"Skip to main contentSkip to article
      Journal of Financial Economics
  Volume 137, Issue 2, August 2020, Pages 470-490
  Institutional allocations in the primary market for corporate bonds☆
  Author links open overlay panelStanislava Nikolova, Liying Wang, Juan (Julie) Wu
  Show more
  Share
  Cite
  https://doi.org/10.1016/j.jfineco.2020.02.007
  Get rights and content
  Abstract
  Using 2002–2014 insurer transactions, we provide the first empirical evidence on underwriters' allocation practices in the primary market for corporate bonds. Since bonds are often underpriced, allocations generate for investors an estimated $41 billion of   first-day profits. These profits increase with proxies for investors' information production during the bookbuilding process and, more strongly, with investors' prior trading with underwriters. Information production has a larger impact when asymmetric information   is higher, while prior trading has a larger impact when the issuer-underwriter agency problem is more severe. When there is more competition for allocations, prior trading further increases an insurer's first-day profits.  
  Introduction
  Following the 2007–2009 financial crisis, US corporate bond offerings have become increasingly oversubscribed yet remain underpriced. Although the average underpricing per offering is low, because bond offerings are frequent and large, the first-day profits   to primary market investors can be substantial. During our 2002–2014 sample period, we estimate these profits to be $41 billion. Since newly issued bonds account for 17.5% of the corporate bond market's capitalization in this time period, access to first-day   profits from underpriced offerings can significantly boost corporate bond investors' performance.2  
  While prior studies have documented the existence and determinants of underpricing in corporate bond offerings, there is surprisingly little research on how first-day profits from underpriced offerings are allocated among potential primary market investors.3   In contrast, a large body of literature examines underwriter allocation practices in the market for initial public offerings (IPOs) of equity.4 Researchers' focus on equity IPOs to the exclusion of bond offerings is surprising for at least three reasons. First,   issuance activity in the corporate bond market is significantly larger than that in the equity market. During our 2002–2014 sample period, underwriters placed $8.757 trillion of corporate bond offerings compared to only $0.745 trillion of equity IPOs and $1.807   trillion of seasoned equity offerings.5 Second, investors have increasingly complained about limited access to the primary market for corporate bonds, claiming that "allocations always come down to favours," which has triggered a Securities and Exchange Commission   (SEC) investigation into the allocation practices of some of the largest corporate bond underwriters.6 Finally, the lack of pre-offering public information about private firms and the fact that equity IPOs are a one-time event in a firm's history have made   it difficult for equity IPO studies to empirically investigate the cross-sectional implications of theories of underwriter favoritism (e.g., Jenkinson, Ljungqvist, 2001, Lowry, Michaely, Volkova, 2017). The richness of corporate bond issuers' data and the   differences between stock and bond markets present a unique opportunity to overcome these data limitations.  
  In this study we provide the first empirical evidence on corporate bond underwriters' allocation practices. Using 2002–2014 data from insurers' regulatory filings with the National Association of Insurance Commissioners (NAIC), we shed light on the determinants   of profitable allocations in a long time-series and a broad cross-section of 5,341 investment-grade (IG) offerings brought to the market by 76 lead underwriters. Since regulations in the US do not require public disclosure of primary market allocations, prior   studies of US equity IPOs either use small samples of proprietary data or infer allocations from institutional quarterly holdings disclosures.7 Instead, we use a novel approach that relies on the detailed information about each trade in the NAIC data to determine   the primary market allocation of a corporate bond offering to an insurer. Specifically, we identify as an allocation an insurer's purchase of a bond on the offering date at the offering price from the offering's underwriters.8  
  We analyze insurers' corporate bond allocations to understand why underwriters favor certain investors with larger first-day profits. Equity IPO studies argue that this happens mainly for reasons related to asymmetric information, as in bookbuilding theories,   and agency problems, as in the profit-sharing view.9 Under bookbuilding theories, underwriters collect pre-market demand information from investors during the bookbuilding process. Because underwriters both set the offering price and determine allocations,   they can reward investors' information production with larger first-day profits (e.g., Benveniste, Spindt, 1989, Benveniste, Wilhelm, 1990, Spatt, Srivastava, 1991, Sherman, Titman, 2002). Under the profit-sharing view, underwriters underprice the offering   more than necessary and then distribute the resultant first-day profits to their best clients (Loughran and Ritter, 2002).  
  Our analyses of corporate bond allocations to insurers produce several new findings. First, insurers purchase a significant amount of newly issued bonds in the secondary market and do so at a price higher than the price they could have paid had they been allocated   the bonds in the primary market. On average, insurers acquire 18% of the par value of IG bonds through primary market allocations but then obtain an additional 12% (or 8% if we exclude insurers already allocated the bonds in the primary market) through secondary   market purchases within 90 days of the offering. This comes at a significant cost, as the price paid by insurers in the secondary market is higher than that paid in the primary market. On average, the secondary market price within 90 days of the offering is   31 basis points (bps) higher than the offering price.  
