Financial Experti of Directors*
Tate‡
A. Burak Güner Ulrike Malmendier† Geoffrey
UC Berkeley and NBER UC Los Angeles Barclays Global Investors,
San Francisco
March 16, 2007
Journal of Financial Economics, forthcoming
Abstract
The composition and functioning of corporate boards is at the core of the academic and policy debate on optimal corporate governance. But does board composition matter for corporate decisions? In this paper, we analyze the role of financial experts on boards. In a novel panel data t on board composition, we find that financial experts significantly af-fect corporate decisions, though not necessar
ily in the interest of shareholders. First, when commercial bankers join boards, external funding increas and investment-cash flow nsitivity diminishes. But, the incread financing affects mostly firms with good credit and poor investment opportunities. Second, investment bankers on the board are associated with larger bond issues, but also wor acquisitions. Third, we find little evi-dence that financial experti matters for compensation policy or for experts without af-filiation to a financial institution. The results suggest a tradeoff between outside incen-tives (e.g. bank profits) and the incentive to maximize firm value. Requiring financial ex-perti on boards, as mandated by regulatory proposals, may not benefit shareholders if conflicting interests are neglected.
* We thank Alexander Ljungqvist, Ernst Maug, Stefan Nagel and Philip Strahan for helpful comments. We also thank the minar audiences at the NBER Summer Institute Corporate Governance meeting, WFA, Fall 2004 NBER Corporate Finance program meeting, at the University of Pennsylvania, Stanford, Cornell, University of Notre Dame, UC Irvine, Emory University, UT Austin, Columbia University, USC and Uni-versity of North Carolina. Justin Fernandez, George Georgiev, Vaibhav Gupta, Nicole Hammer, Rumit Kanakia, Jared Katff, Catherine Leung, Chenkay Li, Scott McLin, Felix Momn, Neeraj Shah, Ning Xu, and Nancy Yu provided excellent rearch assistance. Earlier versions of this paper were titled “The Impact of Boards with Financial Experti on Corporate Policies.”
† Department of Economics, 549 Evans Hall #3880, Berkeley CA 94720-3880, ph.: 510-642-0822, fax: 510-642-6615, email: ulrike@econ.berkeley.edu.
大学生学期总结‡ Entrepreneurs Hall C420, UCLA Anderson School of Management, Box 951481, Los Angeles, CA 90095-1481, ph: 310-825-3544, fax: 310-206-5455, email: geoff.tate@anderson.ucla.edu.
Much of the recent corporate-governance debate revolves around the composition of cor-porate boards. The question is which types of directors can be expected to rve the inter-ests of shareholders. Following the recent wave of accounting scandals, regulators have stresd the need for more financial experts on boards. The implicit assumption is that “an understanding of generally accepted accounting principles and financial statements” will lead to better board oversight and rve the interest of shareholders.1
Financial experts, however, might affect firm policies beyond more accurate dis-closure and better audit committee performance. Directors spend a significant portion of their time on advising rather than monitoring (Adams and Ferreira (2003)). This influ-ence can be problematic if directors are affiliated with financial institutions and pursue the interests of tho institutions rather than maximizing shareholder value. Affiliation has raid concerns in veral areas of financial intermedia
tion, such as analyst recom-mendations and IPO allocations.2 Nevertheless, recent regulatory efforts to increa fi-nancial experti on boards do not preclude such conflicts. As Sarbanes-Oxley (SOX) en-acts a very broad definition of financial experti,3 bankers remain a common type of fi-nancial expert on corporate boards.4
In this paper, we ask whether directors with financial experti influence corpo-rate policies and whether affiliation hampers their advisory role. We analyze both internal investment (capital expenditure) and external investment (mergers and acquisitions), us-ing a novel data t on the board composition of 282 companies over 14 years.
Our analysis proceeds in three steps. First, we study the impact of financial ex-perts on internal investment and on the financing of investment with bank loans. We ex-amine all financial experts, but focus on commercial bankers. One hypothesis is that, if
1 See Section 407 of the Sarbanes-Oxley (2002) Act (definition of audit committee financial experts). Simi-larly, all major stock exchanges have introduced listing requirements on director financial literacy. See also the governance survey by Jenn, Murphy, and Wruck (2004), e.g. recommendation R-36.
2 Malmendier and Shanthikumar (2007); Reuter (2005).
鞍山玉佛苑3 The original SOX proposal had a more restrictive definition of financial expert, which would have only included CPAs and people with direct accounting experience. The objection from the business community (e.g. “Blue Ribbon Commissions”), which led to the softening of the definition, was that the pool of experts is too small, especially if bankers were to be excluded. (See Tenorio, 2003; Stuart, 2005). In our sample, accountants make up only 0.5% of directors. CFOs make up 1%.
