Subprime governance: Agency costs in vertically integrated banks and the 2008 mortgage crisis

• DOI: http://doi.org/10.1002/smj.2481
• Author: Claudine Gartenberg,Lamar Pierce
• Published date: 01 February 2017
• Date: 01 February 2017

Strategic Management Journal

Strat. Mgmt. J.,38: 300–321 (2017)

Published online EarlyView 12 February 2016 in WileyOnline Library (wileyonlinelibrary.com) DOI: 10.1002/smj.2481

Received 11 May 2015;Final revision received14 October 2015

SUBPRIME GOVERNANCE: AGENCY COSTS IN

VERTICALLY INTEGRATED BANKS AND THE 2008

MORTGAGE CRISIS

CLAUDINE GARTENBERG1*and LAMAR PIERCE2

1Management and Organizations, NYU Stern School of Business, New York, New

York, U.S.A.

2Olin Business School, Washington University in St. Louis, St. Louis, Missouri,

U.S.A.

Research summary: This study uses the 2008 mortgage crisis to demonstrate how the relationship

between vertical integration and performance crucially depends on corporate governance. Prior

research has argued that the vertical integration of mortgage origination and securitization

aligned divisional incentives and improved lending quality. We show that vertical integration

improved loan performance only in those rms with strong corporate governance and that this

performance-integration relationship strongly decreases and actually reverses as governance

quality decreases. We interpret these ndings as suggesting that the additional control afforded

by vertical integration can, in the hands of poorly monitored managers,offset gains from aligned

divisional incentives. These ndings support the view that corporate governance inuences the

strategic outcomes of a rm, in our case, by inuencing the effectiveness of boundary decisions.

Managerial summary: One of the unanswered questions of the 2008 mortgage crisis is why

some rms produced toxic mortgages and others did not. Many have argued that vertically

integrated banks—banks that both originated and securitized mortgages —had incentives to

monitor themselves and thereby avoid overaggressive lending and outright fraud. Yet many

of the worst lenders, such as Washington Mutual and New Century Financial, were in fact

integrated. This study shows that the behavior of these rms critically depended on their corporate

governance. We nd that poorly monitored executives used their additional control over the

integrated businesses to issue low quality loans that supported short-term growth. Our results

suggest that governance is a crucial prerequisitefor nancial services, particularly for rms whose

managers control multiple, interrelatedbusinesses. Copyright © 2015 John Wiley & Sons, Ltd.

INTRODUCTION

In this article, we provide evidence that the rela-

tionship between vertical integration and perfor-

mance depends on corporate governance. We study

this role of governance in the context of one

Keywords: vertical integration; corporate governance;

transaction cost economics; mortgage securitization;

boundaries of the rm

*Correspondence to: Claudine Gartenberg, 40 West 4th Street,

Tisch 709 New York, NY 10012, U.S.A. E-mail: cgartenb@

stern.nyu.edu

Copyright © 2015 John Wiley & Sons, Ltd.

of the most important market failures in recent

history— the 2008 American housing crisis. Recent

research in corporate nance on the underlying

causes of the crisis has argued that the vertical

integration of mortgage origination and securiti-

zation aligned divisional incentives, and thereby

helped banks avoid severe lending quality problems

(Demiroglu and James, 2012; Purnanandam, 2011).

Yet, many of the worst performing lenders, such as

Washington Mutual, Wachovia, and New Century

Financial, were in fact integrated, having pursued

deliberate strategies to control the vertical chain.

These lenders subsequently collapsed spectacularly

Agency Costs in Vertically Integrated Banks and the 2008 Mortgage Crisis 301

once the housing market weakened in 2008. What

explains the poor performance by integrated rms?

