Capital Commitment and Illiquidity in Corporate Bonds

• Published date: 01 August 2018
• Author: WILLIAM MAXWELL,STACEY JACOBSEN,HENDRIK BESSEMBINDER,KUMAR VENKATARAMAN
• DOI: http://doi.org/10.1111/jofi.12694
• Date: 01 August 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 4 •AUGUST 2018

Capital Commitment and Illiquidity in Corporate

Bonds

HENDRIK BESSEMBINDER, STACEY JACOBSEN, WILLIAM MAXWELL,

and KUMAR VENKATARAMAN∗

ABSTRACT

We study trading costs and dealer behavior in U.S. corporate bond markets from 2006

to 2016. Despite a temporary spike during the financial crisis, average trade execution

costs have not increased notably over time. However, dealer capital commitment,

turnover,block trade frequency, and average trade size decreased during the financial

crisis and thereafter. These declines are attributable to bank-affiliated dealers, as

nonbank dealers have increased their market commitment. Our evidence indicates

that liquidity provision in the corporate bond markets is evolving away from the

commitment of bank-affiliated dealer capital to absorb customer imbalances, and

that postcrisis banking regulations likely contribute.

THE LIQUIDITY OF THE CORPORATE bond market has attracted substantial atten-

tion from practitioners, regulators, and academics in recent years. The financial

crisis of 2007 to 2009 saw the broad deterioration of liquidity in both equity

(e.g., Anand, Irvine, Puckett, and Venkataraman (2013)) and corporate bond

∗Hendrik Bessembinder is at the W.P. Carey School of Business, Arizona State University.Stacey

Jacobsen, William Maxwell, and Kumar Venkataraman are at the Cox School of Business, South-

ern Methodist University. We thank seminar participants at Arizona State University, Southern

Methodist University,Indiana University, Goethe-University Frankfurt, Board of Governors of the

Federal Reserve, Federal Reserve Bank of New York, Bank of Canada, University of Oklahoma,

University of Mannheim, University of New South Wales, University of Texas at San Antonio,

University of Sydney, Washington State University, University of Washington, Texas A&M Uni-

versity, Louisiana State University, University of Melbourne, and University of Zurich, as well as

conference participants at the 2017 Arizona State University Ph.D. Alumni conference, 2016 Lone

Star Conference, Notre Dame 2016 Conference on Current Topics in Financial Regulation, 2016

Banque De France-Toulouse School of Economics conference, 2017 Western Finance Association

conference, and University of Tennessee Smokey Mountain Finance Conference for their interest

and comments. Thanks are also due to Stefan Nagel, an anonymous Associate Editor, two anony-

mous referees, as well as Indraneel Chakraborty,Michael Goldstein, Steve Joachim, Elliot Levine,

Darius Miller, Michael Piwowar,Akhtar Siddique, Erik Sirri, Rex Thompson, James Weston, and

Steve Zamsky for their valuable comments. None of the authors received financial support specific

to this project. The authors thank the Finance Industry Regulatory Authority (FINRA) for provision

of the data and in particular, Alie Diagne, Ola Persson, and Jonathan Sokobin for their support

of the study. FINRA screened the paper to ensure that confidential dealer identities were not

revealed. Bessembinder, Maxwell, and Jacobsen have no conflicts of interest to report. Venkatara-

man is a visiting economist at the Office of Chief Economist at FINRA, and acknowledges financial

support for other projects.

DOI: 10.1111/jofi.12694

1615

1616 The Journal of Finance R

(e.g., Dick-Nielsen, Feldhutter, and Lando (2012), Friewald, Jankowitsch, and

Subrahmanyam, (2012)) markets. However, while equity market liquidity re-

covered after the financial crisis (Anand et al. (2013)), corporate bond market

liquidity has become a widespread concern in recent years. For example, in

2015 Daniel Gallagher, Commissioner of the U.S. Securities and Exchange

Commission (SEC), noted that “A lack of liquidity in corporate-bond markets

could pose a ‘systemic risk’ to the economy,”1and a 2016 Greenwich Associates

study reports that among 400 credit investors interviewed, more than 80%

indicated that reduced liquidity in corporate bonds limits their investment

strategies.2

Concerns regarding corporate bond market liquidity have been attributed

by some to postcrisis regulatory initiatives. For example, Pacific Investment

Management Company (PIMCO) asserts that “the combination of immediate-

post-crisis capital and liquidity regulations and a lower return environment has

made banks less able and willing to function as market makers.”3However, not

all observers are convinced that liquidity in the corporate bond markets has

deteriorated. Indeed, some argue that concerns about bond market illiquidity

comprise a “myth” and arise from traditional bond dealers’ desire to maintain

their “privileged market position.”4Janet Yellen, former chair of the U.S. Fed-

eral Reserve, has stated “It’s not clear whether there is or is not a problem”

