Can Challenge Be Required? The EBA's Revised Internal Governance Guidelines
The EBA's revised internal governance guidelines require bank boards to challenge management, but a regulator can only prescribe the structures, roles and records that make challenge possible. This text examines which mechanisms decide whether those structures change board decisions, and what they cost.
- The European Banking Authority's revised Guidelines on internal governance, published on 5 November 2026, require bank boards to challenge management, but they can only prescribe the structures, roles and records that make challenge possible. - Separating the chair from the CEO moves formal control of the board's agenda, but real authority stays with whoever holds the information. - Requiring risk and compliance opinions before a new product is approved moves challenge into the search for options, where it can still change the decision. - Documented challenge is evidence that challenge occurred, not proof of it, and the requirement to document can produce records written for the supervisor. - The guidelines are most likely to improve decisions where they change who sees information and when, and least likely where they only add records.
On 5 November 2026, the European Banking Authority (EBA) published its final report on revised Guidelines on internal governance (EBA/GL/2026/10). The guidelines implement Article 74(3) of the Capital Requirements Directive and align the earlier 2021 version with CRD VI and the Digital Operational Resilience Act. They apply two months after publication in all official EU languages.
Most of the text concerns structures: who chairs the board, which committees exist, how duties are mapped. Several provisions, however, concern behavior. The supervisory function should "constructively challenge and critically review proposals and information". The chair should "ensure that dissenting views can be expressed". Decision making "should not be dominated by a single member or a small subset of its members".
This raises a decision governance question. A regulator can prescribe roles, reporting lines and records. Can it prescribe challenge, which is a behavior that occurs, or fails to occur, in a meeting the regulator does not attend?
This text examines what the revised guidelines change in how bank boards make decisions, which mechanisms determine whether prescribed structure produces challenge, and what the requirements cost. It builds on earlier texts in the series on accountability in larger groups and on the relationship between corporate governance and decision governance.
**Case Study: One Product, Two Boards**
Quellmar Bank is a fictional mid-sized retail bank in the European Union, with 1.2 million customers and a board of nine members, six of them non-executive. Management proposes a buy-now-pay-later product for online retailers. The product team expects it to reach 150,000 customers in two years. It also carries credit risk on customers the bank has not served before, and conduct risk if customers misunderstand the repayment terms.
The bank has adapted its governance to the revised guidelines. A non-executive member chairs the board. The risk management and compliance functions give a documented opinion before any new product is approved. Committee minutes record agendas, main results and conclusions.
In Scenario A, the adaptation is formal. The new chair served as the bank's CFO until three years ago and works closely with the CEO. The agenda is drafted by the CEO's office and approved by the chair without changes. The compliance opinion is requested after the product team has signed a letter of intent with a technology partner. The board pack runs to 240 pages and arrives four days before the meeting. Two non-executives raise questions about default rates. Management answers that the risk function has reviewed the model. The minutes record that "members discussed credit risk assumptions" and that the board approved the product unanimously.
In Scenario B, the same rules are applied differently. The chair sets the agenda and asks for the product to come to the board twice: once when three alternative designs are still open, and once for approval. The chief risk officer presents the risk opinion directly, without management editing it. Before the first meeting, each non-executive submits two written questions. At the second meeting, one member votes against, citing the conduct risk of the repayment terms. The minutes record the objection and the reasons, and the product is approved with a lower credit limit for new customers.
Both boards comply with the structural requirements. Only one of them challenged the decision in a way that could change it. The rest of this text examines why.
**What the Guidelines Govern**
Read as decision governance, the guidelines act on four parameters of board decisions: decision rights, participation, the timing of information, and the decision record. They do so mainly through rules-in-form, that is, written rules about who does what. Whether they change rules-in-use, the rules people actually follow, depends on the mechanisms in the next section.
The main provisions, with the paragraph of the guidelines, what each requires, the parameter it governs and the stage of the decision process at which it acts, are as follows.
- **Decision making not dominated by one member or a small subset (para. 22).** Requires dispersed control of board decisions. Governs decision rights, at the Decision stage. - **Supervisory function constructively challenges proposals (paras. 33, 35(b)).** Requires non-executives to review and question management proposals. Governs participation, at the Search and Decision stages. - **Chair should be non-executive, as a general principle (para. 38).** Requires agenda and debate to be led by someone other than management, with mitigating measures where the chair has executive functions. Governs decision rights and power, at all stages. - **Chair encourages open discussion and dissent (para. 37).** Requires conditions in which objections can be voiced. Governs participation, at the Search and Decision stages. - **Committees document agendas, main results and conclusions (para. 61).** Requires a record of what was considered and concluded. Governs memory and transparency, at the Decision stage. - **Mapping of duties and individual statements of roles (paras. 71, 75(a)).** Requires each person's responsibilities to be written down. Governs decision rights and accountability, at all stages. - **Risk and compliance involved before new products are approved (paras. 181, 183).** Requires a prior, documented opinion from control functions. Governs the timing of information, at the Explanation and Search stages. - **Objections of the head of risk management formally documented (para. 216).** Requires dissent from a control function to be on the record. Governs memory and accountability, at the Decision stage.
