafterpolicies meant to reduce interest rates were implemented. A related literature studies the effect of more specific policy interventions, like the Troubled Asset Relief Program. See, for instance, Black and Hazelwood (2013), Duchin and Sosyura (2012), and Li (2013). 3 Jones,Lang,andNigro(2005)documentthedeterminantsoftheproportionofaSNCloanretainedbyanagent bank over time. 4 Fordetails,seethe“Risk-BasedCapitalStandard: AdvancedCapitalAdequacyFramework-BaselII”(Federal Register Vol.72, No.235, December 7, 2007), which defines the probability of default for a wholesale (non-retail) obligor as follows: “For a wholesale exposure to a nondefaulted obligor, the [bank]s empirically based best estimateofthelong-runaverageone-yeardefaultratefortheratinggradeassignedbythe[bank]totheobligor, capturing the average default experience for obligors in the rating grade over a mix of economic conditions (including economic downturn conditions) sufficient to provide a reasonable estimate of the average one-year default rate over the economic cycle for the rating grade.” 4

characteristicsthatstrenghtenthisincentive. Forinstance, severalstudiesshowthatfundmanagers generallyhaveanincentivetoincreaserisk-takinginordertoout-ranktheirpeers, whichistypically attributed to a rapid increase in monetary or reputational benefits as performance relative to their peers improves.5 Similar incentives may apply to the managers ofstructured finance vehicles, which areincludedinouranalysis. Forotherlendertypes, theincentivetoinvestinriskierassetscancome from the structure of their balance sheet. Becker and Ivashina (forthcoming) find that insurance companies with binding capital ratios are more likely to engage in a search for yield. Our results also suggest that finance companies engage in a search for yield. Finance companies increased their funding through (normally fixed-rate) bonds and intermediate notes substantially in the last quarter of 2010, after the 10-year Treasury rate dropped significantly through 2010, suggesting that finance companies may have tried to take advantage of then-historically low rates by issuing longer-term debt. However, the 10-year Treasury rate decreased further into 2012. This drop may have generated a gap between asset yields and interests on liabilities, because the assets held by finance companies can often be prepaid and borrowers have a stronger incentive to refinance after a significant decline in interest rates. As a consequence, finance companies may have attempted to reduce the gap by increasing the credit risk of newly acquired assets.6 The discussion so far has focused on the credit risk of syndicated loans. As noted by Ivashina and Sun (2011), strong investor demand throughout the 2000s resulted in a compression of 5 Early studies attributed the incentive to out-rank peers to the convex relation between fund inflows and past performance, which implies that a fund would increase assets under management (AUM) if the higher risk translated into positive returns, but would not lose much AUM if the higher risk led to negative returns (see, for instance, Chevalier and Ellison, 1997). Spiegel and Zhang (2013) show that this convexity is an artifact of omitted heterogeneity in fund characteristics and suggest that the incentive to out-rank peers can originate from managerial career concerns, like termination risk or compensation. Qiu (2003) shows that the incentive is stronger for funds with performance just below the median, and for those trailing the top performers. Kempf, Ruenzi,andThiele(2009)discussaricherframeworkinwhichtheeffectsofterminationriskdependonthestate of the economy. 6 DatafromtheFederalReserve’sG20statisticalreleaseshowthatfundingthroughbondsandintermediatenotes at the end of 2010 increased to 59% from 45% in the previous quarter. The increase in the share of bonds and intermediate notes at the end of 2010 was nearly three times as large as the second-largest quarterly increase between 1995 and 2010. The 10-year Treasury rate decreased from 4% early in 2010 to 2.5% mid-year before bouncingbackto3.5%bytheendoftheyear. The10-yearTreasuryratefellfurtherthrough2011toreach1.5% in 2012. Consumer loans held by finance companies were 45% of assets at the end of 2010. 5

syndicated loan spreads, and a natural question is whether the risk-taking patterns we find are also reflected in loan spreads. Pricing information is not available in the SNC data, but we are able to match about one third of the syndicated loan originations for which default risk data are available in SNC with spreads from Thomson Reuters Loan Pricing Corporation’s DealScan. Controlling for nonbank loan share, we find that lower Treasury rates result in marginally higher spreads, which is not consistent with the hypothesis that demand from a search for yield leads to spread compression in the primary syndicated term-loan market. Our results are subject to several caveats. First, they are not necessarily representative of overall risk taking by the intermediaries we consider, because we study only a portion of their portfolios and the increased risk could be hedged with more stringent covenants or by trading other financial instruments. However, our conclusions are not driven by within-group risk transfers that leave the exposure of the parent company to syndicated loans unchanged, because we consolidate activity in the syndicated loan market at the parent-company level. Second, we do not observe the individuals or institutions who invest in CLOs/CDOs, hence we do not know where the credit risk taken by these lenders ultimately resides. For instance, a pension fund may be exposed to syndicated-loan risk through intermediate investment funds even if it does not participate in this market directly. While excessive risk taking can facilitate the build-up of imbalances that set the stage for future financial distress (Borio and Zhu, 2012), we should also emphasize that any increase in risk taking attributable to monetary policy must be evaluated against the benefits of an accommodative monetarypolicy. Ingeneral,ahealthiereconomyimplieslowercreditrisk. Inaddition,theliterature on the syndicated loan market has highlighted the fact that loan supply is adversely affected by negative liquidity and capital shocks to lenders, which accommodative monetary policy can help alleviate. Ivashina and Scharfstein (2010), for example, find that banks with more liquidity 6

problems—those with larger potential drawdowns and those with less access to deposit financing and more reliance on short-term debt—cut lending to large borrowers more significantly during the 2008 financial crisis. Interest rates on syndicated loans also increased in proportion with the losses thatbanksexperiencedfromsubprimeloans, asdiscussedinSantos(2011). Chodorow-Reich(2014) also finds broad benefits of monetary policy for banks and life insurance companies in the aftermath of the 2008 financial crisis. 2. Shared National Credits Program Data The Shared National Credits Program was established in 1977 by the Board of Governors of the Federal Reserve System, the Federal Deposit Insurance Corporation, and the Office of the Comptroller of the Currency to provide an efficient and consistent review of large syndicated loans. Before 1999, information was gathered for loans with a committed or disbursed amount of at least $20 million shared by two or more unaffiliated supervised institutions. Currently, the program covers any loan in excess of $20 million that is shared by three or more supervised institutions. Bank regulators review a SNC loan based on information provided by a designated bank—usually an agent bank. One or more agent banks are generally responsible for recruiting a sufficient number of loan participants, negotiating the contractual details, preparing adequate loan documentation, and disseminating financial documents to potential participants. Once the loan is made, agent banks are also responsible for loan servicing, usually for a fee. While bank regulations require participants to assess a borrower’s credit risk independently, syndicate members typically provide an assessment similar to that of agent banks. The SNC program offers two data outputs: one at an annual frequency, that has been covered widely in the literature, and one at a quarterly frequency, that has become available only recently 7

and offers more loan-specific information which is not available from the first output. While we rely only on the quarterly output, we find it instructive to describe both in some detail. Annual SNC reviews are conducted each May using data provided by agent banks, typically as of December 31 of the prior year, and sometimes as of March 31 of the review year. SNC program examiners assign credit ratings to these loans (in descending order: pass, special mention, and classified), and further characterize loans with a classified rating into three sub-categories: substandard, doubtful, and loss. The SNC program publishes review summaries every year, and the results of the 2013 SNC reviewwerepubliclyreleasedonOctober 10, 2013.7 The2013 SNCdatabase covered approximately 9,300 syndicated loans to 5,800 borrowers, for a total of $3 trillion in drawn credit and unused commitments (for a given loan, commitment is the maximum amount of credit lenders agree to provide; throughout the paper we refer to drawn credit as loan “utilization”). Figure 1 shows the evolution of loan commitments and loan utilizations over time. Revolving credits are the bulk of commitments, while term loans are the bulk of actual utilizations. Beginning in the fourth quarter of 2009, federal regulators began collecting syndicated loan data on a quarterly basis from the 18 banks with the most active syndicated loan businesses, which account for about 90% of the market. These quarterly reporters also provide a detailed assessment of each loan’s credit risk through the Basel II parameters used to calculate regulatory capital, such as the probability of default (PD), loss given default, and exposure at default. In our analysis, we use the quarterly SNC data over the sample period 2010:Q1–2013:Q4 because the calculation of our main dependent variable requires lagged holdings. The reported PDs are estimated in compliance with the Basel II Advanced Internal Ratings-Based requirements. For a non-defaulted obligor, the PD is the bank’s estimate of the 7 The results are available at www.federalreserve.gov/newsevents/press/bcreg/20131010a.htm. 8