(with liquidity), and added that “it’s a question that needs further study.”5

In this paper, we analyze liquidity and key aspects of dealer behavior in

the corporate bond market over the 2006 to 2016 period. We are particularly

interested in examining market quality in the years following the financial

crisis, and in evaluating potential explanations for the changes observed. To

do so, we use an enhanced version of the Trade Reporting and Compliance

Engine (TRACE) database of U.S. corporate bond transactions, made available

by Finance Industry Regulatory Authority (FINRA). In addition to the standard

TRACE data, the data that we study include masked dealer identities, which

allow us to directly assess activity at the dealer level, as well as unmasked

trade sizes and transactions in privately traded 144A bonds.

We document that, despite an increase during the financial crisis period,

average customer trade execution costs for corporate bonds have not increased

markedly over time. In particular, we find that the average one-way trade

execution cost during the 2014 to 2016 period averaged 0.42%, as compared to

0.40% during the 2006 to 2007 precrisis period.

1http://www.bloomberg.com/news/articles/2015-03-02/corporate-bond-market-poses-systemic-

risk-sec-s-gallagher-says.

2https://www.greenwich.com/press-release/2017-liquidity-starved-bond-investors-could-get-

relief-block-trading-solutions-and.

3http://www.barrons.com/articles/a-look-at-bond-market-liquidity-1440103954.

4“Overlooking the other sources of liquidity,” Wall Street Journal, July 26, 2015, available at

http://www.wsj.com/articles/overlooking-the-other-sources-of-liquidity-1437950015.

5http://blogs.wsj.com/economics/2015/07/15/fed-chairwoman-janet-yellens-report-to-congress-

live-blog/.

Capital Commitment and Illiquidity in Corporate Bonds 1617

However, average trading costs could mask shifts in the composition of trad-

ing. Investment-grade bonds and large-issue-size bonds, which tend to be more

liquid, each grew as a proportion of overall trading, as did trades that are most

likely to be facilitated by electronic venues. Further, execution costs for com-

pleted trades do not capture search costs or the implicit costs associated with

trades that were desired but not completed. We therefore consider a number of

additional measures, including dealers’ capital commitment measured at the

intraday, overnight, and weekly horizons; turnover; average trade size; block

trade frequency; and principal volume.

We use the term “capital commitment” to refer to dealers absorbing customer

order imbalances into their own inventories. Capital commitment is particu-

larly important to the functioning of markets where buyers and sellers arrive

sporadically and search costs are relatively high. Corporate bond trading oc-

curs largely in a telephone- and instant message–oriented dealer market with

limited pretrade transparency, and days or weeks can elapse between trades

in individual bonds. However, some corporate bond trading, particularly in

recently issued bonds of larger issue sizes, is electronically facilitated. While

a decline in dealer capital commitment may indicate a decrease in liquidity

in the overall market, a decline induced by growth in electronic trading may

reflect reduced search costs, which would imply an improvement rather than

deterioration of liquidity.

We find, not surprisingly, that all measures of dealer capital commitment de-

clined during the financial crisis. Potentially more informative, most measures

of dealer commitment for the overall market have not reverted to precrisis

levels. Indeed, many measures have continued to decline in recent years. All

of the measures that we consider point to significantly lower dealer capital

commitment in the recent 2014 to 2016 period, and many point to lower capital

commitment in the recent period than in the financial crisis period itself.

We consider several possible explanations for the observed decline in overall

dealer capital commitment in the recent period. First, postcrisis reforms in

bank regulation, such as the Volcker Rule and Basel III requirements, while

focused on banking rather than market-making activities, may have affected

dealers’ willingness or ability to commit capital to the provision of liquidity in

the corporate bond market. Second, the U.S. Treasury department estimates

that electronic platforms (the most important of which are “request for quo-

tation” systems rather than limit order books) have captured 15% or more

of customer-to-dealer market share in recent years, with the electronic share

higher for investment-grade than for high-yield bonds. Third, corporate bond

exchange-traded funds (ETFs) allow investors to gain exposure to corporate

bond returns without directly trading in the dealer market.6However, market

participants who create and redeem shares trade in the underlying, relatively

6Though bond ETFs have grown rapidly from 0.2% (net asset value to bonds outstanding) to

4.1% over our sample period, they remain substantially smaller than equity market ETFs. Bond

ETF information is available to academics from ICI.org.