The Fed’s 2018 asset cap on Wells Fargo made governance the price of growth and outside reviewers the gate

On February 2, 2018, the Federal Reserve restricted Wells Fargo from growing beyond its end-2017 total assets until board oversight and risk management improved. In our reading, the order turned governance into a measurable cost and made third-party reviews the route out.

Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.

Two-minute summaryThe dek and the first sentence of each section, quoted from the article

On February 2, 2018, the Federal Reserve restricted Wells Fargo from growing beyond its end-2017 total assets until board oversight and risk management improved. In our reading, the order turned governance into a measurable cost and made third-party reviews the route out.

  1. The week On Friday, February 2, 2018, the Federal Reserve Board announced an enforcement action against Wells Fargo & Company.
  2. The meter: a two-quarter average against a fixed baseline The first part was the measure itself.
  3. The conditions: two plans, each a governance specification The second part was what the firm had to write.
  4. The reviewers: independent experts as the gate The third part was the review.
  5. The incentives: pay, accountability and the board’s own information The fourth part was about behavior.
  6. How this applied across institutions The order bound one firm.
  7. Objections, in order We take five objections, numbered in the order we address them.
  8. What a leader could check These checks follow from the four parts.
  9. With hindsight The cap lasted far longer than the order’s first review date suggested.

The week

On Friday, February 2, 2018, the Federal Reserve Board announced an enforcement action against Wells Fargo & Company. Its press release said the Board would restrict the firm’s growth until it sufficiently improved its governance and controls. Until then, the firm would be restricted from growing any larger than its total asset size as of the end of 2017. The release said the restriction would not require the firm to cease current activities, including accepting customer deposits or making consumer loans.

The action took the form of an order to cease and desist, issued upon consent. The Board required each current director to sign it. The press release said that, concurrently with the action, Wells Fargo would replace three board members by April and a fourth by the end of the year.

The order’s recitals set out the Board’s view. They said the firm had pursued a business strategy that emphasized sales and growth without ensuring that senior management had established an adequate risk management framework, which resulted in weak compliance practices. They said the Federal Reserve Bank of San Francisco and the Board had previously identified risk management deficiencies, including in compliance risk management, that the firm had yet to correct fully. They cited the Federal Reserve’s Regulation YY, which requires large bank holding companies to maintain a firmwide risk management framework. The firm consented to the order without a hearing, solely to settle the matter.

The order’s central requirements came in sequence. Within 60 days, the board had to submit a written plan to further enhance its effectiveness in oversight and governance, acceptable to the Reserve Bank. Within the same period, the firm had to submit a written plan to further improve its firmwide compliance and operational risk management program. After adopting and implementing those plans, the firm had to complete an independent third-party review by no later than September 30, 2018, and later a second review of whether the improvements were effective and sustainable.

The growth restriction was precise. It applied from the last day of the first full calendar quarter after the order. From then, without prior written approval, the firm could take no action that would push its average total consolidated assets above the December 31, 2017, figure. The average covered the current and preceding quarter. The figures came from specified lines of the FR Y-9C, the consolidated financial report bank holding companies file with the Federal Reserve. The restriction would stay until the plans were accepted and implemented, the first third-party review was completed to the Reserve Bank’s satisfaction, and the firm was notified in writing that the conditions were met.

The Board also sent letters. The press release said letters to each current director confirmed that the board, during the period of compliance breakdowns, did not meet supervisory expectations. Separate letters went to two former senior leaders. The letter to the board said Wells Fargo Bank had disclosed, in September 2016 settlements, that it had opened several million potentially unauthorized retail customer accounts between 2011 and 2015. It said more recent disclosures concerned charges to hundreds of thousands of borrowers for unneeded guaranteed auto protection or collateral protection insurance. It said the firm’s performance management and compensation programs had played a material role in its compliance breakdowns.

The action followed a broader Federal Reserve proposal. On August 3, 2017, the Board had requested comment on supervisory expectations for boards of directors. That proposal identified attributes of effective boards for the largest firms, such as setting a clear strategic direction, supporting independent risk management and holding management accountable. It said that, for the largest firms, supervisors would use those attributes in evaluating governance and controls.