long-run average one-year default rate for the rating grade that the bank assigns to the obligor, capturing the average default experience for obligors in the rating grade over a mix of economic conditions, including downturns. For a defaulted obligor, the PD is equal to 100%. In terms of PD comparability across banks, banks calculate PDs independently, but they all need to comply with the provisions of Basel II. Figure 2 compares commitments and utilizations in the annual SNC data with commitments and utilizations in the quarterly SNC data; it shows that the quarterly data are only slightly less comprehensive than the annual equivalent. Requiring the availability of PDs reduces the sample by about30%intermsofloancommitments, andbyabout50%intermsofutilizations. Thereasonfor the significant amount of loans with missing PDs is that only banks in the early stages of adopting Basel II regulations must report the Basel II parameters, while other banks simply have the option to report. Once banks begin providing PDs for a given loan, they must continue doing so. We apply several filters to the data in order to minimize the impact of recording errors. Some banks appear to have reported PDs of zero for loans for which they did not have PD values, and we set zero PDs to “missing” unless we are able to match them with an expected default frequency (EDF)fromMoody’sthatislowerthan50basispoints. Somebanksalsoappeartohaveerroneously reported PDs of 100% for certain loans. We replace a 100% PD with a “missing” value if the loan is rated “pass,” has no charge-off associated with it, is not past due, and did not have a legitimate PD of 100% in the prior quarter. If leads and lags of a missing PD differ by only 1 basis point, we fill in the missing PD value with the average of its lead and lag. We do so also when two consecutive values are missing and the neighboring non-missing PDs are at most 1 basis point apart. For some loans, PDs in a given quarter are materially different from the PDs in the previous and subsequent quarters. If PDs in the previous and subsequent quarters differ by only 1 basis point and if the current reported PD is materially different from the previous and subsequent PDs (either greater 9

than 5 times or less than 1/5 their average), we replace those PDs with the average of the previous and subsequent PDs. Finally, a small number of loans have no PDs but do have information on expectedcreditloss(ECL),lossgivendefault(LGD),andexposureatdefault(EAD).Inthesecases, we calculate PDs according to the following formula: PD = ECL/(EAD×LGD). We should note that our conclusions carry through in the absence of these filters. Table 1 shows summary statistics for the default risk of term loans according to rating grades and lender types. About 80% of loans are classified as “pass,” with a median PD of 78 basis points, while about 3.5% of the loans receive the two lowest grades, whose median PD is 100%. In order to reduce the impact of loans with high default probabilities on the estimation, we cap PDs at 35%, which is the largest value that Moody’s assigns to EDFs (the statistical and economic significance of the results is slightly stronger without capping PDs). The rightmost column of Table 1 shows that this filter affects the median PD of only the riskiest loan categories, which is now 35% rather than 100%. ThebottompanelofTable1reportstheaverageloanshareofdifferentlendertypes,wherethe average is weighted by loan amount (see Appendix for details on lender classification). Banks and bank holding companies (BHCs) domiciled in the United States hold a 22% loan share, on average, while foreign banks hold slightly more than 18%. The largest share is held by U.S. investment funds and other lenders, with about 30%, while the share of CLOs/CDOs is 17%. Insurance companies and pension funds are the smallest lenders, with 3.5% of each syndicate on average. The second column of Table 1 shows loan shares when focusing on loans with non-missing PDs. Banks and BHCs now hold about 55% on average, and the shares of all other lenders are somewhat smaller than when considering all loans, with CLOs/CDOs standing at 11% rather than 17%. In terms of risk taken by the various intermediaries, banks invest in loans with a weighted median PD of about 0.70%, while all the other lenders invest in substantially riskier loans, with weighted median PDs 10

of about 4%. Investment funds and other lenders domiciled in the United States have the highest weighted median PD (8%). In Table 2 we evaluate changes in the composition of the syndicate shortly after origination. Banks, in particular, may facilitate the functioning of the syndicated loan market by originating loans that they intend to sell to other intermediaries quickly afterward. In the table we focus on loans that are in our data at origination and one and two quarters later. We then calculate the average loan share for each intermediary type, weighted by syndicate amounts, at origination and after one and two quarters. Consistent with the hypothesis that banks originate some loans to facilitate market functioning rather than for investment, U.S. and foreign banks reduce their loan shares by 3.4 and 2.7 percentage points, respectively, which is a decline of about 14% of the share at origination. Conversely, all other intermediaries increase their shares. Most of the reallocation happens within the first quarter following origination, because share changes are quite similar when considering shares two quarters after origination. We should point out that the results in this table are likely a lower bound to changes in loan ownership after origination, because banks may, for instance, sell most of their participation before SNC reporting is due at the end of the quarter. 3. Research design Our analysis focuses on how the default risk of investments in the syndicated term-loan market changes when investors expect that U.S. interest rates will remain lower for a longer period of time. We consider term loans because, unlike syndicated loans that provide credit lines, nonbank lenders play a significant role in the term-loan market. The key variable of interest is the loan PD provided by the banks that coordinate each syndicate. Given that we are interested in the credit risk that lenders add to their portfolios, we mostly study the weighted-average PD of portfolio additions, 11

which we define as primary market originations, including renegotiations of existing facilities, and secondary market purchases. The weights are based on each loan’s utilization level; for term loans, there is little difference commitment and utilization. We discuss results for both unbalanced and balanced panels, with the latter only including larger and more sophisticated lenders that are active in the term-loan market in each quarter. Balancing the panel removes participants that add loans to their portfolios only sporadically, and lenderswhoareactiveineverytimeperiodmaypursuedifferentinvestmentstrategiesthantherest. Finally, the SNC data allow us to identify the corporate group to which a certain lender belongs, and we typically measure default risk at the level of the corporate parent. As a consequence, the results are not driven by within-group risk transfers that leave the risk exposure unchanged at the highest decision-making level, where strategic investment decisions are made. We now discuss two issues that are important for the interpretation of our results. First, the analysis identifies a time-series relation using only 16 quarterly observations. While the data clearly have a reduced time-series dimension, it is precisely the period we cover that is characterized by persistently low longer-term interest rates, and a longer sample would not necessarily provide the variation we need to identify the effect in which we are interested. Note that, while interest rates were generally low in the sample we study, they varied significantly: for instance, the 10-year Treasury rate ranged between 1.5% and 4%. The dynamics of the 10-year Treasury rate and the three-year-forward10-yearTreasuryrateafterthelastthreerecessions, showninFigure3, highlight the fact that interest rates have stayed low for longer in the aftermath of the recent financial crisis, when the Federal Reserve engaged in unconventional monetary policy initiatives (Krishnamurthy and Vissing-Jorgensen, 2013) Second, interest rates and changes in portfolio default risk could be endogenous, for instance 12

because of an unobserved credit risk factor that is not captured by the macroeconomic variables we use as controls, even though the PDs in the SNC data measure long-run default risk, which dampens their sensitivity to business-cycle shocks. We address potential endogeneity concerns with a specification wherein the dependent variable is the ratio of the default risk of portfolio additions to the default risk of the existing portfolio, which eliminates the impact of an unobserved credit risk factor that linearly affects the PDs of both portfolio additions and the existing portfolio. Most of our results are based on regressions like the following: (cid:88) (cid:88) log(pdA) = α + I β T + I γ X +q +ε , (1) i,t i j j t j j t j,y i,t j⊂J j⊂J where log(pdA) is the natural logarithm of the weighted-average PD for additions to the portfolio i,t of the corporate group headed by institution i in quarter t (we use log-PDs to reduce the effect of skewness). We classify each corporate parent into seven lender types, indexed with j, using a methodology that we describe in Appendix A and that builds on identifiers from the National Information Center database. The seven categories are: U.S. banks and BHCs, non-U.S. banks and BHCs, insurance companies and pension funds, U.S. CLOs/CDOs, non-U.S. CLOs/CDOs, U.S. investment funds and other lenders, and non-U.S. investment funds and other lenders.8 The variable T is the 10-year U.S. Treasury rate; X is a set of other macroeconomic and t t financial variables that includes the European sovereign yield spread (the difference between the ItalianandGermansovereignyields), ameasureofcreditriskforNorthAmericanspeculative-grade companies (the CDX North American High Yield spread, henceforth CDX HY), the variance risk premium (Bollerslev, Tauchen, and Zhou, 2009), and the University of Michigan index of expected inflation. I(j) is an indicator for lender type j, which means that we estimate the sensitivity of 8 We thank Greg Nini for a discussion on the role of CLOs. 13

risk taking to U.S. Treasury rates and to the other macroeconomic variables at the lender-type level. Each variable is an average of within-quarter values, rather than an end-of-quarter value. We also include lender fixed effects (α ) and lender-type/year fixed effects (q ). The latter term i j,y is meant to account for unobserved common factors that affect risk-taking decisions by specific types of lenders, but the results carry through even without q .9 Throughout the paper, we assess j,y statistical significance on the basis of standard errors double-clustered by time and lender according tothemethodologyofCameron,Gelbach,andMiller(2011). Ineachregression,werequirethateach lendertypecoversatleast1%oftheobservations,withtheexceptionofparticipant-levelregressions, where we set the threshold at 0.5% to account for the fact that lenders are not aggregated at the corporate-group level. We expect the β coefficients to be negative for lender types that increase the riskiness j of their syndicated loan portfolio additions when longer-term interest rates are low. We are also interested in the pattern of coefficients across lender types, in particular for CLOs/CDOs, insurancecompanies/pensionfunds, andinvestmentfunds. Thereasonisthatpreviousstudieshave highlighted how these intermediaries make portfolio choices that are suggestive of or directly in line with search-for-yield incentives. First, Ivashina and Sun (2011) find that the spreads at origination of syndicated loans to which CDOs participate as lenders are more susceptible to compression when institutional demand for syndicated loans is high. Second, Becker and Ivashina (forthcoming) show that certain insurance companies are more likely to seek riskier investments. Third, mutual fund managers have an incentive to take higer risk when they experience poor performance relative to their peers (Kempf, Ruenzi, and Thiele, 2009). Fourth, as we show in the introduction, finance companies greatly increased their funding through largely fixed-rate instruments just before the 10-yearTreasuryratedecreasedtoverylowlevelsin2012, whichmayhavegeneratedagapbetween 9 Statistical significance is generally stronger without lender-type/year fixed effects. 14

