Hypotheses - the board level

The following four hypotheses relate to board structure as they are built on the board’s characteristics as a whole and thus hypothesize on the second part of the research question.

5.3.1 Board size

As the board’s role is to ratify, verify and monitor the actions initiated by the top management (see section 4.8.1), a trade-off seems to emerge between resources held on the board and coordination and communication costs between the board’s members. While the capacity to monitor the top management increases with board members, this benefit may be outweighed by the additional costs of decision-making between extra board members. (John, Senbet 1998). In plain terms, the trade-off is between having a large, resourceful board and having a small board, which has lower communication and coordination costs.

Lipton and Lorsch (1992) find that coordination and communication costs increase with the number of members on the board. The additional cost may impede the benefits from having a larger array of resources available in the monitoring function, which Jensen (1993) endorses.

He finds more specifically that boards with more than seven or eight members are unlikely to be effective.

The markets appear to agree with these findings, as Yermack (1996) finds that valuations of firms decrease over a range over board sizes from four to ten members. Furthermore, the study finds that profitability and asset utilization decrease as the board becomes bigger, again in a range from four to ten members. Bhagat and Black (1996) confirm Jensen’s (1993) predictions and confirm Yermack’s findings. Mak and Yi (2001) remark that smaller boards are associated with effective monitoring and less free-riding on behalf of the individual director.

In a traditional finance research stream of thought, the trade-off between resources on the board and increased communication and coordination cost should be solved by the market forces. If Yermack’s findings are applicable in a broader sense, shareholders – interested in maximizing the value of their share and, in principle, nothing else – should punish boards that are too large (as having too large a board would mean forgone value increases of their shares).

However, the very existence of the findings above suggest that market imperfections may exist (John, Senbet 1998) that eliminate or reduce the effect of the arbitrage mechanism, also in this relation. Denis (2001) remarks that board size may be a reflection of firm dynamics and while she concludes that the literature point in one direction only, she finds that board structure in general and size in particular may be a result of other corporate governance

mechanisms. Either way, the empirical studies on board size are greatly in favor of smaller boards, although not completely linear in the sense that a single member on the board is the most effective.

Smaller boards thus seem to be value-creating for the shareholder, or; the communication and coordination costs increase rapidly as the number of board members increase. When the board is poorly coordinated, it can be argued that its power base erodes and leads to less regard for shareholder interests. Hermalin and Weisbach (2003) find exactly this and conclude that smaller boards lead management to act increasingly in alignment with shareholder interests.

In the same review it is found that smaller boards are more likely to remove poor executive managers. Finally, relating these findings to risk, Cheng (2008) notices that firms with larger boards have less volatility on their stock and thus display less risk-taking behavior (which, ceteris paribus, is not in the interest of the shareholders, see section 4.6.1).

The Danish Recommendations on Corporate Governance formulate the recommendation rather vaguely: “The Committee recommends that the supreme governing body have only so many members as to allow a constructive debate and an effective decision-making process enabling all members to play an active role” (DRCG, 5.3.1) (The Danish Commerce and Companies Agency 2005). Therefore, the thesis at hand finds its support in the empirical research, which finds that smaller boards are more aligned with shareholders.

In this thesis, the findings above are formed into the following hypothesis:

Ho: There is no relation between board size (for sizes over 4) and risk-taking in Danish banks.

H7: There is a negative relation between board size (for sizes over 4) and risk-taking in Danish banks.

5.3.2 Incentive programs

As this thesis is a corporate governance-focused study, management is seen as a product of the system generated by (among other things, such as culture, legislation etc.) the owners or, as a representative of these, the board.

Therefore, a deeper literature review on management and its influence in organizations is viewed as outside the scope and interest of the paper at hand, because the main issues in the relation between owners (and the board) and management have already been outlined in section 4.4.1.

What is the focus, however, is which actions the board of directors take in asserting the shareholders’ rights and incentives in the daily management. The literature on incentive programs is rather one-sided, though, and therefore, the hypothesis construction is not one of weighing a trade-off, but one of reviewing the prevailing view among scholars in the field (Core et al. 1999).

Many corporations use incentive programs to mitigate the type-1 agency problem described in section 4.4.1. Two main types of incentive programs exist (Merchant, Van der Stede 2006), stock options (or stock payment) and bonus systems. Usually, stock options increase in value with the corporation’s stock’s value, while bonus programs pay a specified bonus and a percentage of this in relation to the fulfillment of certain, pre-specified measurement goals.

Gore et al. (2010) and Core et al. (1999) define the CEOs incentives as the change in executive rewards brought about by changes in shareholder wealth and this thesis leans upon this definition in relation to stock option payment schemes.

The rationale behind using these is to shift the incentives of the management to those of shareholders, thus reducing the type-1 agency problem, but increasing the type-3 agency problem which is in the interest of the shareholders (Kose, Mehran et al. 2010). Kose et al.

(ibid) find specifically that aligning managerial incentives with shareholders’ interests will exacerbate the shareholder-debtholder conflict in leveraged firms. In particular, managers who are aligned with shareholders will have the risk-shifting incentives i.e. the incentive to undertake excessive risk at the expense of debtholders.