Enforcement orders against banks speak the language of findings, plans and deadlines. The Federal Reserve’s February 2018 order against Wells Fargo added a ceiling on the size of the balance sheet. In our reading, that ceiling did something a fine cannot do. It priced weak governance in terms of the firm’s own business plan, as growth it could not book, for as long as the weakness lasted.

The order also said how the ceiling would come off. It was not enough for the firm to say it had improved. Plans had to be accepted, implemented and then reviewed by outside experts the Reserve Bank found acceptable. That made the reviewers, and the evidence they could examine, the gate between the firm and its growth.

We take the order in four parts. Each is a separate lever, and together they cover how the order turned governance into economics:

  • The meter: how the cap measured the firm’s size, and what it left the firm free to do.
  • The conditions: the two plans the firm had to write, and what they had to contain.
  • The reviewers: the third-party reviews that decided when the cap could be lifted.
  • The incentives: the compensation, accountability and reporting changes the order demanded.

Our argument is that the order converted governance from a supervisory opinion into a cost with a meter attached. Every quarter the cap stayed in place, the firm bore it in forgone balance sheet growth. The way out ran through documents and independent reviews, so the quality of the firm’s evidence decided how long it paid. For any bank, the order is a reminder that governance has an economic value regulators can enforce. After the four parts, we look at how this applied across different institutions, at objections, and at what a leader could check.

Why look at an enforcement order through an economics lens? Because the order’s design made the firm’s management face a trade-off every quarter. Spending more on governance, controls and evidence could shorten the cap. Spending less would extend it. The order gave the board a financial reason to fund remediation properly, and made the cost of not doing so visible in the firm’s own growth.

A note on our reading. We describe the order and its conditions from the Federal Reserve’s own documents. The order was a consent order that settled the matter without findings after a hearing. Statements in it about the firm are the Federal Reserve’s, and our analysis of their economic effect is our own.

The meter: a two-quarter average against a fixed baseline

The first part was the measure itself. The order did not cap deposits, loans or any one business. It capped total consolidated assets, as a two-quarter average, at the December 31, 2017, figure. That design had consequences for how the firm could manage its balance sheet.

A whole-firm number

Total consolidated assets is the broadest measure of a bank holding company’s size. It includes loans, securities, trading assets, cash and every other asset across the consolidated group. Capping it meant growth in one business had to be offset elsewhere. A new loan funded by a new deposit added to assets. To stay under the cap, the firm had to reduce some other asset, or decline the new one.

That turned balance sheet allocation into an internal market. Businesses that wanted to grow had to compete for space under a fixed ceiling. Treasury and finance functions had to decide which assets earned their place. The cap did not tell the firm which businesses to shrink. It made the firm choose.

Averaging smooths, and constrains

The order measured the average of the current and preceding quarter. Averaging protected the firm from a single day’s spike, for example from client deposits arriving at quarter end. It also meant a sustained increase could not be hidden by trimming assets on the last day of a quarter. The measure looked at the firm’s size over time, not at a snapshot.

What it did not stop

The press release was explicit that the firm could continue to accept deposits and make consumer loans. The cap restricted growth, not activity. Customers could still open accounts and borrow. The firm’s task was to serve them within a fixed total, which meant reducing other assets as client business grew.

That distinction mattered for customers and counterparties. A growth cap is not a closure order. Depositors were not told to leave and borrowers were not cut off. In our reading, the cost fell on the firm’s growth and its shareholders, not on its existing customers’ access to services.

Approval as a valve

The cap applied unless the Reserve Bank, with the concurrence of the Board’s supervision director, gave prior written approval. That gave the supervisors a valve. Any request to exceed the cap would require the firm to make a case to the same supervisors judging its governance. A firm in that position had a strong reason to keep its supervisors confident.

Planning under a ceiling

A cap with no end date changes planning. Annual budgets, capital plans and business strategies all assume some path for the balance sheet. Under the order, the path was flat until the conditions were met, and no one could say when that would be. Business lines had to plan for growth they might not be allowed to book, or plan for none and risk losing ground if the cap came off sooner.