asset yields and funding costs because a significant fraction of the assets held by finance companies can be refinanced as interest rates decline. Finally, banks may facilitate the functioning of the market by originating riskier loans in order to accommodate the investment objectives of other lenders. 4. Results We first illustrate the credit-risk dynamics of portfolio additions graphically. In Figure 4 we show median residual log-PDs, by lender type, against residual Treasury rates. Residual PDs are calculated with a regression like equation (1) but without the Treasury rate. Residual Treasury rates are calculated by regressing the Treasury rate on the control variables in equation (1). As shown in Figure 4, which is based on the unbalanced panel, the residual log-PD of banks’ portfolio additions are not particularly responsive to residual interest rates. On the other hand, the risk taking of the other intermediaries increased as orthogonalized interest rates bottomed out in late2012andearly2013, especiallyinthecaseofCLOs/CDOs. Figure5reportsthedollaramounts of additions by lender type over time; it indicates that the value of portfolio additions for nonbank intermediaries increased as interest rates started to decline in 2010, decreased through the second half of 2011 as the European crisis worsened, and started climbing rapidly in early 2012. The key results of our regression analysis are shown in Table 3. The negative and statistically significant coefficient on the Treasury rate indicates that CLOs/CDOs invest in riskier syndicated loans when U.S. interest rates are lower, and the economic effect is substantial. For instance, U.S. CDOs/CLOswithaportfolioadditionsPDof3.71%(whichisthetimeseriesaverageofthequarterly medianPDsshowninFigure4)areexpectedtoincreasethisPDby1.93percentagepointsto5.64%, 15

a change that is about half as large as the initial PD.10 In the unbalanced panel, we find similar results for investment funds and other lenders and for insurance companies and pension funds, although the statistical significance does not carry over to the balanced panel for the latter group. Banks also have a statistically significant negative coefficient in the balanced panel for additions and in both panels for originations, which we define as loans with an origination date within a given reporting quarter. Still focusing on originations, statistical significance is noticeably weaker for CLOs/CDOs and investment funds relative to portfolio additions, especially in the balanced panel, although both economic and statistical significance increase for insurance companies and pension funds in the balanced panel. As shown in Table 2, banks reduce their loan share by 14%, on average, one quarter after origination; the sensitivity of risk taking to interest rates that we find for bank originations could be driven by loans that banks help arrange but expect to quickly sell to other intermediaries. In the two rightmost columns of Table 3, we investigate whether this is indeed the case. The two columns show regression results for originations in which the bank share declined within one quarter and for originations in which the bank share increased, respectively. The coefficients on Treasury rates for banks are negative and statistically significant only for originations in which banks reduce their share relatively quickly. The overall picture that emerges from the discussion above is of a market where a class of shadow-banking lenders, which help finance a relatively small but significant fraction of loans, increases the riskiness of its investments when interest rates decline, especially through secondary-market purchases. Insurance companies and pension funds also invest in riskier loans, 10 The predicted change is calculated by multiplying 3.71% by the expected percent change in PDs implied by the estimated balanced-panel coefficient on the Treasury rate, which is given by e−0.5·βT −1 (the regressions are in semi-logform). Webaseourcalculationsonaquarterlydecreaseof0.5percentagepointsinthe10-yearTreasury ratebecauseD’AmicoandKing(2013)findthatmedium-tolong-datedyieldsfellbyasmuchas0.5percentage points after large-scale Treasury purchases were announced by the Federal Reserve. 16

includingontheprimarymarket,butonaveragetheyholdasmallfractionofeachsyndicate. Banks, which hold large loan shares, appear to facilitate the functioning of the market by accommodating other intermediaries’ investment preferences and originating riskier loans that they tend to sell soon after origination. In the remainder of this section we explore the sensitivity of our results to omitting quarters of particular economic significance, to different ways of measuring credit risk and interest rate expectations, and to an alternative lender classification. We start with Table 4, where we first exclude the second quarter of 2012 (as shown in Figure 5, U.S. CLOs/CDOs and other nonbank lenders added the lowest amount of loans to their portfolio during the quarter). In a separate set of regressions, we also exclude the second quarter of 2013, when interest rates increased rapidly in response to expectations of a more rapid normalization of the monetary policy stance. While the coefficients are generally somewhat smaller and statistical significance slightly weaker, the results carry through. 4.1 Alternative measures of interest rate expectations The results presented in Tables 3 and 4 focus on whether syndicated lenders add riskier credits to theirportfolioswhenthe10-yearU.S.Treasuryratedeclines. Wenowevaluatetherobustnessofour results to using three other measures of expectations of longer-term interest rates. We first consider the three-year-forward 10-year U.S. Treasury rate, which is the financial market’s expectation of the 10-year rate in three years’ time. The other two measures are built using the term structure of the federal funds rate implied by overnight indexed swaps (OIS), which are derivative contracts of varying maturities whose payoff depends on the future evolution of short-term unsecured interest rates, in our case the federal funds rate. Using this term structure, we calculate the difference 17

between the expected federal funds rate 10 quarters ahead and the current federal funds rate, as well as the number of quarters before the expected federal funds rate reaches 25 basis points. The Treasury rate forward is shown in the right chart of Figure 3, while the measures based on the federal funds rate term structure are shown in Figure 6. In all cases, the dynamics of the measures of interest rate expectations closely follow those of the 10-year Treasury rate. Weuseeachofthesethreemeasuresinplaceofthe10-yearTreasuryratesinasetofregressions identical to equation (1). To be consistent with the results discussed so far, the coefficients on the three-year forward Treasury rate and on the difference between the expected and current federal funds rate should be negative, because both variables decline when investors expect interest rates to remain low for longer. On the other hand, the coefficients on the expected number of quarters beforethefederalfundsratereaches25basispointsshouldbepositive. TheresultsshowninTable5 are similar to those discussed so far, in terms of both statistical significance and relative magnitude. The only exception is that none of the bank coefficients are statistically significant when using the expected number of quarters until the fed funds rate reaches 25 basis points. 4.2 Robustness checks Our sample of syndicated loan data begins in 2010; however, the time series of macroeconomic variables, which is constrained by the availability of the CDX HY spread, goes back to 2004. In the first two columns of Table 6, we obtain the sensitivity of risk taking to interest rates in two steps. WefirstorthogonalizetheTreasuryratewithrespecttotheotherfourmacroeconomicvariablesover the 2004–2013 sample, and then we use the orhogonalized series (T⊥) as the independent variable t 18

in a pooled regression similar to eq. (1): (cid:88) log(pdA) = α + I β T⊥+q +ε . (2) i,t i j j t j,y i,t j⊂J Including Treasury rates that are already orthogonalized relative to macroeconomic state variables is similar to Dell’Ariccia, Laeven, and Suarez (2014)’s use of Taylor residuals to identify exogenous shocks to monetary policy. With the exception of banks, the coefficients are generally larger than in Table 3. Statistical significance is also stronger, especially for non-U.S. banks and BHCs. We should point out that the t-statistics in the second regression do not account for the fact that the orthogonalized Treasury rates series is estimated in the first step, with the consequence that the statistical significance of second-stage coefficients will be overstated to some extent. We now address the endogeneity that could arise from the presence of an unobserved factor that affects both the default risk of additions and the Treasury rate, like an economy-wide default risk factor. For each lender, we study the logarithm of the ratio of gross addition PDs to the PDs of the existing portfolio. If the potentially omitted factor (Ω ) affects the PDs of additions (pdA) t i,t and of the outstanding portfolio (pdO) linearly, then taking the ratio simplifies the factor out and i,t expressesthefactorloadingofnewlyacquiredloansasamultipleofthefactorloadingoftheexisting portfolio: pdA = θA ×Ω , i,t i,t t pdO = θO ×Ω , i,t i,t t 19

and pdA θA i,t i,t = . pdO θO i,t i,t As a result, the dependent variable can be interpreted as the change in the current investment strategy relative to the average investment strategy implied by the existing loan portfolio. We estimate the same regression as in equation (1), where the dependent variable is log(pdA/pdO) i,t i,t and the probabilities of default are, as before, weighted by loan utilization. The results, reported in the third and fourth columns of Table 6, are similar to those shown in Table 3 for banks and CLOs/CDOs, but they are statistically insignificant for investment funds and others. While our discussion so far has been centered on loan purchases, lenders can adjust the riskiness of their portfolios through sales as well as purchases. The fifth and sixth columns of Table 6 show results from regressions where the dependent variable is the log-ratio of addition PDs to disposition PDs, where dispositions are loans that disappear from a lender’s portfolio in a given quarter, and their PDs are also weighted by loan participation amounts. The reported coefficients are in line with the results so far. In the last two columns of Table 6, we study risk taking at the level of individual lenders without consolidating loan portfolios at the corporate-group level. Given that finance companies are often the credit arm of a larger group rather than stand alone credit providers, only in this specification are we able to study them as a separate, albeit small, category. The coefficient shows that finance companies invest in riskier syndicated loans when interest rates decline. In this specification we also focus on individual banks and exclude BHCs from the sample, which allows us to reduce the impact of subsidiaries not involved in core banking activities. The coefficient on banks is 40% lower than in Table 3 and is less statistically significant, indicating a weaker relation 20