凉菜菜谱
4 According to CFO Magazine’s analysis (Stuart, 2005), roughly 30% of boards (among the Fortune 100) identified a director who is “an evaluator of financial statements, such as a banker or investor,” an execu-tive who merely supervid the finance or accounting function, or a director with “no discernible profes-sional experience in finance” as their financial expert for the purpo of compliance with SOX.
firms are financially constrained due to information asymmetries (Myers and Majluf (1984); Fazzari, Hubbard, and Petern (1988)), bankers enable firms to finance value-creating projects they would otherwi forgo. In that ca, bankers decrea the nsitiv-ity of investment to the availability of internal funds.5 An alternative hypothesis is that bankers provide loans even when it is not in the inte
rest of shareholders. Lending to un-constrained firms with low default risk – even if they have no value-creating projects – may increa bank profits. And it may enable empire-building or overconfident managers (Jenn and Meckling (1976); Jenn (1986); Malmendier and Tate (2005)) to divert funds or to overinvest. Such arrangements are more likely if the lending bank is repre-nted on the board so that the board is unlikely to intervene. In this ca, investment-cash flow nsitivity also decreas, but mostly in financially unconstrained firms.
We test the hypothes empirically. We find that when commercial bankers join the board of a firm, the firm displays less investment-cash flow nsitivity and obtains larger loans. The effect depends on affiliation, i.e., directors who banks have a lending relationship with the firm. Moreover, lending increas only for firms that are least finan-cially constrained, such as firms with investment-grade debt. The firms also appear to have wor investment opportunities and lower profitability. Constrained firms receive no such assistance. Thus, the additional lending appears to benefit creditors rather than shareholders. We find no measurable impact of other, unconflicted financial experts.
Second, we consider external investment (i.e. mergers and acquisitions) and fi-nancing with public curities. Here, we focus on investment bankers. We find that firms with investment bankers on thei
r boards undertake wor acquisitions. In the 5 days around takeover bids, they lo 1% more than firms without investment banker directors. They also lo significantly more value over the three years following an acquisition. Firms with investment bankers on their board are also associated with larger bond issues, in particular if the director’s bank is involved in the deal. And, while investment bankers on the board generally em to reduce underwriting fees, this helping hand is not visible when their own bank is involved in the deal. Thus, like commercial bankers, investment bankers have a significant impact on corporate decisions, but not necessarily in the inter-
我的爱金枝玉叶5 Consistent with this story, Hoshi, Kashyap, and Scharfstein (1991) find that investment is less nsitive to cash flow in Japane firms with keiretsu membership. Ramirez (1995) finds that firms with J.P. Morgan executives on their boards displayed lower investment-cash flow nsitivity at the turn of the 20th century.
est of shareholders. Again there is little evidence that non-conflicted financial experts improve firm policies (or have any impact at all).
Third, we test for the influence of financial experts on decisions where their out-side incentives are not in conflict with the interests of shareholders, such as CEO com-pensation.6 We find little evidenc
太空安全e that financial experts affect decisions without clear conquences for their home institutions.
Our results challenge the view that more financial experti on corporate boards unambiguously improves firm policy. Rather, the benefits have to be weighted against the costs due to misaligned incentives. Experts are associated with policies that may create value for their financial institutions, but we find little evidence of benefits to sharehold-ers. Our findings do not imply that financial experts destroy shareholder value on net. Companies may benefit from financial experti on their board through veral other avenues. Aggrawal and Chadha (2003), for example, find that having directors with a CPA, CFA, or similar degree on audit committees translates into fewer earnings restate-ments. DeFond, Hann, and Hu (2005) document a positive stock market reaction to the appointment of directors with accounting knowledge to the audit committee (though not to the appointment of other financial experts). Moreover, our data consists of large, ma-ture US firms; small early-stage firms, for example, may benefit from the financial exper-ti of venture capitalists (Hellman and Puri (2000, 2002), Kortum and Lerner (2000)).
A key concern for any analysis of director effects is the endogeneity of board composition, a point made both theoretically and empirically by Hermalin and Weisbach (1998 and 1988), among others.7 In particular, the causality may be rever, and firms’ financing needs may determine the bo
ard reprentation of financial institutions.8 Our de-tailed data allows us to better address the concerns for veral reasons.
First, the fourteen-year time ries provides sufficient variation in board composi-tion to identify commercial banker effects even after controlling for company fixed ef-fects. Thus, the estimated effect does not reflect time-invariant firm characteristics.
6 For a discussion of the role of directors in tting compensation and of financial experti (e.g. in under-standing the value of option grants which are not expend) e Holmström and Kaplan (2003).