That uncertainty had its own cost. Hiring, technology investment and product launches tied to balance sheet growth became harder to justify. A firm in that position would favor activities that earned fees without adding assets, and would look harder at assets that earned little. The cap shaped strategy as well as size.

A cost that is hard to count

Forgone growth is real and hard to measure. No one can observe the balance sheet the firm would have had without the cap. Estimates depend on assumptions about markets, competitors and the firm’s own strategy. That makes the cap’s cost a matter of judgment, which is part of its force. A board cannot treat it as a known line item and move on. It remains an open question every quarter, and the only way to close it is to meet the conditions.

Who owns compliance with the cap

Compliance with the cap needed an owner. Treasury managed the balance sheet. Finance produced the regulatory report. Business lines generated the assets. The board answered to the supervisors. A breach could start in any of them. A firm under a cap needs a forecast of the two-quarter average, updated in time to act on, and a named executive who can tell a business line to stop adding assets.

The measure as a reporting artifact

The cap was defined by lines on the FR Y-9C, the firm’s regulatory report. That tied a governance remedy to a financial reporting process. The accuracy and timeliness of the report became part of compliance with the order. A reporting error that overstated assets could look like a breach. One that understated them could hide one.

The conditions: two plans, each a governance specification

The second part was what the firm had to write. Paragraph 2 required a board effectiveness plan within 60 days, acceptable to the Reserve Bank. Paragraph 3 required a plan to improve firmwide compliance and operational risk management, on the same timetable. Each plan had a long list of minimum contents.

The board’s plan

The board’s plan had to address how the board would improve its own effectiveness. That included making sure the firm’s strategy and risk tolerance were clear, aligned and within its risk management capacity. It included making sure the board’s composition and structure supported that strategy, and that no board role went unfilled for an undue period after a departure.

The plan also had to address how the board would oversee and hold senior management accountable. It had to set out what the board would do to make sure management maintained an independent firmwide risk management function, with stature, authority and resources. For compliance and operational risk, it needed a management structure independent of the business lines, with separate, independent reporting lines to the Chief Risk Officer and to the board or a board committee.

Risk data and escalation

Two items in the board’s plan were about information. The firm needed an effective risk identification and escalation framework that identified, aggregated, evaluated and reported material risk issues. It also needed a comprehensive and effective risk data governance and management framework. The plan had to provide reporting that would let the board oversee management’s risk management and judge the effectiveness of the compliance and operational risk functions.

Those are data requirements, written into an enforcement order. A board cannot oversee what it cannot see. The order required the firm to build the pipes that bring risk information to the board in a form the board can use.

The firm’s risk management plan

The second plan had to cover testing and validation of compliance and operational risk controls, including consumer compliance. It had to show how compliance requirements would be built into business processes, control design and change management. It had to strengthen operational risk management, including a key risk indicator program tied to risk tolerance, and a stress loss forecasting method that quantified operational risk under stress.

It also required a review of policies, procedures and practices across the firm for remediating customers harmed by the firm’s products, services or employee misconduct. Deficiencies found in that review had to be addressed. So the plan reached back to customers, as well as forward to controls.

Indicators and stress losses

Two items in the second plan were quantitative. A key risk indicator program had to be well controlled and linked to the firm’s risk identification and risk tolerance processes. A stress loss forecasting method had to quantify operational risks under stress. Both turn operational and compliance risk into numbers a board can track and a supervisor can test.

Building them takes data the firm may not have had in usable form: incident records, loss events, control failures, complaints and their causes. The indicators are only as good as those records. A firm that had never captured its operational losses consistently would need to start before its indicators could mean anything.

Compliance in change management

The plan also had to integrate compliance and operational risk requirements into business processes, control design and change management. That is a statement about when risk is considered. A new product, a new sales incentive or a system change would need a compliance review before launch, not after a problem appeared. For a large firm, that means a gate in every change process, with someone empowered to stop a launch.