between risk taking and interest rates. In a final robustness check, we present results based on a fully manual classification of the lenders in the balanced sample. The unbalanced sample includes a much larger cross-section of lenders, but many are private credit providers with nondescript names and it is difficult to obtain information on their activities from public sources. As discussed in Appendix A, we do not sort lenders on the basis of whether they are domiciled in the United States or not, and we use six categories: banks and BHCs, nonbank financials, insurance companies and pension funds, retail-orientedassetmanagers,CLOs/CDOs,andnon-financials. Weshouldhighlightthatanumber oflendersclassifiedasinvestmentfundsandotherlendersinthemainanalysisare,inourassessment, tranched securitizations and should thus be considered CLOs/CDOs. This change likely explains why neither of the two categories derived from investment funds and other lenders has statistically significantcoefficientsinTable7,wherewepresenttheresultsbasedonthenewclassification. These two categories are: nonbank financials, which mainly include private investment organizations and wealth management companies, and retail-oriented asset managers, which typically comprise funds available to retail or institutional investors and which are associated with asset managers with an established retail presence. The remainder of the results in Table 7 are similar to those presented previously for banks and BHCs, insurance companies and pension funds, and CLOs/CDOs. 4.3 Treasury rates and the pricing of syndicated loans The discussion so far has focused on the credit risk of syndicated loans. As noted by Ivashina and Sun (2011), strong investor demand throughout the 2000s resulted in a compression of syndicated loan spreads, and a natural question is whether the risk-taking patterns we find are also reflected in loan spreads. Pricing information is not available in the SNC data, but we are able to manually match about one-third of the loan originations with PDs in SNC to spreads from the commercial 21

data set Thomson Reuters DealScan. We consider originations because Thomson Reuters DealScan only provides loan spreads at origination. Table 8 shows summary statistics for the spreads and PDs of the loans on which we focus in this section. Weevaluatewhetherlonger-terminterestratesaffectloanspreadswithasetofregressionsthat relate log-spreads to selected loan characteristics and several macroeconomic variables, including 10-year Treasury rates. The full specification is (cid:88) (cid:88) log(ais ) = β L + γ M +δS +νS ×T +q +q +ε , (3) i,t c c,i m m,t i i t b,y ab,y i c⊂C v⊂V where ais is loan i’s spread; L a set of loan characteristics that includes each loan’s log PD; i,t c,i M is a set of macro variables that includes the 10-year Treasury rate; S is the loan share held by m,t i nonbank lenders; and q and q are borrower-industry/year and agent-bank/year fixed effects, b,y ab,y respectively. The rationale for including nonbank loan share as a covariate is that the presence of certain nonbank intermediaries (CDOs) makes spreads more sensitive to lenders’ willingness to provide credit (Ivashina and Sun, 2011). To the extent that low Treasury rates increase this willingness to lend, their effects on spreads could be stronger for loans with a higher fraction of nonbank lenders. As shown in Table 9, default risk and the fixed effects explain more than half of the variation in loan spreads, and, as expected, higher PDs are associated with higher spreads. Longer duration andsmallerloanamountsalsocontributetohigherspreads, asdoesriskaversion, whichismeasured with the variance risk premium. The coefficient on the 10-year Treasury rate is not statistically significant, unless the nonbank loan share and its interaction with the Treasury rate are included, in which case the coefficient is negative. The economic effect is also relatively small, because a 0.5 percentagepointquarter-on-quarterdecreaseintheTreasuryratewouldraisethemedianspreadfor 22

investment-gradeloansfrom150to160basispoints.11 Thenonbankshareenterswithapositiveand statistically significant coefficient, which may reflect the higher riskiness of loans held by nonbank intermediaries (we control for PDs, but loan spreads may also capture other dimensions of credit risk, like loss given default). The interaction between the Treasury rate and the nonbank share has a small and statistically insignificant coefficient. While the relation between Treasury rates and PDs suggests that lenders increase their risk taking when interest rates are low, the relation between Treasury rates and loan spreads seems to imply that the appetite for riskier credits translates, if anything, into higher loan spreads. These two findings, however, are not necessarily at odds. As noted above, loan spreads may embed more accurate information about default risk than PDs do, for instance because they incorporate lenders’ expected loss given default. In this case, higher spreads would reflect the same increased risk taking that we find when studying PDs. A second possibility is that the increase in risk taking that goes with low Treasury rates may be accompanied by a change in the characteristics of borrowers. For instance, lenders may be more willing to provide credit to companies for which, keeping default risk constant, information is more difficult to gather, and they may demand higher spreads to compensate for this additional business cost (see Easley and O’Hara, 2004, for the effect of information availability on the cost of capital). 5. Conclusions We use supervisory data to study the risk-taking behavior of banks and nonbank lenders in the primary and secondary U.S. syndicated loan market during the recent period of low interest rates. We find that certain financial intermediaries—in particular, structured finance vehicles—increase 11 The regressions are in semi-log form, and the expected percent change in the loan spread for a decrease of the 10-year Treasury rate of 0.5 percentage points in a quarter is given by e−0.5·βT −1. 23

the riskiness of their syndicated loan portfolios when longer-term interest rates are expected to remain low. We mainly focus on the 10-year Treasury rate as an explanatory variable, but the results are robust to using alternative measures for interest rate expectations. We also study the relation between Treasury rates and loan spreads for a sub-sample of loans for which we were able to source pricing information. We find that spreads increase modestly when interest rates are low. Theseresultsareconsistentwitha“searchforyield”bylendersinthesyndicatedloanmarket. In light of the evidence that unconventional monetary policy put downward pressure on longer-term interest rates, our results are also consistent with the existence of a risk-taking channel of monetary policy. Our findings should be interpreted in light of several caveats. First, we focus only on part of an intermediary’s portfolio – syndicated term loans – and the additional risk may be small relative to the overall portfolio, or the intermediary may be actively hedging the additional risk. Second, the dynamics of loan pricing and of recovery rates may be such that lenders are appropriately compensatedfortheadditionalriskorexpectedlossesremainstable. Finally, anyeffectofmonetary policy on risk taking must be evaluated against the broader benefits of accommodative monetary policy. The syndicated loan literature, for instance, has highlighted the fact that loan supply is adversely affected by negative liquidity and capital shocks to lenders (see Ivashina and Scharfstein, 2010, and Santos, 2011). 24

A. Lender classification We assign each lender to one of seven categories on the basis of the Entity Type code provided by the National Information Center database, which is made available in the SNC data.12 We then refine the classification manually using lender names, as discussed below. The number of lenders in each category is shown in parentheses. Entity Type codes are shown in italics. 1) U.S. banks and bank holding companies (20): BHC (Bank Holding Company), FHD (Financial Holding Company) if the lender is domiciled in the United States, FCU (Federal Credit Union), FSB (Federal Savings Bank), NAT (National Bank), NMB (Non-member Bank),SAL(Savings&LoanAssociation),SMB (StateMemberBank),orSSB (StateSavings Bank). 2) Non-U.S. banks and bank holding companies (24): FBH (Foreign Banking Organization as a Bank Holding Company), FHD (Financial Holding Company) if the lender is domiciled outside the United States, FBK (Foreign Bank), FBO (Foreign Banking Organization), or FHF (Financial Holding Company/Foreign Banking Organization). 3) Insurance companies and pension funds (14): the lender name includes “Insurance,” “Reinsuarance,” “Assurance,” “Reassurance,” “Retirement,” “Pension,” or “Pensioen,” as long as the Entity Type is not BHC or FHD and the name does not contain “401” or “Superannuation.” Note that we do not distinguish U.S./non-U.S. insurance companies or pension funds, and that 401k-style pension funds (including Australian superannuations) are considered investment funds because they are similar to tax-advantaged mutual funds. We also manually search lender names to ensure that the 25 largest U.S. insurance companies by 12 See page 14 of the data dictionary at http://www.ffiec.gov/nicpubweb/nicweb/DataDownload.aspx for details on this variable. 25

assets as of the third quarter of 2014 (based on data from SNL Financial) are classified as insurance companies. 4) U.S. CLOs/CDOs (80): the lender name includes “CDO” or “CLO,” the Entity Type is not BHC or FHD, and the lender is domiciled in the United States. 5) Non-U.S. CLOs/CDOs (29): thelendernameincludes“CDO”or“CLO,”theEntityType is not FBH or FHD, and the lender is domiciled outside the United States. 6) U.S. investment funds and others (108): DEO (Domestic Entity Other); if, for lenders domiciled in the United States, the lender name contains “Fund” and the Entity Type is not BHC or FHD, and the lender is not classified as CLO/CDO or insurance company/pension fund; and if, for lenders domiciled in the United States, the lender name contains “401” and theEntityTypeisnotBHC orFHD,andthelenderisnotclassifiedasCLO/CDOorinsurance company/pension fund. 7) Non-U.S. investment funds and others (24): FEO (ForeignEntityOther); if, forlenders domiciled outside the United States, the lender name contains “Fund” and the Entity Type is notFBH orFHD,andthelenderisnotclassifiedasCLO/CDOorinsurancecompany/pension fund; and if, for lenders domiciled outside the United States, the lender name contains “Superannuation” and the Entity Type is not FBH or FHD, and the lender is not classified as CLO/CDO or insurance company/pension fund. The Entity Type variable can also take the value SLCH (Savings and Loan Holding Company), for which we classify lenders manually into one of the categories described above. As a rule, we do not reclassify lenders on the basis of their name if the entity type is BHC, FBH, or FHD because we have found these codes to reliably classify lenders. One specification in Table 6 includes an analysis at the participant level, rather than at the level of the corporate group parent. In this case, we use 26