The five stages are those used throughout the series: Reaction, Explanation, Search, Decision and Action. Most provisions act at the Decision stage, when the board meets to approve. The new-product provisions are the exception: they act earlier, while options are still being formed.
**Why Prescribed Structure May or May Not Produce Challenge**
Challenge is distinct from disagreement. Disagreement is a difference of view. Challenge is a difference of view expressed at a point where it can change the decision, and answered on its merits. Six mechanisms determine whether the structures in the guidelines produce it.
**1. Agenda control**
Whoever sets the agenda decides which questions the board considers and in what order. Separating the chair from the CEO moves this formal right away from management. Research on CEO duality, where one person holds both roles, finds mixed effects on firm performance, which suggests that the separation matters less than how the agenda right is used (Krause, Semadeni, & Cannella, 2014). At Quellmar in Scenario A, the separation changed the title but not the drafting of the agenda.
Governance answer: assign the chair explicit control over which items come to the board, when, and in how many readings. Bring high-consequence decisions to the board while alternatives are still open, not only for approval.
**2. Information asymmetry and real authority**
Formal authority is the right to decide. Real authority is effective control over the decision, and it tends to sit with the party that has the information (Aghion & Tirole, 1997). Non-executive members depend on management for most of what they know about the bank (Adams, Hermalin, & Weisbach, 2010). A board that can only question what management chooses to show it holds formal authority while management holds real authority.
Governance answer: give non-executives direct access to the control functions, as in Scenario B, where the chief risk officer presented without management editing. Limit board packs to what the decision requires, and state what was left out.
**3. Timing of challenge**
Challenge after commitment is ratification. Once a bank has signed with a partner, built a system or announced a product, the cost of reversing grows, and objections become questions about implementation. The guidelines' requirement that risk and compliance give an opinion before a new product is approved places challenge in the Explanation and Search stages, where it can still change the option set.
Governance answer: require control-function opinions before external commitments are made, not only before board approval. Record the date of each opinion relative to the date of commitment.
**4. Social dynamics in the boardroom**
Directors who privately doubt a strategy often stay silent because they believe the others support it. This pluralistic ignorance has been linked to boards persisting with a strategy after poor performance (Westphal & Bednar, 2005). Board effectiveness depends on process as well as composition: on cognitive conflict, effort norms, and the use of members' knowledge (Forbes & Milliken, 1999). The guidelines assign the chair responsibility for open discussion, but they cannot create the conditions in which dissent feels safe.
Governance answer: collect members' views in writing before the meeting, as Quellmar did in Scenario B, so that each member commits to a position before seeing the others'. Have the chair invite dissent explicitly before a vote.
**5. The record as evidence**
Records change behavior because people who expect to justify a decision tend to reason more carefully (Tetlock, 1992). Records also allow later review: an organization can learn from a decision only if it remembers why it was made. But a record can also be written for its reader. Verification can become a ritual, producing evidence that controls exist rather than controls that work (Power, 1997). The Scenario A minute, "members discussed credit risk assumptions", meets a documentation requirement and says nothing about what was decided on those assumptions.
Governance answer: record the alternatives considered, the objections raised and how each was answered, not only the conclusion. Use the records in later reviews of outcomes, so that they are read by someone other than the supervisor.
**6. Contrived versus authentic dissent**
Assigned or scheduled dissent is less effective than dissent from a member who actually disagrees. In experiments, authentic dissent stimulated more and better thinking than a devil's advocate arguing the same position (Nemeth, Brown, & Rogers, 2001). A board that adds a standing "challenge" item to its agenda may produce the form of dissent without its effect.
Governance answer: do not measure challenge by the number of objections in the minutes. Measure whether decisions changed after challenge: conditions added, limits lowered, options dropped.
**Summary of the mechanisms**
The mechanisms fall into two groups. Agenda control, information and timing determine whether challenge is possible. Social dynamics, records and authenticity determine whether it occurs. The guidelines act strongly on the first group and weakly on the second. The second group is therefore where decoupling is most likely: the organization adopts the prescribed structure while its practice continues unchanged (Meyer & Rowan, 1977).
**What the Requirements Cost**
The guidelines are not free, and their costs fall unevenly.
