The large increase in leveraged term loan originations in the syndicated lending market recently has called into question whether banks are sufficiently prepared for a significant downturn in economic activity stemming from problems in the corporate sector. Banks face risks from a variety of channels in this dynamic market, which has become more reliant on an originate-to-distribute model over the past two decades with the rapid growth of participation in these syndicated loans from institutional investors such as CLOs and mutual funds (Aramonte, Lee, and Stebunovs (forthcoming) and Irani, Iyer, Meisenzahl, and Peydro? (2018)). First, with some portion of the term loans remaining on banks’ balance sheets, banks are susceptible to losses from increases in loan defaults and mark-to-market losses from declines in the secondary loan prices. 2 Second, during the loan syndication process-that is, before a loan is originated-banks are exposed to sudden insufficient demand by other nonbank investors for these loans. In such cases, banks are forced to retain large share of these loans on their balance sheets. 3 This so-called “pipeline risk,” especially as it relates to committed funds prior to launching a deal, has been well documented in the industry and in the literature (Bruche, Malherbe, and Meisenzahl (2017)). This note looks carefully at the transition of ownership of syndicated term loans immediately after a deal is launched based on the Shared National Credit data. 4 We shed further light on who eventually ends up investing in these loans and the consequences of the materialization of pipeline risk during the origination process forcing banks to initially hold larger loan shares. 5

Figure 1: Spread distribution for SNC loans originated post-2008 by rating category

We first link the Shared National Credit data to ratings and pricing data from supervisory data and proprietary data using the same methods used in Calem, Correa, and Lee (forthcoming) and Cohen et al. (2018). We can then categorize loans by their risk profiles. We use the internal borrower ratings of the administrative agent (“agent”)-the bank in charge of collecting and dispersing interest payments and relevant information. Figure 1 shows that the ratings and loan spreads are highly correlated. In the left panel of Figure 2, we then categorize loans according to the ratings. If we define leveraged loans as noninvestment-grade loans (BB or lower), the total leveraged loans outstanding is about $1.2 trillion as of the end of 2018. The large increase over the past atic increase in B-rated loans. In the right panel, we categorize loans according to the pricing of the loan. 6 Following industry norm, we also classify a loan to be a “highly leveraged loan” if the issuance spread is 225 basis points or more above LIBOR. By this definition, the outstanding amount http://www.americashpaydayloan.com/pawn-shops-sd/ of leveraged loans is slightly lower. 7 In both cases, the leveraged segment appears to have grown dramatically in the past decade.

Figure 2: Term loans outstanding by risk characteristic

Figure 3 shows types of lenders for loans of various riskiness using data as of the end of 2018. 8 Loans that are either above investment grade or have spreads lower than 225 basis points, which are on average larger in loan size, are primarily held by banks. In contrast, further down the risk profile, loans with ratings of B or spreads between 325 and 425 basis points are primarily held by CLOs. 9 Mutual funds, on the other hand, appear to invest more evenly across different types of leveraged loans. Finally, we can see alternative lenders, such as hedge funds, investing in some of the highest risk/spread loans. These types of holding patterns have been relatively consistent throughout the past decade (not shown).