the approach described previously to classify lenders, with the exception that we do not include U.S. bank holding companies in category 1. We also include finance companies (Entity Type FNC), whose classification we review manually on the basis of lender names. In Table 7 we present results using a manual classification of the lenders in the balanced sample, in which we research information on each lender’s activities. We construct one category each for “Banks and bank holding companies” and “CLOs/CDOs,” irrespective of whether they are domiciled in the United States or not, and we reclassify U.S./non-U.S. investment funds and others into “Nonbank financials,” “Retail-oriented asset managers,” and “Non financials.” Retail-oriented asset managers include companies with an established presence in the retail market, but their funds that participate in the syndicated-loan market are not necessarily plain-vanilla mutual funds. The six categories of the manual classification, highlighting the differences relative to the main classification, are the following: A) Banks and bank holding companies (44): The first two categories of the main classification are included in this category. B) Nonbank financials (18): These lenders appear to be private investment companies or wealth management companies, and were classified as 6 or 7. C) Insurance companies and pension funds (15): A small number of lenders previously classified as 6 are reclassified as insurance companies and pension funds, including insurance companies and investment funds for company pension plans. Public pension funds are now classifiedasnon-financialslenders,becauseplanfundingandinvestmentchoicescanbeaffected by political considerations, and their sponsors have taxing power. D) Retail-oriented asset managers (63): All these lenders were previously classified as either 6 or 7, and include investment funds associated with asset managers with an established retail 27

presence. E) CLOs/CDOs (142): All lenders in categories 4 and 5 are in this category. 33 lenders that were previously classified as investment funds and others are now classified as CLOs/CDOs. These are lenders whose names do not contain the keywords used in the main classification and that we judge to be tranched securitizations, normally using brief descriptions of the structure that are attached to notices of publicaly available rating changes. F) Non financials (17): This category includes charitable foundations, private trusts, health insurance companies, and public treasury departments and pension plans. We consider health insurance companies non-financial lenders because premia tax deductability and employer plan sponsorship may provide these companies with different pricing power and investment incentives than life, property, and casualty insurers. As discussed above, we consider public pension funds non-financial lenders because their sponsors are directly exposed to political considerations and have taxing power. The large majority of lenders were classified as 6 or 7, and a small number were classified as 4. 28

References Altunbas, Y., L. Gambacorta, and D. Marques-Ibanez (forthcoming): “Does monetary policy affect bank risk-taking?,” International Journal of Central Banking. Becker, B., and V. Ivashina (forthcoming): “Reaching for yield in the bond market,” Journal of Finance. Black, L., and L. Hazelwood (2013): “The effect of TARP on bank risk-taking,” Journal of Financial Stability, 9(4), 790–803. Bollerslev, T., G. Tauchen, and H. Zhou (2009): “Expected stock returns and variance risk premia,” Review of Financial Studies, 22(11), 4463–4492. Borio, C., and H. Zhu (2012): “Capital regulation, risk-taking and monetary policy: A missing link in the transmission mechanism?,” Journal of Financial Stability, 8(4), 236–251. Buch, C., S. Eickmeier, and E. Prieto (2014): “In search for yield? Survey-based evidence on bank risk taking,” Journal of Economic Dynamics and Control, 43, 12–30. Cameron, A., J. Gelbach, and D. Miller(2011): “Robustinferencewithmultiwayclustering,” Journal of Business & Economic Statistics, 29(2), 238–249. Campello, M. (2002): “Internal capital markets in financial conglomerates: evidence from small bank responses to monetary policy,” Journal of Finance, 57(6), 2773–2805. Chevalier, J., and G. Ellison(1997): “Risktakingbymutualfundsasaresponsetoincentives,” Journal of Political Economy, 105(6), 1167–1200. Chodorow-Reich, G. (2014): “Effects of unconventional monetary policy on financial institutions,” Brookings Papers on Economic Activity, Spring, 155–204. D’Amico, S., W. English, D. Lopez-Salido, and E. Nelson (2012): “The Federal Reserve’s large-scale asset purchase programmes: rationale and effects,” Economic Journal, 122(564), F415–F446. D’Amico, S., and T. King (2013): “Flow and stock effects of large-scale treasury purchases: evidence on the importance of local supply,” Journal of Financial Economics, 108(2), 425–448. Degryse, H., and S. Ongena (2007): “The impact of competition on bank orientation,” Journal of Financial Intermediation, 16(3), 399–424. Dell’Ariccia, G., L. Laeven, and G. Suarez (2014): “Bank leverage and monetary policy’s risk-taking channel: evidence from the United States,” Working paper, Columbia University. Di Maggio, M., and M. Kacperczyk (2015): “The unintended consequences of the zero lower bound policy,” Working paper, Columbia Business School. Duchin, R., and D. Sosyura (2012): “Safer ratios, riskier portfolios: banks’ response to government aid,” Working paper. Easley, D., and M. O’Hara (2004): “Information and the cost of capital,” Journal of Finance, 59(4), 1553–1583. 29

Gorton, G., andG.Pennacchi(1995): “Banksandloansales: marketingnonmarketableassets,” Journal of Monetary Economics, 35(3), 389–411. Ioannidou, V., S. Ongena, and J. Peydro´ (2015): “Monetary policy, risk-taking and pricing: evidence from a quasi-natural experiment,” Review of Finance, 19(1), 95–144. Ivashina, V., and D. Scharfstein (2010): “Bank lending during the financial crisis of 2008,” Journal of Financial Economics, 97(3), 319–338. Ivashina, V., and Z. Sun (2011): “Institutional demand presure and the cost of corporate loans,” Journal of Financial Economics, 99(3), 500–522. Jimenez, G., S. Ongena, J. Peydro´, and J. Saurina (2014): “Hazardous times for monetary policy: what do twenty-three million bank loans say about the effects of monetary policy on credit-risk taking?,” Econometrica, 82(2), 463–505. Jones, J., W. Lang, and P. Nigro (2005): “Agent bank behavior in bank boan syndications,” Journal of Financial Research, 28(3), 385–402. Kempf, A., S. Ruenzi, and T. Thiele (2009): “Employment risk, compensation incentives, and managerial risk taking: evidence from the mutual fund industry,” Journal of Financial Economics, 92(1), 92–108. Krishnamurthy, A., and A. Vissing-Jorgensen (2013): “The ins and outs of LSAPs,” Symposium on Global Dimensions of Unconventional Monetary Policy, Federal Reserve Bank of Kansas City, pp. 57–111. Li, L. (2013): “TARP funds distribution and bank loan supply,” Journal of Banking & Finance, 37(12), 4777–4792. Maddaloni, A., and J. Peydro´ (2011): “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), 2121–2165. Paligorova, T., and J. Santos(2013): “MonetaryPolicyandBankRisk-Taking: Evidencefrom the Corporate Loan Market,” Working paper. Qiu, J. (2003): “Termination risk, multiple managers and mutual fund tournaments,” European Finance Review, 7(2), 161–190. Santos, J.(2011): “Bankcorporateloanpricingfollowingthesubprimecrisis,”ReviewofFinancial Studies, 24(6), 1916–1943. Spiegel, M., and H. Zhang (2013): “Mutual fund risk and market share-adjustes fund flows,” Journal of Financial Economics, 108(2), 506–528. 30

Figure 1: Annual Shared National Credit: Commitment and utilization trends by loan type The charts show the time series of commitments and utilization, by loan type, in the annual SNC data. OOvveerraallll ccoommmmiittmmeenntt,, bbyy llooaann ttyyppee VVoolluummee ((iinn bbiilllliioonnss)) AAnnnnuuaall 33000000 Revolving Credit Term Loans Other 22550000 22000000 11550000 11000000 550000 00 11999933 11999955 11999977 11999999 22000011 22000033 22000055 22000077 22000099 22001111 22001133 OOvveerraallll uuttiilliizzaattiioonn,, bbyy llooaann ttyyppee VVoolluummee ((iinn bbiilllliioonnss)) AAnnnnuuaall 11660000 Revolving Credit Term Loans 11440000 Other 11220000 11000000 880000 660000 440000 220000 00 11999933 11999955 11999977 11999999 22000011 22000033 22000055 22000077 22000099 22001111 22001133 Note: SNC program reviews are conducted annually in May using data provided by agent banks, typically as of December 31 of the prior year, and sometimes as of March 31 of the review year. 31

Figure 2: Shared National Credit: Commitment and utilization trends, quarterly vs. annual data The charts show the time series of annual and quarterly commitments and utilization by loan type. The charts also report volumes according to whether probabilities of default (PDs) are available. SNC Commitment Trends Billions 3500 Annual SNC (all banks) Quarterly (18 banks) Quarterly, PD not missing 3000 2500 2000 1500 1000 500 0 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 SNC Utilization Trends - Term Loans Billions 1000 Annual SNC (all banks) Quarterly (18 banks) Quarterly, where PD not missing 800 600 400 200 0 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 N o te: SNC program reviews are conducted annually in May using data provided by agent banks, typically as of December 31 of the prior year, and sometimes as of March 31 of the review year. 32

Figure 3: Interest rates during the last three recessions The charts report 10-year Treasury rates (left) and three-year-forward 10-year Treasury rates (right) through the three most recent recessions. The horizontal axes show quarters from the end of the recessions. 10-Year Treasury Rate 10-Year Treasury Rate Three Years Forward Percent Percent 10 10 1990 1990 2001 2001 2007 2007 8 8 6 6 4 4 2 2 0 - 6 - 3 0 3 6 9 12 1 5 1 8 - 6 - 3 0 3 6 9 12 1 5 1 8 Quarters Quarters 33