  Second, we show that both information production and a trading relationship with an offering's underwriters are rewarded with more profitable allocations, though the economic impact of the latter is significantly larger. In particular, insurers with more expertise   in bonds similar to the one being offered, measured as the percent of their prior-year holdings in the same industry, receive more of the offering's first-day profits. To the extent that an insurer's industry expertise is related to its information contribution   during the bookbuilding process, this finding suggests a link between information production and profitable allocations. Even more importantly, insurers with a stronger trading relationship with an offering's underwriters, measured as the percent of the lead   underwriters' prior-year trading volume that comes from the insurers, receive more profitable allocations. In terms of economic significance, a one standard deviation increase in an insurer's prior trading with the underwriters increases its average first-day   profits by an estimated $0.8 million per year. This impact is five times that of a one standard deviation increase in our proxy for information production. Unobservable time-invariant insurer characteristics are not responsible for our findings, which are   robust to the inclusion of insurer fixed effects. Excluding the most active insurer traders, the largest bond offerings, or offerings brought to the market on a compressed timeline also leaves our conclusions unchanged. Furthermore, underwriter-by-underwriter   regressions reveal considerable consistency in the importance of prior trading for larger first-day profits across the biggest corporate bond underwriters.  
  Third, we provide evidence of cross-sectional variation in the association between investors' first-day profits and their information production and trading relationship with underwriters. This has been difficult to do for equity IPOs because finding convincing   ex ante proxies for asymmetric information and agency problems in that setting has been challenging. Using corporate bond data, we are able to show that in offerings characterized by more asymmetric information between underwriters and investors, the importance   of information production for obtaining more profitable allocations increases. When underwriters lack underwriting experience in the industry of the new offering, or when they are not core dealers in the secondary market, they allocate more of the first-day   profits to investors who produce more information. The relation between information production and profitable allocations is also stronger when the issuer's existing bonds are less actively traded or are traded at widely divergent prices and when the issuer's   equity is private. We also show that investors' trading relationship with an offering's underwriters contributes less to first-day profits when the agency problem between issuers and underwriters is less severe. When an offering is of a frequent corporate   bond issuer, the association between investors' prior trading and profitable allocations is weaker. This suggests that underwriters' desire to win corporate bond issuers' repeated underwriting business may temper the severity of the issuer-underwriter agency   problem.  
  Finally, we show that when competition for allocations is high, insurers can secure even larger first-day profits through a stronger trading relationship with the underwriters. We use two approaches to identify offerings with high competition for allocations.   The first approach focuses on offerings that help insurers "reach for yield," which Becker and Ivashina (2015) show to be in high demand by insurers because of capital regulations. The second approach assumes that offerings characterized by larger secondary   market purchases soon after issuance are more sought after by insurers in the primary market. We show that in offerings with high competition for allocations, an insurer's trading relationship with the underwriters further increases its first-day profits,   while more information production does not. This suggests that, when competition for corporate bond allocations is high, underwriters prioritize even more the interests of their most important clients over those of other primary market investors.  
  Several reasons might explain why, in our setting, information production appears as a less important determinant of profitable allocations than a trading relationship with the underwriters. We focus on newly issued bonds that carry a credit rating, which is   an independent third-party assessment of the bond's risk. In addition, 72% of bonds in our sample are issued by public firms with equity market information already available, and 96% are issued by repeat borrowers in the public debt markets. These features   of the corporate bond offerings in our sample likely make information production by primary market investors less important to underwriters and less rewarded through profitable allocations. This contrasts with the importance of information production in the   underwriting process for equity IPOs, which are characterized by high levels of information asymmetry (e.g., Lowry, Schwert, 2004, Hoberg, 2007, Hanley, Hoberg, 2010, Lowry, Officer, Schwert, 2010).  
  By providing the first large-sample evidence on institutional allocations in corporate bond offerings, our study contributes to the literature in two important ways. It adds to research on whether and why underwriters systematically favor some primary market   investors. Equity IPO studies argue that underwriter favoritism is related to asymmetric information and agency problems. We show that the same two drivers are at play in the corporate bond market, though the latter appears to dominate. Furthermore, exploiting   the richness of corporate bond issuers' data and the differences between equity and debt markets, we are able to overcome some of the data limitations of equity IPO studies and provide new evidence on the cross-sectional implications of the bookbuilding and   profit-sharing theories of underwriter favoritism.  
