Stanford’s David F. Larcker and Brian Tayan have a new paper called Loosey-Goosey Governance: Four Misunderstood Terms in Corporate Governance in which they appear to think they’ve discovered what everyone has understood forever — that there are limits to structural solutions and that checklists of best practices are not especially helpful.
We were very clear about this when we founded The Corporate Library in 2000 (later GMI and now a part of MSCI). Our ratings of boards were not based on what they said they did or on “resume independence” but on the quality of the decisions they made, especially in those areas where there were potential conflicts of interest between management and shareholders. For example, managers, including CEOs, would like less risk and volatility in their incentive compensation, while investors would like a tighter link between pay and performance. If conflicts like those were not resolved in favor of investors, the board got a lower rating. Over time, our ability to find indicators of litigation risk, liability risk, and investment risk were so reliable that we were able to meet the ultimate market test of providing the ratings to D&O liability insurers as well as institutional investors.
Larker and Tayan appear to think it is insightful that there is no proven benefit from independent board chairs. That has been shown many times already. The reason is not that independence is unimportant; it is that in the US “independent board chairs” or splits between CEO and Chairmen are not especially meaningful in the aggregate. That does not mean that it is not a useful change when a company is failing. It means that calling someone an independent chair does not actually mean that the chair is actually independent. We have used the example many times of the board at Juno Lighting, which consisted of the CEO, the CFO, the company’s lawyer, the company’s investment banker, and another guy who insisted he was independent but who was a pal of the CEO who played in the same jazz band. That is not the kind of connection that will show up in an SEC filing, but in reality there is no way to call him “independent.” As long as insiders control the nomination process and even a 99 percent no vote cannot boot someone off the board, the term “independent” has very little meaning. This is even clearer when the insiders determine when the CEO and chair positions should be split and who should serve as chair.
If investors are over-focused on these kinds of indicators (as they tend to be due to the cost/benefit of this kind of information) perhaps we should consider what kind of weight would be attached to those initiatives if other options were available — like, say, removing board members on, say, comp or nominating or audit committees.
“Best practices” have two functions. First, when things are not going well, they suggest changes that may be helpful. Second in a “comply or explain” model, they can shift the burden of proof so that executives and directors need to explain why their version of governance is better. If, for example, every board member has a substantial proportion of his or her net worth invested in the company, some of the conventional “best practices” may be superfluous. If a board is made up exclusively of white males, we’d be willing to hear the reason why, and confident that having to answer it would be a worthwhile exercise.
“Why can we still not answer the question of what makes good governance?” Larker and Tayan ask. Answering that question is not difficult, but looking at structures and check-lists is the wrong place to start. Good governance means optimal risk management, determined by evaluating the pay/performance link over the long term compared to the peer group and the market as a whole, strategic planning, management of conflicts of interest, acquisitions and divestments of non-core assets, transparency of financial reporting (including explicit quantifiable goals for any “stakeholder” priorities), CEO succession planning, etc.
Speaking of incentives, we’d like to know more about who funded this research.