Figure 4: The default risk of gross portfolio additions by lender type The charts show the time-series evolution of the residual log-probabilities of default of gross portfolio additions, by lender type, using the unbalanced panel. The orthogonalized Treasury rate is also shown in the chart. Residual PDs are calculated with a regression like equation (1) but without the Treasury rate. Residual Treasury rates are calculated by regressing the Treasury rate on the control variables in equation (1). U.S. Banks and Bank Holding Companies Foreign Banks and Bank Holding Companies Residuals Residuals Median of log(PD) residuals Orthogonalized 10-year Treasury rate 1 1 0 0 -1 -1 2010 2011 2012 2013 2010 2011 2012 2013 Insurance Companies and Pension Funds CLOs and CDOs Residuals Residuals U.S. median of log(PD) residuals 1 Foreign median of log(PD) residuals 1 Orthogonalized 10-year Treasury rate 0 0 -1 -1 2010 2011 2012 2013 2010 2011 2012 2013 U.S. Investment Funds and Other Lenders Foreign Investment Funds and Other Lenders Residuals Residuals 1 1 0 0 -1 -1 2010 2011 2012 2013 2010 2011 2012 2013 34

Figure 5: Portfolio additions by lender type The charts show the amount of portfolio additions (in $ billion) by lender type, using the unbalanced panel. U.S. banks and bank holding companies Foreign banks and bank holding companies 60 35 50 30 25 40 20 30 15 20 10 10 5 0 0 2010 2011 2012 2013 2010 2011 2012 2013 Insurance companies and pension funds CLOs and CDOs 6 25 Total U.S. 5 20 4 15 3 10 2 5 1 0 0 2010 2011 2012 2013 2010 2011 2012 2013 U.S. investment funds and other lenders Foreign investment funds and other lenders 60 18 16 50 14 40 12 10 30 8 20 6 4 10 2 0 0 2010 2011 2012 2013 2010 2011 2012 2013 35

Figure 6: Monetary policy expectations from OIS quotes The left chart shows the spread between the expected federal funds rate 10 quarters ahead and the current federal funds rate. The right chart shows, at a given point in time, the number of quarters before the federal funds rate is expected to reach 25 basis points. Expected federal funds rates at various time horizons are obtained from overnight indexed swaps (OIS) quotes. Spread Between the Expected and Current Quarters Before Liftoff Federal Funds Rates Percent Quarters 3.0 8 2.5 6 2.0 1.5 4 1.0 2 0.5 0 0.0 2009 2010 2011 2012 2013 2014 2009 2010 2011 2012 2013 2014 36

Table 1: Loan summary statistics by rating grade and lender type ThetoppanelshowsthepercentofloansthathavebeenassignedagivenratinggradebySNCexaminers,thepercent of loans with a given rating grade that have a probability of default (PD), and the median PD by rating grade. The bottom panel shows the weighted average lender share in a syndicated loan, weighted by loan amount, according to whether the loan has a PD or not. The table also shows the weighted median PD, with the weight given by the participation amount. Percent of Loans Percent of Loans (with PD) Median PD Loan Rating Grade 35% ceiling Pass 76.2 79.5 0.78 0.78 Special Mention 9.5 8.0 5.58 5.58 Substandard 11.2 9.1 29.41 29.41 Doubtful 2.0 2.0 100 35 Loss 1.1 1.4 100 35 Av. loan share Av. loan share (with PD) Weighted median PD Lender type 35% ceiling U.S. banks and BHCs 22.4 29.9 0.71 0.71 Non-U.S. banks and BHCs 18.3 26.3 0.65 0.65 Insurance cos./Pension funds 3.5 2.3 4.26 4.26 U.S. CLOs/CDOs 10.3 6.7 3.25 3.25 Non-U.S. CLOs/CDOs 6.7 4.4 4.50 4.50 U.S. inv. funds and others 30.3 24.4 7.95 7.95 Non-U.S. inv. funds and others 8.7 5.9 4.82 4.82 Table 2: Lender market share at origination and one or two quarters after origination The table reports the market share of each lender type at origination and after one or two quarters. The sample includes loans whose origination date is within the reporting quarter. Share held by lender type (in %) Share held by lender type (in %) At orig. +1 qrt ∆ ∆% At orig. +2 qrts ∆ ∆% U.S. banks and BHCs 25.7 22.3 -3.4 -13.2% 25.7 22.2 -3.5 -13.5% Non-U.S. banks and BHCs 18.1 15.4 -2.7 -14.7% 18.1 15.1 -2.9 -16.3% Insurance cos./Pension funds 3.6 3.9 0.4 10.4% 3.6 3.9 0.3 9.5% U.S. CLOs/CDOs 8.3 9.2 0.9 11.1% 8.7 9.7 1.1 12.3% Non-U.S. CLOs/CDOs 5.8 6.7 0.9 15.4% 5.8 7.1 1.2 21.3% U.S. inv. funds and others 30.3 32.7 2.4 8.0% 30.1 32.3 2.2 7.2% Non-U.S. inv. funds and others 8.3 9.8 1.5 17.4% 8.1 9.7 1.6 19.5% 37

stluser niaM :tekram naol detacidnys eht ni gnikat ksiR :3 elbaT .3 noitceS ni dessucsid noisserger lenap eht gnisu detamitse ,epyt rednel yb ,etar yrusaerT raey-01 eht ot ksir tiderc oiloftrop fo ytivitisnes eht swohs elbat ehT eht ,daerps ngierevos naeporuE eht era stneicffieoc detropernu htiw srossergeR .tnerap etaroproc eht ta soiloftrop naol fo DP gol eht si elbairav tnedneped ehT A hcaeninwohssDPehtfoegarevaseires-emitehtsi DP .noitaflnidetcepxefoxedninagihciMfoytisrevinUehtdna,muimerpksirecnairaveht,daerpsYHXDC A detciderpehtsi DP∆ .lenapdecnalabehtnietaryrusaerTehtnotneicffieoctnacfiingisyllacitsitatsahtiwsepytrednelroftroperewhcihw,strahcs’4erugiFfo A A egnahc tnecrep detcepxe eht yb DP gniylpitlum yb detaluclac si DP∆ .stniop egatnecrep 5.0 fo etar yrusaerT eht ni esaerced ylretrauq a rof sDP ni egnahc scitsitats-t .stceffe dexfi raey/epyt-rednel dna rednel edulcni snoisserger ehT .)mrof gol-imes ni era snoisserger eht( 1− Tβ·5.0−e yb nevig si hcihw ,DP eht ni ,slevel %01 dna ,5 ,1 eht ta ecnacfiingis lacitsitats etacidni ∗ ,∗∗ ,∗∗∗ slobmys ehT .rednel dna retrauq yb deretsulc-elbuod era srorre dradnatS .sesehtnerap ni .3102 ot 0102 sedulcni elpmas ehT .epyt rednel detacidni eht sa defiissalc snoitavresbo fo tnecrep eht wohs scitsitats-t eht ot stpircsrepuS .ylevitcepser yb ,snoitanigirO ni egnahc erutuf snoitanigirO snoitiddA erahs naol knab A A )lenap decnalabnu( decnalaB .labnU DP∆ DP .laB .labnU esaercnI esaerceD 840.0 ∗444.0- ∗∗886.0- ∗∗373.0- 84.0 73.1 ∗∗706.0- 201.0sCHB dna sknab .S.U 6.3)41.0( 4.4)30.2-( 8.21)74.2-( 7.3)02.2-( 7.6)46.2-( 5.3)77.0-( 215.0 ∗∗894.0- 362.0- ∗∗013.0- 570.0 730.0sCHB dna sknab .S.U-noN 1.4)12.1( 9.3)98.2-( 4.71)64.1-( 8.3)41.2-( 0.8)15.0( 6.3)02.0-( 175.0- ∗573.1- ∗∗140.1- ∗330.1- 044.0- ∗∗024.0sdnuf noisneP/.soc ecnarusnI 6.9)49.0-( 8.9)19.1-( 3.2)54.2-( 6.01)87.1-( 7.4)36.1-( 5.01)32.2-( 585.0 585.0- 868.0- 845.0- 39.1 17.3 ∗∗∗238.0- ∗∗696.0sODC/sOLC .S.U 7.71)82.1( 0.51)12.1-( 3.32)47.1-( 7.31)62.1-( 8.62)46.4-( 6.21)78.2-( 672.0- 037.0- 975.0- ∗917.0- 35.1 84.3 ∗∗∗527.0- ∗∗∗766.0sODC/sOLC .S.U-noN 6.21)46.0-( 7.11)53.1-( 7)09.0-( 8.01)19.1-( 7.9)98.3-( 5.9)56.3-( 673.0- ∗650.1- 335.0- ∗968.0- 862.0- ∗204.0srehto dna sdnuf .vni .S.U 8.73)45.0-( 2.04)68.1-( 4.13)11.1-( 4.14)58.1-( 1.63)70.1-( 7.34)30.2-( 967.0- ∗250.1- *705.1- ∗∗378.0- 50.2 55.5 ∗∗636.0- ∗∗∗583.0srehto dna sdnuf .vni .S.U-noN 8.41)05.1-( 1.51)09.1-( 8.5)77.1-( 9.51)12.2-( 0.8)71.2-( 6.61)11.3-( 851,21 343,02 673,1 702,82 487,4 789,93 snoitavresbO 35.0 24.0 54.0 14.0 73.0 04.0 2R .jdA 38