- **Time to decision.** Two readings of a product, prior control-function opinions and written questions add weeks to approval. For a product in a fast-moving market, delay can cost more than the risk it reduces. - **Fixed compliance costs.** Mapping duties, maintaining signed role statements and documenting committee work cost roughly the same in a small bank as in a large one. The EBA's impact assessment states that the final text "removes unnecessary prescriptiveness and simplifies a number of requirements" compared with the consultation draft, and the guidelines allow proportionate application, for example combining the risk and compliance functions in smaller institutions. - **Defensive records.** A requirement to document challenge creates an incentive to document in a way that protects the board rather than informs later review. Over time, this can displace the goal of the record, from learning to defense. - **Loss of unity of command.** A combined chair and CEO can decide and act quickly, and in a crisis this may matter more than independent oversight. The evidence on separation does not show a consistent performance gain (Krause et al., 2014). - **Slower reversal of bad proposals.** More participants and more stages can also make a weak proposal harder to stop once it has passed several reviews, because each review lends it legitimacy.
The requirements are worth their cost when decisions are high in consequence, hard to reverse and dependent on information that management controls. New products with unfamiliar risks, large credit exposures and changes to the group structure fit this description. Routine, reversible and low-value decisions do not, and applying the same procedure to them spends board attention that the high-consequence decisions need.
**Conclusion**
The revised guidelines prescribe the conditions for challenge, not challenge itself. A regulator cannot attend the meeting, so it governs what it can observe: who chairs, who is consulted, what is written down. These are rules-in-form. Whether a bank's rules-in-use change depends on how the board uses the rights the guidelines give it.
Three design implications follow for a board or a supervisor.
1. Act on information and timing first. The provisions most likely to change decisions are those that change who sees information and when: direct access to control functions and opinions before commitment. 2. Evaluate challenge by its effects. A decision that changed after an objection is evidence of challenge. A minute that records discussion is evidence of a meeting. 3. Match the procedure to the decision. Reserve the full procedure for decisions that are high in consequence, hard to reverse and dependent on information management controls, and say explicitly which decisions are exempt.
Quellmar's two boards met the same requirements. The difference between them was not in the rules but in whether the chair, the control functions and the non-executives used them while the decision could still change.
**Definitions**
- **Challenge:** A difference of view expressed at a point where it can change the decision, and answered on its merits. - **Decision rights:** The allocation of who may propose, review, approve and override a decision. - **Formal authority:** The right to decide. - **Real authority:** Effective control over a decision, usually held by the party with the relevant information (Aghion & Tirole, 1997). - **Rules-in-form:** Written rules that state who should do what. - **Rules-in-use:** The rules that participants actually follow (Ostrom, 2005). - **Decoupling:** Formal adoption of a structure while practice continues unchanged (Meyer & Rowan, 1977). - **Management function and supervisory function:** In the EBA guidelines, the parts of the management body that run the institution and that oversee and challenge management.
**References**
- Adams, R. B., Hermalin, B. E., & Weisbach, M. S. (2010). The role of boards of directors in corporate governance: A conceptual framework and survey. *Journal of Economic Literature, 48*(1), 58–107. - Aghion, P., & Tirole, J. (1997). Formal and real authority in organizations. *Journal of Political Economy, 105*(1), 1–29. - European Banking Authority. (2026, November 5). *Final report on Guidelines on internal governance under Directive 2013/36/EU* (EBA/GL/2026/10). [Link](https://www.eba.europa.eu/sites/default/files/2026-11/3c8b1810-c335-499a-803f-6d795ce3d196/Final%20report%20on%20Guidelines%20on%20internal%20governance.pdf) - Forbes, D. P., & Milliken, F. J. (1999). Cognition and corporate governance: Understanding boards of directors as strategic decision-making groups. *Academy of Management Review, 24*(3), 489–505. - Krause, R., Semadeni, M., & Cannella, A. A. (2014). CEO duality: A review and research agenda. *Journal of Management, 40*(1), 256–286. - Meyer, J. W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth and ceremony. *American Journal of Sociology, 83*(2), 340–363. - Nemeth, C., Brown, K., & Rogers, J. (2001). Devil's advocate versus authentic dissent: Stimulating quantity and quality. *European Journal of Social Psychology, 31*(6), 707–720. - Ostrom, E. (2005). *Understanding institutional diversity*. Princeton University Press. - Power, M. (1997). *The audit society: Rituals of verification*. Oxford University Press. - Tetlock, P. E. (1992). The impact of accountability on judgment and choice: Toward a social contingency model. *Advances in Experimental Social Psychology, 25*, 331–376. - Westphal, J. D., & Bednar, M. K. (2005). Pluralistic ignorance in corporate boards and firms' strategic persistence in response to low firm performance. *Administrative Science Quarterly, 50*(2), 262–298.