Academic

Bill Carney on Broughman's Independent Directors

By Gordon Smith, Christine, Hurt, Vic Fleischer, Fred Tung, Lisa Fairfax, David Zaring

Sorting out how investors in closely held firms arrange their governance under conditions of bounded rationality is a wide open topic. It was ignored by law & econ writers who concentrated on the Berle & Means and agency cost models for the past 70 years. The few who wrote about conflicts in a closely held firm usually did so without either a theory or evidence beyond anecdotes. I attempted to describe the fundamental conflicts among co-investors and the contractual devices available to resolve them in 1987, in an article that gained no traction at that time.1 Now we have the beginnings of a rich literature, to which this is a good contribution. The new tradition that is developing, as evidenced by this paper, involves really hard empirical work ? not just employing existing databases, but actually using questionnaires and interviews to obtain private information. That young scholars are willing to do this hard work is encouraging for the future of academic work ? it will stay closely in touch with reality.

This paper demonstrates that the role of the independent director in the VC-financed firm may be distinctly different from the director?s role in the publicly held corporation. Monitoring fades to unimportance where investors can monitor directly, and independence is defined here not as independence from management, but from two classes of investors - entrepreneurs and Vcs. Similarly, the advisory role of independent directors has little power where the investors are knowledgeable and involved.

Initially I worried that this study, like others, takes a static look at board governance. That is, the mix might very well depend on what stage of development the firm is at, and how many rounds of VC financing it has had. Each round may introduce new VC directors, and increase the uncertainty about whether conflict will arise. But this problem was solved by looking at allocation of board seats for all rounds of financing. The paper suggests that the first round may be more relevant for this study because it involves only one class of preferred, but I?m not convinced that?s true. Nevertheless, the results hold for all rounds, to my surprise. I know the numbers would be small, but a separate look at board composition for each round might disclose some trends that would be interesting.

The paper doesn?t deal with some of the contractual devices that VCs obtain to protect themselves from entrepreneurial opportunism. While I understand the criticism of the limits of contracting, the paper might be richer (and I guess less focused) if it looked at these devices in connection with board composition.

The model sets up the conflict well, but it?s a bit constrained. The model suggests that the VC will only get the original investment back plus accrued dividends, but VCs can strike much tougher deals than that. Requiring a liquidation value of several times the original investment assures that entrepreneurs won?t get anything at all unless there is a very large payoff for the firm, although it doesn?t eliminate the moral hazard of the high risk investment alternative. Consider the following outcome from a high-risk strategy if the VC holds participating preferred, where the VC gets the first $120, and 50% of the remaining $180 in the winning alternative:

If the VC?s return is based on a multiple of its original investment prior to an exit event, the results will be similar. These devices can control risky choices; contracting is not without its responses to some of these problems.

Finally, the right of the VC to convert into common means that the VC?s return isn?t fixed if the entrepreneur causes a spectacular rise in value through a risky strategy. convert to common or fully participating preferred stock which, after being paid its par value and accrued dividends, shares pro rata with the common in the remaining assets. The example given doesn?t deal with the whole range of risky outcomes, or the range of contractual strategies. The outcomes are likely more open-ended than the model, so it?s possible some outcomes from risky projects are better than the model?s limit. Consider the following outcome from a high-risk strategy if the VC converts to common, and the payoff is $500:

Here the VC?s payoff exceeds its fixed payment rights on the preferred, in exchange for allowing the entrepreneur to choose a risky project. There are other devices as well ? requiring a supermajority approval of budgets is one, and setting limits on individual expenses by managers without board approval is another.

VCs typically bargain for the right to put their shares to the company if an exit event (either an IPO or a sale) isn?t achieved within a few years (five is used in the forms provided by the National Venture Capital Association. In practice, redemption rights are not often used; however, they do provide a form of exit and some possible leverage over the Company. Further, the mandatory conversion when a company goes public generally is only triggered if the IPO price returns a multiple of the original investment to the VC.

The paper also doesn?t address the protection the VC gets from the usual custom of serial investments. The first round of financing is virtually never expected to get to the goal of profitability. If the VC is to be expected to participate in a second round, the entrepreneur cannot choose the high risk strategy. It is only in the final period that the entrepreneur would be free to do this, and it?s possible that by then the VCs have the lion?s share of the investment, as well as board seats.

Part IV discusses the role of fiduciary obligations and suggests that they require directors to serve the best interests of the corporation, which presumably would lead to efficient decisions for independent directors. But this oversimplifies the description of the law. Corporate law generally provides that directors owe fiduciary duties only to common stockholders, and not to the preferred. Preferred is generally seen as strictly a creature of contract, as the paper acknowledges. If independent directors are advised of these rules, how can venture capitalists be assured that the independent directors will favor the efficient outcome at the expense of the common? The paper?s answer is that directors owe duties to the firm to maximize its value, but that of course doesn?t really provide a compelling answer when the two groups are in serious conflict. If the law were about maximizing firm value, the Independent Director would choose the efficient outcome. But would a director in the middle seek advice of counsel? Are VCs naive in expecting firm wealth-maximizing decisions from independent directors? These questions could result in a spin-off study of the dynamics of board decisions when these types of conflicts arise. All good papers raise more questions, and this one succeeds.

What explains the expectations of both parties that the independent director will support the most efficient choice? The paper suggests reputational concerns might be sufficient to make the independent director an honest broker, but consulting with counsel about legal obligations when conflicts arise might lead to a pro-entrepreneur bias. The analogy of arbitrators, who have multiple repeat dealings in disputes, isn?t very persuasive, where independent directors are uninfluential where the VCs and entrepreneurs agree, and where directors have far fewer opportunities to serve, much less resolve conflicts. Here modeling doesn?t advance analysis of this problem. The paper acknowledges that the existing literature doesn?t address the issue of independent director bias, and acknowledges that VCs may have networks of independent director candidates who owe their repeated selection to the VC. Entrepreneurs are far less likely to have similarly extensive networks. The paper attempts to address the problem by examining whether both parties had a role in the selection of the independent director or both had a prior relationship with the individual (the de facto coding). It?s hard to know if this really captures the dynamics of the relationships and the loyalties. I?m not sure that more can be done in this fascinating area, however.

I not only learned a good deal from the paper, it whetted my appetite for more. That?s a sign of a successful effort.

¹Carney,

Full post as published by Conglomerate on July 28, 2008 (boomark / email).