”murtnat repat“ eht fo tceffe ehT :tekram naol detacidnys eht ni gnikat-ksiR :4 elbaT .3 noitceS ni dessucsid noisserger lenap eht gnisu detamitse ,epyt rednel yb ,etar yrusaerT raey-01 eht ot ksir tiderc oiloftrop fo ytivitisnes eht swohs elbat ehT eht era stneicffieoc detropernu htiw srossergeR .tnerap etaroproc eht fo level eht ta soiloftrop naol fo tluafed fo ytilibaborp gol eht si elbairav tnedneped ehT edulcni snoisserger ehT .noitaflni detcepxe fo xedni nagihciM fo ytisrevinU eht dna ,muimerp ksir ecnairav eht ,daerps YH XDC eht ,daerps ngierevos naeporuE etacidni ∗ ,∗∗ ,∗∗∗ slobmys ehT .rednel dna retrauq yb deretsulc-elbuod era srorre dradnatS .sesehtnerap ni scitsitats-t .stceffe dexfi raey/epyt-rednel dna rednel rednel detacidni eht sa defiissalc snoitavresbo fo tnecrep eht wohs scitsitats-t eht ot stpircsrepuS .ylevitcepser ,slevel %01 dna ,5 ,1 eht ta ecnacfiingis lacitsitats .3102 ot 0102 sedulcni elpmas ehT .epyt 2Q:3102 gnidulcxE 2Q:2102 gnidulcxE ,snoitanigirO snoitiddA ,snoitanigirO snoitiddA egnahc erahs knab yb egnahc erahs knab yb )lenap decnalabnu( .laB .labnU )lenap decnalabnu( .laB .labnU esaercnI esaerceD esaercnI esaerceD 111.0 ∗264.0- ∗∗916.0- 480.0- 140.0- ∗024.0- ∗∗516.0- 390.0sCHB dna sknab .S.U 9.3 )23.0( 4.4)48.1-( 7.6)33.2-( 4.3)75.0-( 5.3)41.0-( 2.4)88.1-( 8.5)27.2-( 4.3)86.0-( 185.0 ∗∗284.0- 212.0 940.0 913.0 ∗∗634.0- 061.0 020.0 sCHB sknab .S.U-noN 4.4)34.1( 0.4)75.2-( 0.8)93.1( 7.3)62.0( 9.3)68.0( 8.3)66.2-( 2.7)49.0( 6.3)01.0( 422.0- ∗794.1- ∗184.0- ∗614.0- 175.0- ∗∗∗494.0- 460.0- ∗∗172.0sdnuf noisneP/.soc ecnarusnI 2.9)36.1( 7.9)56.1-( 7.4)36.1-( 3.01)09.1-( 6.9)49.0-( 8.9)50.3-( 9.4)43.0-( 4.01)82.2-( 346.0 637.0- ∗∗∗087.0- ∗∗955.0- 585.0 242.0- ∗∗∗077.0- ∗∗216.0sODC/sOLC .S.U 5.81)57.1( 4.51)89.1-( 8.62)79.3-( 9.21)83.2-( 7.71)82.1( 2.51)39.0-( 7.72)75.4-( 7.21)26.2-( 211.0- 348.0- ∗∗∗527.0- ∗∗∗676.0- 672.0- 732.0- ∗∗∗527.0- ∗∗∗316.0sODC/sOLC .S.U-noN 6.21)83.0-( 7.11)36.1-( 7.9)25.3-( 6.9)17.3-( 6.21)46.0-( 8.11)90.1-( 5.11)80.4-( 6.9)22.3-( 820.0- ∗951.1- 032.0- 353.0- 414.0- ∗824.0- 431.0- ∗652.0srehto dna sdnuf .vni .S.U 2.73)50.0-( 9.93)40.2-( 1.63)08.0-( 6.34)65.1-( 8.73)95.0-( 0.04)59.1-( 9.43)17.0-( 7.34)18.1-( 125.0- ∗141.1- ∗916.0- ∗∗853.0- 608.0- ∗254.0- 424.0- ∗∗∗982.0srehto dna sdnuf .vni .S.U-noN 2.41)42.1-( 0.51)41.2-( 0.8)58.1-( 6.61)73.2-( 8.41)25.1-( 3.51)51.2-( 1.8)45.1-( 7.61)11.3-( 924,01 688,71 584,4 541,63 611,21 615,91 502,5 052,83 snoitavresbO 45.0 54.0 53.0 93.0 35.0 64.0 93.0 24.0 2R .jdA 39

snoitatcepxe etar tseretni fo serusaem evitanretlA :tekram naol detacidnys eht ni gnikat-ksiR :5 elbaT noisserger lenap eht gnisu detamitse ,epyt rednel yb ,snoitatcepxe etar tseretni fo erusaem detacidni eht ot ksir tiderc oiloftrop fo ytivitisnes eht swohs elbat ehT detropernu htiw srossergeR .tnerap etaroproc eht fo level eht ta soiloftrop naol fo tluafed fo ytilibaborp gol eht si elbairav tnedneped ehT .3 noitceS ni dessucsid ehT .noitaflni detcepxe fo xedni nagihciM fo ytisrevinU eht dna ,muimerp ksir ecnairav eht ,daerps YH XDC eht ,daerps ngierevos naeporuE eht era stneicffieoc slobmysehT .redneldnaretrauqybderetsulc-elbuoderasrorredradnatS .sesehtnerapniscitsitats-t .stceffedexfiraey/epyt-redneldnaredneledulcnisnoisserger sa defiissalc snoitavresbo fo tnecrep eht wohs scitsitats-t eht ot stpircsrepuS .ylevitcepser ,slevel %01 dna ,5 ,1 eht ta ecnacfiingis lacitsitats etacidni ∗ ,∗∗ ,∗∗∗ eht neewteb daerps ehT .etar yrusaerT raey-01 drawrof-raey-eerht eht si etar yrusaerT drawroF .3102 ot 0102 sedulcni elpmas ehT .epyt rednel detacidni eht sdnufdefdetcepxenodesaberastniopsisab52sehcaeretarsdnufdefdetcepxeehtlitnusretrauqforebmunehtdnaetarsdnufdeftnerrucdnadaehasretrauq-01 .setouq spaws dexedni thginrevo morf deniatbo setar ot sretrauQ tnerruc/detcepxE drawroF ffotfil setar daerps sdnuf def etar yrusaerT .laB .labnU .laB .labnU .laB .labnU 051.0 010.0 ∗284.0- 500.0- ∗∗624.0- 960.0sCHB dna sknab .S.U 7.6)85.1( 5.3)62.0( 7.6)80.2-( 5.3)30.0-( 7.6)12.2-( 5.3)36.0-( 410.0- 510.0 391.0 151.0 840.0 230.0sCHB dna sknab .S.U-noN 0.8)82.0-( 6.3)13.0( 0.8)84.1( 6.3)98.0( 0.8)63.0( 6.3)02.0-( ∗432.0 ∗∗861.0 031.0- ∗503.0- ∗754.0- ∗∗653.0sdnuf noisneP/.soc ecnarusnI 7.4)39.1( 5.01)05.2( 7.4)95.0-( 5.01)97.1-( 7.4)79.1-( 5.01)32.2-( ∗∗∗852.0 ∗∗∗602.0 ∗∗∗045.0- ∗723.0- ∗∗∗537.0- ∗∗∗056.0sODC/sOLC .S.U 8.62)36.6( 6.21)65.3( 8.62)25.3-( 6.21)21.2-( 8.62)33.5-( 6.21)02.3-( ∗∗∗742.0 ∗∗∗051.0 ∗∗494.0- ∗∗∗194.0- ∗∗∗266.0- ∗∗∗875.0sODC/sOLC .S.U-noN 7.9)06.5( 5.9)72.3( 7.9)87.2-( 5.9)31.3-( 7.9)11.5-( 5.9)50.4-( ∗∗561.0 ∗∗861.0 200.0- 471.0- 792.0- ∗∗873.0srehto dna sdnuf .vni .S.U 1.63)62.2( 7.34)55.2( 1.63)10.0-( 7.34)30.1-( 1.63)93.1-( 7.34)92.2-( ∗∗∗382.0 ∗∗∗041.0 942.0- ∗332.0- ∗∗036.0- ∗∗∗733.0srehto dna sdnuf .vni .S.U-noN 0.8)37.3( 6.61)77.3( 0.8)39.0-( 6.61)67.1-( 0.8)96.2-( 6.61)22.3-( 487,4 789,93 487,4 789,93 487,4 789,93 snoitavresbO 83.0 04.0 63.0 04.0 73.0 04.0 2R .jdA 40