  We also add to the literature on the benefits of a trading relationship in the corporate bond market. O'Hara et al. (2018) examine the execution quality of secondary market corporate bond trades and find that insurers receive better prices when they have a   stronger trading relationship with bond dealers. Hendershott et al. (2017) show that large insurers have larger trading networks and thus pay lower transaction costs in the secondary bond market. We complement these studies by showing that the trading relationship   between investors and underwriters benefits investors in the primary market as well through more profitable allocations.  
  Section snippets
  Institutional background and hypothesis development
  In this section, we first describe the underwriting process for corporate bonds. We then develop our two hypotheses as to why underwriters choose to allocate certain offerings to some investors and not others.  
  Data
  In this section, we describe our data sources, main variable definitions, and sample construction. We then present sample summary statistics.  
  Primary market allocations versus secondary market purchases
  To better understand insurers' buying activity in newly issued IG bonds, we begin our analysis with a univariate comparison of insurers' primary market allocations and secondary market purchases. To do so, we identify in the NAIC data any purchases that do   not meet our definition of a primary market allocation and are made within 30, 60, or 90 days of a bond's offering date. For each offering, we then aggregate the par value bought and average the price paid, separately for primary market  
  Determinants of first-day profits
  To shed light on why some insurers are able to access newly issued bonds through the primary market while others are not, in this section we investigate the drivers of corporate bond underwriters' allocation practices. In particular, we examine whether an insurer's   first-day profits from an offering are related to that insurer's potential information production and trading relationship with the offering's underwriters by testing the hypotheses detailed in Section 2.25  
  First-day profits and underwriter-investor asymmetric information
  Our results so far suggest that for the average corporate bond offering, the effect of investors' trading relationship on profitable allocations is empirically stronger than that of their information production. This may reflect the fact that the information   investors provide during the bookbuilding process cannot be directly observed and as a result may be more problematic to measure than their trading relationship with the underwriters. However, if our proxy does reflect information  
  First-day profits and the issuer-underwriter agency problem
  In this section we investigate the potential cross-sectional differences in the impact of trading relationship on first-day profits. While equity IPOs are a one-time event in a firm's history, many firms issue bonds repeatedly over time.36  
  First-day profits and competition for allocations
  When an offering is oversubscribed and more investors compete for an allocation, underwriters have more discretion when distributing the first-day profits from the offering among primary market investors. We expect that our information production and trading   relationship proxies should have an even stronger impact on investors' first-day profits when competition for allocations is high.  
  We use two approaches to classify offerings as characterized by high competition for allocations among  
  Individual underwriter analysis
  In this section, we investigate whether our finding that an insurer's trading relationship with an offering's underwriters dominates its information production as a determinant of first-day profits is consistent across underwriters. We focus on the ten underwriters   with the largest number of offerings in the sample for two reasons. First, for each of these ten underwriters, we have a sufficient number of observations to reliably estimate the marginal effect of InfoProd and TrdRel. Second, since    
  Conclusion
  In this paper we provide the first empirical evidence on the allocation practices of corporate bond underwriters. We use a comprehensive data set of insurer trades to identify insurers' primary market allocations of investment-grade corporate bonds issued during   the 2002–2014 period. Since these bonds are underpriced by an average of 32 bps, being allocated the bonds in the primary market provides insurers with large first-day profits. We explore the cross-sectional variation within insurers to    
  References (52)
  A.E. Sherman et al.
  Building the IPO order book: underpricing and participation limits with costly information
  J. Financ. Econ.
  (2002)
  J.R. Ritter et al.
  Affiliated mutual funds and the allocation of initial public offerings
  J. Financ. Econ.
  (2007)
  M. O'Hara et al.
  The execution quality of corporate bonds
  J. Financ. Econ.
  (2018)
  M. Nimalendran et al.
  Do today's trades affect tomorrow's IPO allocations?
  J. Financ. Econ.
  (2007)
  C.J. Muscarella et al.
  A simple test of Baron's model of IPO underpricing
  J. Financ. Econ.
  (1989)
  M. Lowry et al.
  Is the IPO pricing process efficient?
  J. Financ. Econ.
  (2004)
  A.P. Ljungqvist et al.
  IPO allocations: discriminatory or discretionary?
  J. Financ. Econ.
  (2002)
  X. Liu et al.
  Local underwriter oligopolies and IPO underpricing
  J. Financ. Econ.
  (2011)
  Q. Hao
  Laddering in initial public offerings
  J. Financ. Econ.
  (2007)
  K.W. Hanley et al.