7 For an extensive review of the literature e Hermalin and Weisbach (2003).
8 Stearns and Mizruchi (1993), Pfeffer (1992), and Booth and Deli (1999) interpret the correlation between firm leverage and board prence of bankers as evidence of firms hiring financial directors for their debt market experti.
Second, we are able to instrument for the board prence of commercial bankers, using pre-sample shocks to the supply of banker directors. During the banking crisis in the late 1970s and early 1980s,
executives of commercial banks became less suitable candidates for corporate boards. As a result, open board positions between 1976 and 1985 were less likely to be filled with commercial bankers than in other decades, while the overall rate of board appointments remained the same. We instrument for the number of commercial bankers on the board with the number of current directors hired during the crisis period. All results replicate. A placebo instrument, the number of directors ap-pointed between 1966 and 1975, fails to replicate the results, corroborating our analysis.
Finally, we can identify and remove company-year obrvations in which lec-tion concerns are most vere, such as years with major acquisitions and the first years of a banker’s tenure. We show that such years do not drive our results.
For other financial experts, we cannot u fixed effects in most cas. There is, for example, insufficient within-firm variation in acquisitions. We also do not have an in-strumental variable strategy. As in previous literature, we rely on cross-ctional identifi-cation and must be cautious about the interpretation of the findings. The endogeneity concerns are ameliorated by the discrepancy between high-frequency corporate decisions and low-frequency board turnover. (Average director tenure is nine years.) Costs of ter-mination and arch costs make it impractical to adjust board composition for every pol-icy. Thus, even if a director is chon to implement a specific
吃鸡肉的好处policy, firms must consider many policy dimensions. For example, a director hired for her debt market experti will also influence executive compensation and acquisition policies. But, her impact on the latter decisions may conflict with the preferences of the CEO or the shareholders.
Our analysis complements a growing literature relating board characteristics to firm performance (Fich and Shivdasani (2006); Perry and Peyer (2005)) and extends the analysis to specific corporate policies. Our paper relates most cloly to Kroszner and Strahan (2001a and b), who also study conflicts of interest when commercial bankers sit on corporate boards. They find that banker directors are less common in smaller, more volatile firms, where conflicts are most vere, and in distresd firms, where legal con-straints, such as equitable subordination, deter bankers. Our results indicate that conflicts of interests still matter in large, stable firms. Consistent with this interpretation, Kracaw
人口老龄化英语and Zenner (1998) find a negative stock price reaction to bank loans if an affiliate of the lending bank sits on the board of the borrower. Morck and Nakamura (1999) show that banker directors emphasize policies that favor creditors over shareholders in a datat on Japane bank ties.
The remainder of the paper is organized as follows. First, we describe the data (Section I). In Sectio
n II, we investigate the effect of financial experti on internal in-vestment and financing policies. In Section III, we study acquisition and public issuance decisions. In Section IV, we evaluate the impact of financial experti on policies which preclude conflicts of interests. In Section V, we conclude.
I. Data
We analyze a sample of publicly traded companies from 1988 to 2001.9 We build on the datat of Hall and Liebman (1998) and Yermack (1995), merged with CEO demograph-ics from Malmendier and Tate (2005). To be included in the original Hall-Liebman sam-ple, a firm has to appear at least four times on one of the lists of largest US companies published by Forbes magazine from 1984 to 1994. We exclude financial firms.
We hand-collect biographical information on all board members of the compa-nies using annual proxy statements (1988−1997) and the IRRC databa (1998−2001). We code each outside director’s main employment into one of the following categories10: (1) commercial bank executive, (2) investment bank executive, (3) executive of a non-bank financial institution, (4) finance executive (CFO, Accountant, Treasurer, or Vice President for Finance), (5) “finance” professor (including finance, economics, account-ing, and business), (6) consultant, (7) lawyer, (8) executive of a non-fin
ancial firm that falls outside the categories, and (9) non-corporate worker (including careers in acade-mia, nonprofit or civil activist organizations, and politics).
We take additional steps to refine the first two categories, which play a key role in the analysis. First, if the description of a bankers’s employer is vague or missing, we identify the bank from the FDIC list of US chartered commercial banks and the邓丽君简介
9 Our sample period excludes SOX-induced changes. While it would be interesting to evaluate the impact of SOX in an extended data t, such an analysis remains currently infeasible since our estimations would require multiple post-SOX obrvations (after the extended compliance deadlines in 2005). The homogene-ity of pre-SOX data allows us to exploit the full time ries for the estimation of financial-expert effects.
10 The employee falls into more than one category in a few cas, such as banks that are both (1) and (2).