We propose that weak corporate governance in

vertically integrated banks led to agency prob-

lems of the sort that are particularly pronounced

in information-intensive industries. Within these

rms, “soft” information— such as intangible bor-

rower risk factors— is passed internally between

divisions and is difcult for the outside market

to validate (Gartenberg, 2014; Pierce, 2012). A

deep literature in strategy and economics shows

that such information asymmetry makes aspects of

corporate governance such as executive compensa-

tion (Finkelstein and Hambrick, 1988; Harris and

Bromiley, 2007; Jensen and Murphy, 1990; Sanders

and Hambrick, 2007), board structure (Dalton et al.,

1998; Johnson, Hoskisson, and Hitt, 1993; West-

phal and Fredrickson, 2001), and investor compo-

sition (Hoskisson et al., 2002; Schnatterly, Shaw,

and Jennings, 2008) crucial for aligning man-

agerial actions with shareholder interests. With-

out sufcient monitoring by boards (Baysinger and

Hoskisson, 1990) or outside investors (Bushee,

1998; Thomsen and Pedersen, 2000), top managers

have wide discretion to either shirk responsibilities

or implement strategy and policies that enable them

to achieve compensation, status, and other personal

goals at the expense of shareholder value.

Vertical integration extends top managers’

span of control over the supply chain. Given this

increased control, we propose that weak corporate

governance allows two costly types of agency prob-

lems. First, managers can pursue self-interested

goals by distorting the activities of each division

as well as the terms of exchange and information

passed between them. Indeed, while recent work

suggests that vertical integration did in fact align

divisional incentives in banks (Demiroglu and

James, 2012), this alignment under weak gover-

nance may have served managerial rent-seeking

rather than shareholder value. Weak corporate gov-

ernance enables the top management team to struc-

ture compensation systems, reporting hierarchies,

and culture within the organization to support the

managers’ goals of myopic growth and excessive

risk (Werner, Tosi, and Gomez-Mejia, 2005). In this

way, the increased coordination, shared language,

and knowledge that is argued to produce benets by

some organizational scholars (Grant, 1996; Kogut

and Zander, 1992; Macher, 2006; Nahapiet and

Ghoshal, 1998), can produce the value-destroying

distortion highlighted by other scholars (Bidwell,

2012; Eccles and White, 1988; Nickerson and

Zenger, 2004; Osterloh and Frey, 2000).

Second, weak governance can also allow passive

CEOs or entire top management teams to insuf-

ciently monitor their organizations (Bertrand and

Mullainathan, 2003; Harris and Helfat, 1997; Hart,

1983), allowing self-interested managers inside the

organization to misrepresent, distort, and withhold

information for their own interests (Bidwell, 2012;

Eccles and White, 1988; Nickerson and Zenger,

2004; Osterloh and Frey, 2000; Pierce, 2012;

Shleifer and Vishny, 1997; Williamson, 1985).

As Williamson (1985) repeatedly argues, when

high-powered managerial incentives exist with

the rm, as is common in the banking industry,

the imperfect monitoring and intervention of top

managers and owners is frequently insufcient to

restrain this distortionary behavior.

We study how corporate governance changes the

role of vertical integration by examining the quality

of loans issued between 2000 and 2007 by mortgage

lenders that vary in both integration levels and gov-

ernance characteristics. We construct a rm-year

measure of lending quality as the incremental like-

lihood that a mortgage defaults if it is originated

by that lender, controlling for the loan characteris-

tics observable by external market participants. If a

lender chooses to supply its securitization unit by

unobservably lowering loan quality, this distortion

is captured by this metric.

We rst show that, on average, vertically inte-

grated lenders issue higher quality loans than

nonintegrated rms, replicating earlier research

(Demiroglu and James, 2012). We then show that

this average effect masks signicant differences

between integrated rms with strong and weak

governance. Although integrated rms with strong

governance have the lowest mortgage default like-

lihood, this relationship between integration and

default likelihood strongly increases as governance

weakens. In rms with the weakest governance,

greater integration is actually associated with

worse quality lending, the opposite of their

strong-governance counterparts. We also examine

specic governance dimensions and nd evidence

that both shareholder and board characteristics

moderate the relationship between integration and

performance, with inconsistent results on exec-

utive compensation. Our results suggest that the

advantages of vertical integration are offset in rms

with a weak governance structure, which is likely

a function of both external and internal monitoring.

Copyright © 2015 John Wiley & Sons, Ltd. Strat. Mgmt. J.,38: 300–321 (2017)

DOI: 10.1002/smj