Escalation that reaches the board

The board’s plan required a framework that identified, aggregated, evaluated and reported material risk issues, with plans to address them and progress against those plans. Aggregation is the hard part. Individual complaints, control failures and employee reports can each look minor. Only when they are combined do patterns appear. The letter’s account of sales practice misconduct described a pattern that management reported to the board in general terms. An escalation framework is meant to surface such patterns with enough detail for the board to act.

Acceptability as a standard

Both plans had to be acceptable to the Reserve Bank. That standard gave the supervisors the decision. A plan that met every listed item on paper might still not be acceptable if the supervisors judged it insufficient. The firm had to write plans that would persuade the supervisors, beyond meeting the list.

Paragraph 6 added discipline. Plans needed timelines with specific deadlines for each component. Once approved, they had to be adopted within 10 days and could not be amended without approval. Paragraph 7 added quarterly progress reports from the board. The plans became binding programs, with dates the supervisors could check.

Writing a plan supervisors will accept

A plan the supervisors accept has to be specific. Vague commitments to strengthen or enhance a function give a reviewer nothing to test. A plan with named owners, defined deliverables, dates and measures of success gives the supervisors confidence and gives the reviewer a checklist. Paragraph 6’s requirement for timelines with specific deadlines pointed the same way.

A plan is also a promise. Once approved, it could not be amended without the Reserve Bank’s approval. A firm that committed to dates it could not meet would face a choice between missing them and asking to change them. Both would cost credibility with the supervisors who decided when the cap came off.

Customer remediation as a program

The requirement to review how the firm remediated harmed customers made remediation a program in its own right. Identifying affected customers, calculating what they were owed, paying them and recording the result are operational tasks with data at their core. A firm that could not identify affected customers from its records could not remediate them, however willing it was.

That has a cost dimension the order did not state but implied. Remediation is paid from the firm’s earnings. Doing it well costs money up front. Doing it badly risks further findings and a longer cap.

Sixty days

Sixty days is short for plans of this scope. A firm could not design a new risk data framework in that time. It could set out how it would, with milestones. The order’s structure assumed the plans would describe programs that took much longer to complete. The cap’s duration would depend on how long those programs took and how convincingly the firm showed they worked.

The reviewers: independent experts as the gate

The third part was the review. Paragraph 4 required two independent reviews by third-party experts acceptable to the Reserve Bank. The first, the Initial Risk Management Review, had to be completed by an appropriate date no later than September 30, 2018. It would cover the board’s improvements and the enhancements to compliance and operational risk management. The second would assess whether the improvements were effective and sustainable once integrated into business as usual.

Why reviewers were the gate

The cap could not be lifted until, among other things, the first review was completed to the Reserve Bank’s satisfaction. That put the reviewer’s work at the center of the firm’s economics. Every month a review took longer, or reached a conclusion the supervisors did not accept, was another month under the cap.

That gave the review a value far beyond its fee. The firm’s cost of a slow or unconvincing review was measured in forgone growth, not consulting hours. That changes how a firm should think about buying such a review.

Buying independence

The order required reviewers to be acceptable to the Reserve Bank. A firm choosing a reviewer had to pick one the supervisors would trust, which in our reading argued against firms with conflicts or without the needed expertise. Thirty days before each review began, the firm had to submit a scope of work plan acceptable to the Reserve Bank. So the supervisors shaped both the reviewer and the work.

For procurement, that is unusual. The buyer pays, but a third party, the supervisor, decides whether the purchase is acceptable. The firm’s procurement process had to account for that: building the supervisor’s expectations into the selection criteria, the statement of work and the timeline.

Evidence the reviewer could test

A reviewer can only test what the firm can show. Plans, board minutes, risk reports, control test results, escalation records and compensation decisions were all evidence. If the firm’s records were thin, the reviewer could not conclude that the improvements were working. The second review’s question, about effectiveness and sustainability, needed records over time.

So the firm’s record keeping was part of its path out. A firm that built its governance changes with evidence in mind, recording decisions and tests as it went, gave the reviewer something to examine. A firm that improved its practices without documenting them would struggle to prove it.

