The Effects of Internal Board Networks: Evidence from Closed-End Funds

Several recent studies document the importance of social networks in corporate governance. Most of this research focuses on networks between board members and the CEO, or between directors and outside parties. My study, The Effects of Internal Board Networks: Evidence from Closed-End Funds, recently published in the Journal of Accounting and Economics, is the first to examine a different aspect of social networks: internal connections between members of an individual firm’s board of directors. These internal networks might affect governance in two ways. On one hand, we might argue that shared backgrounds can improve communication and facilitate decision making, therefore improving firm value and monitoring quality. On the other hand, the shared backgrounds of directors may reduce the likelihood of dissent as a result of fewer opposing viewpoints. This argument follows the theoretical model of board communication presented by Malenko (2014), which highlights the improved quality of governance associated with having diverse preferences on the board and with incentivizing directors to openly communicate opposing viewpoints. In this model, diverse interests strengthen the director’s incentives to incur the costs of dissent.

Using a sample of closed-end funds, I find evidence supporting Malenko (2014). A more closely connected board is associated with lower firm values and higher costs, both direct and indirect, borne by shareholders. Direct costs take the form of the expense ratio, which is higher for funds with connected boards. Indirect costs show up as a result of poor monitoring and oversight, as evidenced by an increased likelihood of rights offerings and more frequent cases of deceptive disclosure practices. Directors personally benefit from these shareholder costs, as internally connected boards have higher levels of director compensation and lower turnover among directors.

Boards of closed-end funds present an interesting and advantageous dynamic for studying these questions. The responsibilities of a fund’s board are largely identical to any other corporation; directors monitor the operations of the fund and are expected to represent the interests of the shareholder in fund decision-making. Funds are subject to the same governance regulations as industrial corporations and have similar governance structures. These funds, however, present a superior setting for testing governance issues. The closed-end fund discount provides a direct, accurate, and observable measure of the disparity between market value and the net asset value (NAV) of the underlying assets. In a study of industrial corporations, we must use Q-ratios to estimate this disparity, which only allow for annual or quarterly estimation and are also reflective of a variety of unobservable firm characteristics such as inconsistent accounting values, misstated liabilities, and varying growth opportunities. Because asset values of closed-end funds are observed and accurately measurable, closed-end funds provide a setting that alleviates these issues. I find that the presence of internal networks is associated with lower fund market values (that is, larger discounts from net asset value). For the average board consisting of 8.5 directors, each additional internally connected director is associated with an 88 basis point increase in fund discounts.

The closed-end fund setting also provides an easily observable measure of board effectiveness. One of the board’s primary responsibilities to shareholders is minimizing the expense ratio. A higher expense ratio therefore implies a less effective board. I find that expense ratios are positively related to the presence of internal board networks, consistent with the hypothesis that these boards are not acting in shareholder interests. For the average board consisting of 8.5 directors, an additional internally connected director is associated with a 5.2 basis point increase in the expense ratio. I next test whether directors benefit from these higher expense ratios. The sample of closed-end funds provides a conveniently homogeneous sample of directors. Since these directors are all serving in the same industry and overseeing similar firms, they are expected to undertake similar tasks and are likely to have similar skills and qualifications. Additionally, directors in these funds are legally only allowed to be paid in cash, providing a straightforward valuation of the total compensation package that minimizes measurement error. I find that internal board networks are positively and significantly related to director compensation. Each additional connected director relates to a 7% increase in director compensation.

One of the most common tools to address underperformance in a closed-end fund is a share repurchase, which allows shareholders to redeem a limited number of shares, typically when trading at a large discount on the market, for a price approaching NAV. Because this provides shareholders with a free in-the-money put option on their shares, it is always beneficial for the shareholder. Further indicating the agency concerns associated with internal board networks, I find that a closely-connected board is associated with a lower likelihood of a share repurchase. Additionally, I find evidence that these boards are more likely to issue additional shares in the form of a rights offering, shown in prior research to decrease fund premiums and be an additional source of rent extraction from shareholders.

The ineffective monitoring rising from a lack of dissent may also affect shareholders through the fund’s disclosure efforts, as directors have incentive to inflate fund performance figures. To measure this, I look to cases of SEC enforcement for financial misrepresentation. I find that internally connected boards are associated with a higher likelihood of SEC enforcement.

If these networks do encourage self-dealing behavior as evidenced by prior results, we would expect lower turnover for board seats in firms where networks are strong; directors should be more reluctant to leave when they are extracting valuable benefits from the position. When new appointments do occur, we would expect an existing network to have an incentive to grow stronger and minimize potential dissenting voices by appointing a director who is involved in the network as well. I show that both of these are true: boards with strong internal networks are associated with lower turnover and an increased likelihood of new director appointments being connected to an existing director.

The study improves our understanding of boardroom dynamics while raising important concerns about how relationships among board members may affect the quality of the board’s monitoring.

The complete paper is available for download here.