skcehc ssentsuboR :tekram naol detacidnys eht ni gnikat-ksiR :6 elbaT ehT .3noitceSnidessucsidnoissergerlenapehtgnisudetamitse,epytrednelyb,etaryrusaerTraey-01ehtotksirtidercoiloftropfoytivitisnesehtswohselbatehT naeporuEehterastneicffieocdetropernuhtiwsrossergeR .tnerapetaroprocehtfolevelehttasoiloftropnaolfotluafedfoytilibaborpgolehtsielbairavtnedneped dnaredneledulcnisnoissergerehT .noitaflnidetcepxefoxedninagihciMfoytisrevinUehtdna,muimerpksirecnairaveht,daerpsYHXDCeht,daerpsngierevos lacitsitats etacidni ∗ ,∗∗ ,∗∗∗ slobmys ehT .rednel dna retrauq yb deretsulc-elbuod era srorre dradnatS .sesehtnerap ni scitsitats-t .stceffe dexfi raey/epyt-rednel ehT .epyt rednel detacidni eht sa defiissalc snoitavresbo fo tnecrep eht wohs scitsitats-t eht ot stpircsrepuS .ylevitcepser ,slevel %01 dna ,5 ,1 eht ta ecnacfiingis na era slaudiser etar yrusaerT eht taht tcefler ot detsujda ton era egats dnoces eht ni srorre dradnats ,noitamitse egats-owt eht nI .3102 ot 0102 sedulcni elpmas .yraidemretni na fo oiloftrop eht evael ,retrauq nevig a ni ,taht snaol era snoitisopsiD .elbairav detamitse ,snoitiddA delacs sDP snoitiddA delacs sDP snoitiddA ,snoitiddA level-tnapicitrap sDP snoitisopsid yb sDP .trop .tstuo yb noitamitse egats-owt .laB .labnU .laB .labnU .laB .labnU .laB .labnU ∗783.0- 490.0- ∗∗358.0- 574.0- ∗∗355.0- 040.0- ∗∗295.0- 670.0sCHB dna sknab .S.U 7.7)60.2-( 1.4)06.0-( 8.6)45.2-( 5.3)84.1-( 9.6)54.2-( 5.3)32.0-( 7.6)44.2-( 5.3)35.0-( )level-tnapicitrap ta ylno sknab( 640.0- 763.0- ∗∗∗836.0- ∗∗∗247.0- 390.0 730.0 ∗∗493.0- ∗293.0sCHB dna sknab .S.U-noN 3.1)70.0-( 3.2)75.1-( 1.8)45.3-( 7.3)10.3-( 3.8)65.0( 9.3)51.0( 0.8)71.2-( 6.3)00.2-( ∗994.0- ∗∗704.0- 343.0- ∗∗∗049.0- 233.0- 451.0- ∗∗538.0- ∗∗∗436.0sdnuf noisneP/.soc ecnarusnI 3.5)00.2-( 1.01)02.2-( 4.4)06.0-( 4.01)05.3-( 2.4)98.0-( 3.01)95.0-( 7.4)48.2-( 5.01)81.3-( ∗∗∗238.0- ∗∗196.0- ∗∗271.1- ∗∗387.0- ∗∗∗607.0- ∗∗∗076.0- ∗∗∗330.1- ∗∗∗498.0sODC/sOLC .S.U 7.62)82.4-( 1.21)68.2-( 1.72)28.2-( 0.41)94.2-( 7.72)84.4-( 0.51)92.3-( 8.62)98.4-( 6.21)88.3-( ∗∗∗908.0- ∗∗∗086.0- ∗∗301.1- ∗∗179.0- ∗∗144.0- ∗∗∗156.0- ∗∗∗550.1- ∗∗∗959.0sODC/sOLC .S.U-noN 0.11)13.4-( 5.9)97.3-( 8.9)53.2-( 2.01)76.2-( 0.01)17.2-( 4.01)83.3-( 7.9)96.5-( 5.9)73.5-( 203.0- ∗683.0- ∗∗688.0- ∗∗037.0- 960.0- 392.0- ∗∗095.0- ∗∗∗575.0srehto dna sdnuf .vni .S.U 0.04)02.1-( 6.44)59.1-( 3.63)05.2-( 4.24)07.2-( 0.63)62.0-( 8.14)81.1-( 1.63)25.2-( 7.34)50.3-( ∗∗636.0- ∗∗∗293.0- 836.0- 944.0- 473.0- 372.0- ∗∗∗248.0- ∗∗∗955.0srehto dna sdnuf .vni .S.U-noN 0.8)71.2-( 3.61)81.3-( 5.7)64.1-( 6.51)26.1-( 9.6)50.1-( 0.51)06.1-( 0.8)70.3-( 6.61)24.4-( ∗∗237.0seinapmoc ecnaniF 0.1)14.2-( )ylno level-tnapicitrap( 008,4 747,14 146,4 222,92 426,4 879,13 487,4 789,93 snoitavresbO 33.0 04.0 61.0 70.0 02.0 02.0 13.0 83.0 2R .jdA 41

noitacfiissalc rednel evitanretlA :tekram naol detacidnys eht ni gnikat-ksiR :7 elbaT .3 noitceS ni dessucsid noisserger lenap eht gnisu detamitse ,epyt rednel yb ,etar yrusaerT raey-01 eht ot ksir tiderc oiloftrop fo ytivitisnes eht swohs elbat ehT eht era stneicffieoc detropernu htiw srossergeR .tnerap etaroproc eht fo level eht ta soiloftrop naol fo tluafed fo ytilibaborp gol eht si elbairav tnedneped ehT edulcni snoisserger ehT .noitaflni detcepxe fo xedni nagihciM fo ytisrevinU eht dna ,muimerp ksir ecnairav eht ,daerps YH XDC eht ,daerps ngierevos naeporuE etacidni ∗ ,∗∗ ,∗∗∗ slobmys ehT .rednel dna retrauq yb deretsulc-elbuod era srorre dradnatS .sesehtnerap ni scitsitats-t .stceffe dexfi raey/epyt-rednel dna rednel rednel detacidni eht sa defiissalc snoitavresbo fo tnecrep eht wohs scitsitats-t eht ot stpircsrepuS .ylevitcepser ,slevel %01 dna ,5 ,1 eht ta ecnacfiingis lacitsitats .A xidneppA ni dessucsid stnemngissa rednel launam eht no desab si noitacfiissalc evitanretla ehT .3102 ot 0102 sedulcni elpmas ehT .epyt ffotfil setaR daerps dnuf deF etar drawroF 2Q 3102 oN 2Q 2102 oN elpmas lluF ∗801.0 051.0- 322.0- 772.0- 292.0- ∗103.0sCHB dna sknaB 7.41)78.1( 7.41)97.0-( 7.41)37.1-( 7.41)56.1-( 6.41)17.1-( 7.41)38.1-( 890.0 901.0 033.0- 232.0- 761.0- 152.0slaicnanfi knabnoN 0.6)98.0( 0.6)64.0( 0.6)21.1-( 0.6)17.0-( 0.6)45.0-( 0.6)57.0-( ∗∗391.0 861.0- ∗∗804.0- ∗224.0- ∗152.0- ∗493.0sdnuf noisneP/.soc ecnarusnI 0.5)04.2( 0.5)69.0-( 0.5)83.2-( 0.5)58.1-( 0.5)88.1-( 0.5)49.1-( ∗∗151.0 920.0 091.0- 351.0- 520.0- 961.0sreganam tessa detneiro-liateR 1.12)43.2( 1.12)51.0( 1.12)00.1-( 1.12)95.0-( 9.02)51.0-( 1.12)67.0-( ∗∗∗962.0 ∗∗994.0- ∗∗∗147.0- ∗∗∗667.0- ∗∗∗917.0- ∗∗∗618.0sODC/sOLC 5.74)02.6( 5.74)29.2-( 5.74)50.5-( 5.74)84.3-( 8.74)09.3-( 5.74)61.4-( 981.0 932.0 631.0- 150.0- 522.0 850.0slaicnanfi noN 7.5)82.1( 7.5)97.0( 7.5)93.0-( 7.5)11.0-( 6.5)18.0( 7.5)51.0-( 487,4 487,4 487,4 584,4 515,4 487,4 snoitavresbO 83.0 63.0 73.0 53.0 04.0 73.0 2R .jdA 42

Table 8: Probabilities of default and loan spreads at origination by borrower rating The table shows summary statistics for the distribution of default probabilities and loan spreads over LIBOR at origination, by loan rating. The sample includes loan originations with PDs in the SNC database that we were able to match with Thomson Reuters DealScan data. Default prob. Loan spread percentile percentile Obs 10 50 90 10 50 90 Borrower rating Investment 0.14 0.26 1.00 100 150 275 37 Speculative 0.33 1.82 7.95 200 325 600 145 Not rated 0.16 0.80 3.89 150 275 475 527 43

Table 9: Determinants of loan spreads at origination The table shows regressions of loan spreads over LIBOR at origination on PDs, selected loan characteristics, and on severalmacroeconomicvariables. ThesampleincludeSNCloanswewereabletomatchtoThomsonReutersDealScan loans. Durationismeasuredindays,LBO indicateswhethertheloanisissuedtoprimarilyfinancealeveragedbuyout, No CUSIP is a dummy for loans that have no CUSIP in SNC, Public is a dummy equal to one if the borrower is a listed company, and Nonbank share is the share of the loan held by nonbank intermediaries. Standard errors are clustered by time. Regressions include borrower-industry/year and agent-bank/year fixed effects. dependent variable: log loan spread (1) (2) (3) log(PD) 0.200∗∗∗ 0.177∗∗∗ 0.138∗∗∗ (10.20) (9.40) (9.11) log(duration) 0.109∗∗∗ 0.033 (3.41) (1.06) log(loan amount) -0.030∗∗ -0.077∗∗∗ (-2.93) (-7.42) LBO 0.154 0.094 (1.27) (1.18) No CUSIP -0.096∗∗∗ -0.045 (-3.61) (-1.57) Public -0.197∗∗∗ -0.169∗∗∗ (-5.51) (-5.91) Default spread -0.012 -0.075 (-0.09) (-0.61) Variance risk premium 0.005∗∗ 0.006∗ (2.26) (2.07) CDX HY spread -0.000 0.001 (-0.17) (0.84) European sovereign spread -0.038 -0.078∗∗ (-1.24) (-2.31) Expected inflation 0.067 0.059 (1.26) (1.33) Treasury rate -0.073 -0.130∗∗∗ (-1.32) (-3.04) Nonbank share 0.543∗∗∗ (2.99) Nonbank share × Treas. rate 0.016 (0.23) Observations 709 709 709 Adj. R2 0.53 0.59 0.67 44