Why Do Bank Boards Have Risk Committees?

Though the Dodd-Frank Wall Street Reform and Consumer Protection Act (DFA) passed in July 2010 required bank holding companies with more than $10 billion of assets to have a board risk committee, a majority of the banks required to have a risk committee had one before the legislation. The presumption of the legislators apparently was that having a board risk committee would reduce bank risk-taking. As far as we know, there was no scientific evidence at the time suggesting that requiring the establishment of a risk committee for banks that did not have one would be valuable either for the banks’ owners or for the financial system. We develop a model of whether a bank should have a risk committee and show that for a bank that maximizes shareholder wealth there is no expectation that a board risk committee causes bank risk-taking to decrease. Our empirical analysis finds no support for the proposition that the existence of a board risk committee decreases bank risk-taking. We use unique interview data to assess how bank risk committees work and whether they act as expected with our theory. We find that risk committees play a role that is consistent with our theory except that they also seem to be a way for regulators to monitor and influence risk-taking within banks. Though a well-functioning risk committee can be valuable to a bank’s shareholders, it is also possible for the risk committee to worsen the communication and engagement of a bank’s board. Therefore, having a risk committee only makes sense for banks where risk-taking is sufficiently complex that risk metrics have to be monitored by a specialized committee.

Taking risks is a core activity for banks. Therefore, we would expect the board to pay close attention to how a bank takes risks and to the risks that a bank is exposed to. For the board to make useful decisions concerning the bank’s risk-taking, it is necessary for the risk metrics it monitors to be reliable. While the audit committee makes sure that the accounting metrics are reliable, the accounting metrics are not sufficient to assess risk-taking for complex banks. If a bank mostly makes loans, whether the bank makes the type of loans that helps increase shareholder wealth can be assessed using typical metrics that do not require specialized knowledge. However, as the activities of a bank become more complex, monitoring the risk and risk-taking of a bank requires non-accounting risk metrics. Assessing the reliability of these metrics and assessing the bank management’s adherence to board risk-taking targets becomes a more difficult and more time-consuming exercise. When this exercise becomes sufficiently complex and time-consuming, it makes sense to have most of it conducted outside of a board meeting by members of the board who have specialized knowledge. As such, we would expect larger banks and banks with more complex activities to be more likely to have a risk committee. We find that this was the case before the DFA was enacted.

There is no reason to expect that the existence of a risk committee will cause a bank to become less willing to take risks. In fact, the opposite could happen. For instance, a bank board might be willing for the bank to take more risk after the introduction of a risk committee because it is more confident about the quality of the bank’s risk measurement. We examine whether the performance and risk during the global financial crisis (GFC) of banks vary with the existence of a risk committee in 2006 and whether the performance and the risk of a bank changes after the addition of a risk committee during 2003-2018. We focus on banks that would have had a risk committee had the DFA requirement been in force throughout our sample period. We find no evidence that banks with a risk committee perform better or have less risk. This lack of evidence does not mean that having a risk committee when appropriate does not increase shareholder wealth. It could be consistent with a notion that the risk committee at times has the impact of reducing risk-taking and at other times has the opposite impact. As a result, risk-taking would more likely be at the level that maximizes shareholder wealth. However, the evidence is inconsistent with the view that banks take too much risk without a risk committee and the committee reduces risk-taking.

Finding no evidence that risk is lower when a bank has a risk committee is consistent with our model where the board is focused on maximizing shareholder wealth and uses the risk committee to help it be more effective. In such a situation, the risk committee improves the board’s monitoring of the bank’s risk and risk-taking and provides management with advice. With the alternative theories, the risk committee would more directly attempt to reduce the bank’s risk either because management wants to take too much risk or because regulators want the committee to do so. It is not possible to investigate directly which functions a risk committee performs using traditional datasets. To assess more directly the role that risk committees play, we use a unique dataset, which comprises in-depth interviews with 20 chairs of risk committees in financial institutions. This dataset allows us to understand how these committees perceive their role and how they proceed in trying to fulfill that role.

The financial institutions represented in our panel differ considerably in size. It is clear from the interviews that the workload increases with bank size. The issues that a risk committee at a large institution has to deal with are such that they could not possibly receive the same attention if they were addressed only at a plenary board meeting. It is also quite clear from the interviews that chairs of risk committees believe that it is important for them to have direct access to the leadership of the risk management team and to develop a good working relationship with that team. A concern with the role of risk committees in the DFA is that they could be a way for bank examiners and regulators to push their agenda. We find that regulatory matters are time-consuming for risk committees and impact committees’ agenda considerably. Further, in many cases, the risk committee chair interacts directly with bank supervisors and regulators. It seems clear that bank examiners and regulators influence risk committees’ work. Though having a committee chair that engages with the risk management leadership team and with bank supervisors on his/her own and without the presence of the CEO would seem to be a possible source of tensions, our interviews do not show that such tensions are important.

The complete paper is available for download here.