  Evidence on the strategic allocation of initial public offerings
  J. Financ. Econ.
  (1995)
  View more references
  Cited by (14)
  Overallocation and secondary market outcomes in corporate bond offerings
  2022, Journal of Financial Economics
  Citation Excerpt :
  Wang (2020) studies how underwriters use the information gained during the book building process to adjust offer prices. Using NAIC data that identifies those primary investors that are also insurance companies, Nagler and Ottonello (2021), Nikolova et al.   (2020), and Flanagan et al. (2019) document that the syndicate is strategic in its primary market allocations. Nikolova and Wang (2022) also use the NAIC data, documenting that flipping by insurance companies is not only common, but is highly variable across   offers.  
  Show abstract
  Lifting the veil: The price formation of corporate bond offerings
  2021, Journal of Financial Economics
  Citation Excerpt :
  Brugler et al. (2020) find underpricing decreases after mandated post-trade transparency in secondary markets. Nikolova et al. (2020) study the allocation of a sample of IG bonds to insurance companies. To proxy for investors' information production, they use   insurers' percentage holding of CBs in the same industry as the new offering.  
  Show abstract
  Underpricing in the euro area bond market: New evidence from post-crisis regulation and quantitative easing
  2021, Journal of Financial Intermediation
  Citation Excerpt :
  Our results support the predictions obtained from bookbuilding-based theories of underpricing (e.g. Benveniste and Spindt, 1989) but are inconsistent with liquidity-based explanations. We thus confirm Cai et al. (2007), Brugler et al. (2018) and Nikolova et   al. (2020) who also favor information-based explanations of bond IPO underpricing. Third, we document that the underpricing of euro area bonds has increased considerably during the financial crisis and has remained above the pre-crisis level since, thereby   confirming the results Nagler and Ottonello (2019) obtained for the US market.  
  Show abstract
  The Bond-Pricing Implications of Rating-Based Capital Requirements
  2022, Journal of Financial and Quantitative Analysis
  The perception of corporate innovation and credit spreads in emerging markets: evidence from China
  2023, Technology Analysis and Strategic Management
  Inventory-Constrained Underwriters and Corporate Bond Offerings
  2022, Review of Asset Pricing Studies
  View all citing articles on Scopus
  Recommended articles (6)
  Research article
  Concentration of control rights in leveraged loan syndicates
  Journal of Financial Economics, Volume 137, Issue 1, 2020, pp. 249-271
  Show abstract
  Research article
  Heterogeneous beliefs and return volatility around seasoned equity offerings
  Journal of Financial Economics, Volume 137, Issue 2, 2020, pp. 571-589
  Show abstract
  Research article
  The importance of being special: Repo markets during the crisis
  Journal of Financial Economics, Volume 137, Issue 2, 2020, pp. 392-429
  Show abstract
  Research article
  Prime (information) brokerage
  Journal of Financial Economics, Volume 137, Issue 2, 2020, pp. 371-391
  Show abstract
  Research article
  The Volcker Rule and corporate bond market making in times of stress
  Journal of Financial Economics, Volume 130, Issue 1, 2018, pp. 95-113
  Show abstract
  Research article
  Corporate bond mutual funds and asset fire sales
  Journal of Financial Economics, Volume 138, Issue 2, 2020, pp. 432-457
  Show abstract
  ☆
  We thank William Schwert (the editor), an anonymous referee, Vladimir Atanasov, Jennie Bai, Bo Becker, Allen Berger, Carole Comerton-Forde, Mark Flannery, Ioannis Floros, Carole Gresse, Kathleen Hanley, Jean Helwege, Jerry Hoberg, Jan Jindra, Alexander Ng,   Christo Pirinsky, Jay Ritter, Michael Schwert, Pei Shao, Yoshiki Shimizu, Yuehua Tang, Donghang Zhang, and participants in the 2018 Fixed Income and Financial Institutions Conference, 2018 FMA Conference, 2019 Women in Microstructure Meeting, 2019 FMA European   Conference, 2019 NFA Conference, and seminars at the Australian National University, American University, Babson College, Ludwig-Maximilians-Universität München, Securities and Exchange Commission, University of Florida, University of Melbourne, University   of Nebraska-Lincoln, University of New South Wales, University of South Carolina, University of Wisconsin-Milwaukee, and Wichita State University for insightful feedback that has substantially improved this paper.  
  View full text
  © 2020 Elsevier B.V. All rights reserved.  
  About ScienceDirect
  Remote access
  Shopping cart
  Advertise
  Contact and support
  Terms and conditions
  Privacy policy
  We use cookies to help provide and enhance our service and tailor content and ads. By continuing you agree to the use of cookies.  
  Copyright © 2023 Elsevier B.V. or its licensors or contributors. ScienceDirect® is a registered trademark of Elsevier B.V.    
  "
  https://www.sciencedirect.com/science/article/abs/pii/S0304405X20300404#:~:text=Skip%20to%20main%20content,of%20Elsevier%20B.V.