Scope as a negotiation

The scope plan, due thirty days before each review, had to be acceptable to the Reserve Bank. That made scoping a three-way conversation among the firm, the reviewer and the supervisors. A scope that left out an area the supervisors cared about would not be accepted. A scope that included everything might take too long. Getting the scope right early saved time later, because a review the supervisors found incomplete would not open the gate.

Independence in practice

Independence is easy to claim and harder to show. A reviewer that also advised the firm on designing its remediation would be reviewing its own work. A reviewer whose other business with the firm was large might be seen as reluctant to deliver bad news. A firm choosing a reviewer the supervisors would accept had to look at those relationships first, and be ready to explain them.

Results to both supervisors

Results of each review went to the Reserve Bank and the Board within 30 days of completion. The Reserve Bank, with the supervision director’s concurrence, could require additional reviews by a third party or by internal audit. The firm could not treat a completed review as the end of the process. The supervisors could ask for more.

What the second review asked

The second review came after the improvements were integrated into business as usual. Its question was whether they were effective and sustainable. That is harder to show than adoption. It needs evidence over time: controls that kept working through staff changes, indicators that flagged issues and led to action, board reports that drove decisions. A firm that had adopted new processes but not run them long enough would not have that evidence yet.

Internal audit’s role

The order named internal audit as a possible performer of additional reviews. That points to a function a bank already pays for. An internal audit team with the independence and skill to test governance changes could reduce the firm’s dependence on outside reviewers over time. One without them would leave the firm buying outside reviews for longer.

The incentives: pay, accountability and the board’s own information

The fourth part was about behavior. The order and the letters addressed why the controls had failed, as well as what the controls should be.

Compensation and consequences

The board’s plan had to address any enhancements to the firm’s performance management processes needed to keep compensation and other incentives consistent with risk management objectives. That included appropriate consequences for violations of policy, law and regulation, and for adverse risk outcomes. The Board’s letter to the directors said the firm’s performance management and compensation programs had played a material role in the compliance breakdowns.

That linked pay to the cap. Incentives that rewarded sales without regard to risk had, in the Federal Reserve’s account, contributed to the harm. The order required the firm to change them, and the cap would stay until the governance around them improved.

Directors as signatories

The Board required each current director to sign the order. The press release said letters to each director confirmed that the board, during the period of compliance breakdowns, did not meet supervisory expectations. That put the directors personally on the record as parties to the remedy.

For directors at any bank, the signature is the point. A director who signs an order to improve board oversight has accepted, in writing, a duty to deliver it. The board’s own effectiveness became a subject of supervision, with the cap as the consequence.

Information to the board

The letter to the board described a specific failure. Citing an April 2017 investigation report commissioned by the board’s independent directors, it said that from February 2014 the board and some committees had received assurances from management that sales practice misconduct was being monitored and addressed. Management’s reports, it said, lacked detail and were not accompanied by concrete action plans and metrics. It said the board should have received more detailed and concrete plans.

That point about board information reaches every bank. A board that accepts general assurances on a serious issue, without plans and metrics, is relying on management’s word. The order’s reporting requirements were the remedy: information sufficient for the board to judge for itself.

Board composition and turnover

The press release said the firm would replace three directors by April and a fourth by year end. The board’s plan had to ensure that its composition, governance structure and practices supported the firm’s strategy and risk tolerance. It also had to ensure that no board role went unfilled for an undue period after a director left. Board turnover was part of the remedy, and managing it without gaps was part of the plan.

New directors bring fresh judgment and need time to learn the firm. A board in the middle of a remediation program needs both. Committee chairs, in particular, carry responsibility for the oversight the order required. A board that planned succession for those roles reduced the risk that the plan stalled when a director left.

Quarterly progress reports

Paragraph 7 required the board, or an authorized committee, to submit written progress reports to the Reserve Bank within 30 days after each calendar quarter. The reports had to detail the actions taken to comply and their results. That put the board’s name on a quarterly account of progress. A board that could not describe results, only activity, would be telling the supervisors the program was not yet working.

