Print
E-MailAlberto Manconi is Assistant Professor of Finance at Bocconi University; Ekaterina Neretina is a Ph.D. Student in Finance at Tilburg University and CentER; Luc Renneboog is Professor of Finance at Tilburg University and Research Associate at ECGI. This post is based on their recent paper.
The global bond market is a major source of corporate financing, and has been rapidly growing in recent years, reaching nearly $50 trillion in outstanding value as of 2013. At the same time, the bond underwriting industry has become increasingly more concentrated: In 2013, the ten largest underwriters had a combined market share of about 80%, up from 55% in 2000 and 30% in 1990. Industry practitioners as well as the press have raised concerns that this may give disproportionate power over the issuance process to a few large underwriters, enabling them to extract rents to the detriment of corporate bond issuers. In a recent paper, we address this concern and ask whether, and to what extent, underwriter power has an impact on corporate bond contracts.
The main challenge for any study of bargaining power is that it is unobservable, and often overlaps with the underwriter’s reputation and certification ability. For example, market share or past performance, which at first glance could seem reasonable proxies for bargaining power, could just as well stand for the certification value of the underwriter’s reputation as a signal of bond quality. To disentangle the impact of the bargaining power from certification, we develop a new measure of bargaining power, which relies on limited competition in the bond underwriting market and underwriters’ comparative ability to place bonds. The intuition behind the measure is straightforward. Consider, for instance, two bond issues by two firms Oracle and by AT&T Inc. both taken to the market by J.P. Morgan in 2012. On the first one, our measure indicates that only J.P. Morgan can be expected to successfully take the issue to the market, while on the second there are 10 potential replacements to J.P. Morgan, with Bank of America, Barclays, and UBS among the alternatives. This suggests higher bargaining power of J.P. Morgan over Oracle than AT&T, since Oracle has fewer “outside options” to take her bond to the market.
Building on this intuition, the key feature of our empirical approach is that our bargaining power index varies for a given underwriter at a given point in time across different issuers, allowing us to separate the effects of bargaining power from those of reputation and certification. In our example, whether J.P. Morgan is underwriting an issue by Oracle or AT&T, her reputation and certification ability remain the same. In contrast, our proxy allows its bargaining power to differ across the two issues. Comparing different issues that share the same underwriter at the same point in time, therefore, enables us to absorb the impact of certification and hence identify bargaining power effects. Consistent with our predictions and with the concerns among practitioners and in the press, we find that underwriters with high bargaining power charge higher fees and are associated with higher offering yields. In dollar terms, issuers pay additional issuing costs of $1.5m, or about 7% in comparison to the $22m cost they bear in terms of combined fees and yields on the typical bond in our data. These effects indicate that the costs associated with underwriter bargaining power are not trivial.
The nature of our index and our empirical strategy considerably raise the bar for any alternative explanation for our findings. Having said that, one possibility is that a given underwriter has different certification ability on different issues at the same point in time, e.g. because the issues belong to different bond market segments where she may have more or less expertise. If that were the case, the higher fees and yields might be compensation for certification. We show that this is not likely to be a concern in two ways. First, we find that returns on the issuer’s bonds outstanding drop more when powerful banks underwrite new bonds, consistent with the issuer incurring losses that can harm the position of the existing bondholders. The size of the drop is 64.4bps, or about 35% of the bond average yearly risk premium (185bps). Second, we show that effects on fees and yields do not appear stronger for issuers with higher certification needs (first time on the corporate bond market, low analyst coverage).
A second alternative explanation is “loyalty” between issuers and underwriters, referring to the stylized fact that issuers tend to stick to the same underwriters, possibly because they derive some benefits (observable or otherwise) from the relationship. We rule this out with two checks. First, we compare the effects of power on fees and yields on bonds issued just before or just after the issuer switches to a new underwriter. The estimated effects are similar. In contrast, if our findings were driven by the expected benefits of a future issuer-underwriter relationship, we should observe stronger effects on bonds issued with the new underwriter right after the switch. Second, we test whether a past relationship with the same underwriter on bond, equity, or M&A transactions absorbs the effects of our bargaining power measure. We find that is not the case, suggesting that loyalty is not driving our results.
Our findings document a “dark side” of corporate bond underwriting, where top underwriters are able to leverage their bargaining power to extract rents at the expense of bond issuers. They suggest that lack of competition in the underwriting industry can impose material costs on corporations.
The full paper can be downloaded here.