This thesis does not seek to investigate the level of compensation nor the performance measurement related to this, although literature abounds. Instead, the existence (or lack) of incentive programs as an indication of risk-alignment between the board of directors and management is the object for the study.

Jensen and Murphy (1990) predict theoretically that CEOs are only motivated to act in their shareholders’ best interest if they are offered incentive contracts that pay for ‘performance’.

Milbourn (2003) describes this as the (or one of the) first studies to examine that phenomenon. This implies that CEOs with incentive payment packages will act (increasingly) in accordance with shareholders’ wishes.

As the CEO’s assumed risk-averseness is a central assumption, his pay-based incentives can be explored through the use of the elegant representation by Agrawal and Mandelker (1987), who offer an algebraic outline of the misalignment of incentives²⁴.

The manager’s total wealth is represented by W, which is comprised of three elements:

a) The wealth he derives from his human capital in the principal’s company (his salary):WH b) The wealth he derives from his stockholdings or –options (if any) in that company: WS c) The wealth he derives from holding assets unrelated to the firm: W0

Thus, the manager’s total wealth is summarized in the equation:

W = WH + WS + W0,

with W being variable with σ², which depends on the volatility of WS and W0. Since W0 is assumed to be completely uncorrelated to his actions in the focal firm, the volatility of the manager’s total wealth is equal to the volatility on the stock of the company, because the others are constant. The resulting behavior - ceteris paribus - rests on the manager’s assumed rationality:

If he holds none of the company’s stocks or options, his W is maximized by keeping WH as high as possible, which is best done by making sure the volatility on the income is as low as possible, thus improving the certainty equivalent in future income stream: the smaller the risk of bankruptcy, the higher the likelihood of deriving pay from the company in the future.

If the manager, on the other hand, holds shares, his incentives become less obvious. On one hand, WH is still maximized when keeping default at bay. This conflicts with WS, which is maximized at relatively higher levels of risk. Thus, awarding the manager stocks or options will shift his incentives through the increased risk he bears, but the amount of risk is in a trade-off for the manager with the inversely related movements in WH²⁵. This is somewhat similar to the pay-off proposition to the shareholders, of course, as they too should prefer the company in any solvent state over the bankrupt state.

Agrawal et al. (ibid) thus find the implementation of incentive programs will reduce the type-1 agency problem as the CEO adopts a more risk-seeking behavior.

Therefore, the thesis hypothesizes the following:

24 They assume that the principal is risk-neutral as well and that the manager acts rationally.

25 I.e.: when WS goes up, WH goes down.

Ho: The existence of stock-option payment schemes has no relation to risk-taking in Danish banks

H8: The existence of stock-option payment schemes is positively related to risk-taking in Danish banks

Ho: The existence of bonus payment schemes has no relation to risk-taking in Danish banks

H9: The existence of bonus payment schemes is positively related to risk-taking in Danish banks

5.3.3 The CEO’s tenure versus the board’s tenure

While the board of directors officially is the representative of the owners of the company, multiple studies conclude that CEOs have actual power in the organization, among these Reinganum (1985), Smith and White (1987) and Thomas (1988), who all find that CEOs have strong effects on organizations. The research on the relation between the board’s tenure and the CEO’s tenure is relatively scarce and the studies that do exist point in diverging directions.

In relation to agency theory, it could be implied that CEOs with longer tenure have stronger effects on the organization as a whole (due to an increased advantage in asymmetrical information, see section 4.3.5).

Brickley et al. (1997) and Coles (2001) find that when the CEO has long tenure, he has earned the trust of the shareholders and thus requires less monitoring to act in the interest of them, because he is assumed to be knowledgeable and committed to the company. They argue that the fact that he has held on to his position for a relatively long period implies shareholders alignment – otherwise he would not have held his seat for so long. Schwenk (Schwenk 1993) finds that tenured CEOs formulate more efficient strategies and policies that will enhance the company’s performance, thereby decreasing the type-1 agency problem, as his actions are wealth creating.

On the other hand, Pathan (2009) finds that bank risk-taking is positively related to strong bank boards and negatively related to CEO power²⁶. Walters et al. (2007) find that when the board is weak and/or new, then CEO tenure is negatively associated with performance and

26 The term “a strong board” is defined in this hypothesis as a tenured board.

risk-taking. In the study, the researchers find that the longer the tenure of the board compared to that of the CEO, the better the shareholder alignment of the CEO.

Shakir (2009) finds that CEOs who sit in their position for a long time may become complacent. Also, Shaker suggests that the tenured CEO facing a new board may act self-interested and entrench himself, which is not in alignment with shareholders and which is increasing the type-1 agency problem.

Thus, a hypothesis emerges of this paragraph leaning in the direction of the self-interested behavior from CEOs with longer tenure than the board, as this corresponds best to the agency theory outlined in section 4.6.3 The management.

Ho: The tenure of the CEO versus the tenure of the board does not influence risk-taking in Danish banks

H10: CEOs with longer tenure relative to the board have more power than CEOs with shorter tenure and thus, having a CEO with longer tenure is negatively related to risk-taking in Danish banks.