Cooperation on individual conduct

Paragraph 8 required the firm to cooperate with the Federal Reserve in investigations of whether separate actions should be taken against individuals involved in the underlying conduct. That kept individual accountability open alongside the firm-level remedy.

How this applied across institutions

The order bound one firm. Its design carried lessons for others, depending on size and supervisor.

Large bank holding companies

The order cited Regulation YY’s risk management requirement for large bank holding companies. Peers subject to the same rule faced the same expectation of a firmwide risk management framework, with an independent risk committee. The order showed what supervisors might do when they found that framework wanting: cap growth until it was fixed. For those firms, the economic value of governance was now visible. Their boards could ask a pointed question: if a supervisor imposed a similar condition, what evidence could the firm produce, and how long would it take to satisfy an outside reviewer? The answer would show the gap between governance as practiced and governance as provable.

Regional and community banks

The Regulation YY section the order cited now sits in the regulation’s subpart for bank holding companies with $100 billion or more in assets, so smaller banks are outside it. The specific remedy, a cap on consolidated assets, was tailored to a large firm. But the underlying points applied to any bank: sales incentives need risk controls, boards need detailed information, and remediation plans need evidence. A regional bank’s board could take the letter’s description of thin management reporting as a test of its own.

Credit unions

Credit unions are not supervised by the Federal Reserve, and the order did not apply to them. The governance lessons about incentives, board information and evidence of control effectiveness did not depend on the supervisor.

Broker-dealers and wealth managers

Firms in the securities business face their own rules on sales practices and supervision. The order’s emphasis on incentive design and on escalation of conduct risk to the board is relevant to any firm that pays staff to sell. A broker-dealer affiliated with a bank holding company sits within the holding company’s firmwide risk framework.

Insurers and other financial groups

Insurance groups and other financial firms are supervised under different regimes. The order’s specific remedy did not reach them. Its logic still applies wherever a supervisor can restrict growth or new business until governance improves. Any regulated firm whose supervisor holds that kind of lever has a reason to treat governance evidence as an economic asset.

Fintech partners and service providers

Firms that sell services to large banks saw a different effect. A bank under a growth cap reallocates its balance sheet and scrutinizes its programs. Partnerships that added assets would compete for room under the cap. Vendors of risk data, control testing and governance services would see demand from banks building the capabilities the order described.

Objections, in order

We take five objections, numbered in the order we address them.

1. A growth cap is an unusual remedy, so it says little about other banks

The first objection is that the cap was exceptional. The remedy was exceptional. The underlying expectations were not. Board effectiveness, independent risk management, risk data and incentive design are expectations the Federal Reserve applies to large firms, as its August 2017 proposal showed. The cap showed what failure to meet them could cost. Other firms did not need to expect the same remedy to learn from its design. The question it raised, how a firm would prove its governance works, applies at any size.

2. The cost was only forgone growth, not a real loss

The second objection is that the firm lost nothing it already had. Forgone growth is a real cost to shareholders and to a firm’s competitive position. It is also a continuing cost, paid every quarter the cap remains. A one-time fine ends when paid. A cap runs until the conditions are met. That makes its total cost uncertain at the start, which is itself a burden for a firm trying to plan.

3. Third-party reviewers are a formality

The third objection is that outside reviews are routine. Under this order, they were the gate. The Reserve Bank had to accept the reviewer and the scope, and could require more reviews. A review the supervisors did not find convincing would extend the cap. That made the review a high-stakes purchase, not a routine one. A firm that chose its reviewer on price, rather than on the supervisors’ likely confidence in the work, could save on fees and lose on time.

4. Boards cannot manage operational detail

The fourth objection is that boards should not be drawn into operations. The order did not ask the board to run operations. It asked the board to ensure management built the functions, data and reporting the board needed to oversee them. The letter’s complaint was about the quality of information the board accepted, which is squarely a board responsibility. Asking for plans, owners, metrics and dates is not operational management. It is how a board tests whether management is managing.

5. Compensation design is a business decision

The fifth objection is that pay is management’s choice. It is, within limits. The order treated incentives that rewarded sales without regard to risk as part of the risk management failure. A bank can design its own pay, and supervisors can still judge whether that design undermines controls. The letter’s account of how sales targets and incentives contributed to the harm shows why supervisors treat incentive design as part of risk management.

What a leader could check

These checks follow from the four parts. They are our suggestions, not requirements.

  • Does the board receive risk reports with specific plans, owners, metrics and dates, rather than general assurances, on each material conduct issue?
  • Are sales and performance incentives reviewed by the risk function, with consequences for policy breaches built into pay decisions?
  • Could the firm show an outside reviewer, from its own records, that a governance change is working, beyond showing that it was adopted?
  • Does internal audit have the independence and skill to test governance and risk management changes, so outside reviews are not the only source of assurance?
  • If the firm needed an independent review acceptable to a supervisor, does it know how it would select, scope and schedule one?

Each check also has a measure. A board can count the material issues reported to it with a dated plan and an owner. Risk can count the incentive plans it reviewed and the pay decisions that reflected breaches. Audit can count the governance changes it tested. Those counts do not prove good governance. They show whether the habits that produce evidence of it are in place.

The checks need owners. Board reporting belongs to the corporate secretary and the chief risk officer, with the board committees. Incentive review belongs to human resources and the compensation committee, with risk management. Evidence of effectiveness belongs to each control owner and to the second line. Internal audit belongs to the audit committee. Readiness for an outside review belongs to the chief risk officer and procurement together.

With hindsight

The cap lasted far longer than the order’s first review date suggested. On June 3, 2025, the Federal Reserve announced that Wells Fargo was no longer subject to the asset growth restriction. It said the firm had met all the conditions the 2018 action required for removal. Those conditions included improving governance and risk management and completing a third-party review of the improvements. The Federal Reserve said it had completed its review of the firm’s remediation and the required third-party assessments, and its own assessment of the firm’s governance and firmwide risk management.

The same release said the other provisions of the 2018 order would remain until the firm satisfied the requirements for their termination. So lifting the cap did not end the order. Its other provisions, including the second review on effectiveness and sustainability, remained.

On March 5, 2026, the Federal Reserve Board announced that it had terminated the 2018 enforcement action. It said the firm had been required to demonstrate that its improvements made its governance and risk management program effective, and to complete two third-party reviews. It described the remediation work as having spanned nearly a decade. It noted that the asset cap had been removed in 2025.

Those two releases are consistent with the order’s design as we described it. The cap came off after plans, implementation and an accepted third-party review. The rest of the order ended only after a second review addressed whether the changes worked over time. The reviewers were, in effect, the gate.

The Federal Reserve also turned its board expectations into guidance. On February 26, 2021, it issued SR 21-3, supervisory guidance on board effectiveness, for bank and savings and loan holding companies with $100 billion or more in assets. The guidance described five attributes of an effective board. They include setting a clear direction on strategy and risk appetite and directing management on the board’s information needs. They also include holding management accountable, supporting independent risk management and internal audit, and maintaining a capable composition and governance structure.

Several of those attributes map onto the order’s board plan. The second attribute, directing senior management to provide information sufficient in scope, detail and analysis for sound decisions, reads as the general version of the letter’s complaint about thin management reporting. The guidance said supervisors would use it in assessing governance and controls under the large financial institution rating system.

For a bank leader today, the economics are the lasting point. The order showed that supervisors can attach a running cost to weak governance, and that the cost stops only when evidence convinces outside reviewers and then the supervisors. Investment in governance evidence, such as board reporting, control testing and records of remediation, is therefore an investment in shortening any future remedy, whatever form it takes.

What held up is that governance was given an economic price, and that evidence reviewed by outsiders decided when the price stopped. What the February 2018 documents did not signal was the duration. The order set September 30, 2018, as the latest date for the first review. The cap came off in June 2025, and the order itself ran until March 2026. For any bank, the length of that path is the measure of how hard it is to prove that governance has changed.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • Federal Reserve
  • enforcement
  • governance
  • risk management
  • third-party review