NOTE-1095: Sales Practices Investigation Report

Contribution

The company’s own causal account of the Wells Fargo sales-practices episode, written for its independent directors by outside counsel. It had access to 35 million documents and 100 interviews, and it reviewed 1,341 internal investigation records of terminated employees (pp. 2–3, 35). It answers where the harm came from, how the company explained it to itself at the time, and why the people with authority did not act on what the organization knew.

Key insight

The report rejects the account that the company’s leaders gave at the time, in which a few bad employees caused the harm. In its place it puts a system: unattainable sales goals, enforced by rankings, scorecards and fear for one’s job, sustained by a structure in which the business line ran itself, “run it like you own it”. The report’s own words for the leadership’s error: “Effect was confused with cause” (p. 6).

Assumptions

• The investigator’s position. The board’s independent directors commissioned it, and the board is one of its subjects. Shearman & Sterling was “determined to be independent of Wells Fargo” (p. 2).
• The company’s data. FTI Consulting “relied on data provided to it by Wells Fargo” (p. 26, n. 8) and “did not independently test” the population of potentially unauthorized accounts (p. 26, n. 7).
• Missing witnesses. Tolstedt “declined, on advice of counsel, to be interviewed” (p. 8, n. 1). So did Paterson, the Community Bank’s HR head (p. 80, n. 24), and Bacon, head of Corporate Security (p. 90, n. 27).
• A non-random sample. The 1,341 records were not statistically sampled, and the review “disproportionately focused on managers” (p. 36).

Key results

• Root cause. “The root cause of sales practice failures was the distortion of the Community Bank’s sales culture and performance management system, which, when combined with aggressive sales management, created pressure on employees to sell unwanted or unneeded products to customers and, in some cases, to open unauthorized accounts” (Overview). None of the ingredients is condemned alone: “there is nothing necessarily pernicious about sales goals, a sales-oriented culture or a decentralized corporate structure”, but they “coalesced and failed dramatically here” (p. 4).
• Pressure, not pay. “Employees who engaged in misconduct most frequently associated their behavior with sales pressure, rather than compensation incentives” (p. 7). Plans were “commonly referred to as 50/50 plans”, expected to be met by only half the regions (p. 19). Turnover ran at least 30% a year from 2011 to 2015 (p. 27).
• Misconduct rose with the goals. Sales-practice allegations rose from 288 in Q2 2007 to 1,469 in Q4 2013, and terminations or resignations from 61 to 447, then fell after 2013 as goals were moderated (pp. 33–34). The rate at which new accounts were funded fell from about 90% in 2005 to below 80% in 2012 and exceeded 95% by 2016 (p. 6).
• Scale. About 5,300 employees were fired for sales practice violations between 2011 and March 2016 (p. 16). The CFPB order’s population is “1,534,280 deposit accounts that may not have been authorized” (p. 26, n. 7). Only 1.7% of the unfunded accounts incurred fees that were not reversed (p. 43, n. 18). Removing all the suspect accounts moved the cross-sell metric by at most 0.04 in any quarter (p. 43, n. 19).
• Little explicit direction. Most terminated employees admitted the misconduct and blamed sales pressure. Branch managers “only rarely” said they were explicitly told to misbehave, and in nine of the 1,341 records a district manager was accused of explicitly directing it (p. 37). The report also documents regional leaders in Los Angeles, Florida and Arizona who applied extreme pressure and urged selling regardless of need (pp. 22–26).
• Leadership’s explanation. “It was convenient instead to blame the problem of low quality and unauthorized accounts and other employee misconduct on individual wrongdoers and poor management in the field rather than on the Community Bank’s sales model” (p. 5). Stumpf’s “reaction invariably was that a few bad employees were causing issues” (p. 10). In 2015 he wrote “This is not systemic” (p. 56). In May 2015 the Risk Committee was told that “the root cause was intentional employee misconduct, not systemic issues arising from sales goals or compensation” (p. 105).
• Structure and control functions. The Chief Risk Officer “did not have any line authority or directive power to enforce changes on the lines of business” (p. 60). The control functions took a “transactional” approach, case by case (p. 13). Corporate HR had “no coordinated effort … to track or size the sales practice issues” (p. 79). Audit rated the Community Bank’s culture “Strong” in 2013 (p. 94). The term “customer harm” was “narrowly construed to mean only financial harm such as fees and penalties” (p. 14).
• Board. Sales practices were not reported to the board as a noteworthy risk until 2014 (p. 15). The board found its own conduct could have been “improved in three respects”: earlier centralization of risk, more concrete plans demanded of management, and more pressure on Stumpf to replace Tolstedt (pp. 16–17).
• Sanctions and remedies. Product sales goals were eliminated from October 1, 2016 (p. 51), and incentives were rebuilt around customer service (pp. 8, 52). Risk and HR were centralized, 4,100 risk employees in 2016 and 1,100 more in 2017 (p. 12). Five Community Bank executives were fired for cause (pp. 8–9, Overview). The board recorded forfeitures and clawbacks of $19millionplusabout$47.3 million in options for Tolstedt (p. 9), about $41millionplusa$28 million clawback for Stumpf (p. 10), and about $32millionforeightOperatingCommitteemembers(p.14).Intotalthatis"morethan$180 million” (Overview).

Signals and authority

The report traces, repeatedly, the path from the people who knew to the people who could act:

• Known early, not escalated. In August 2004 Internal Investigations reported that “The incentive to cheat is based on the fear of losing their jobs for not meeting performance expectations” and recommended cutting or eliminating sales goals. “There is no evidence that the report and its recommendations were further escalated” (pp. 89–90).
• Reaching the top without being acted on. Stumpf “received numerous customer and employee complaints about sales practices and sales pressure, which he or his assistants referred to appropriate subordinates without further follow-up” (p. 55). Told that about 1% of Community Bank employees were fired each year, he “reacted positively” and does not appear to have “initiated any follow-up investigation or inquiry” (p. 55).
• Dissent overruled. Regional leaders’ objections to the goals were viewed “as simply a cover-up for poor management” (p. 41). Tolstedt told Stevens “not to talk to Loughlin anymore and that she needed to ‘toe the line’” (p. 65).
• Reported, then filtered. Bacon wrote in 2012 of “a concerning trend in the area of Sales Integrity – from the increase in EthicsLine reports, to the increase in executive complaint letters”. Russ Anderson told him his reporting made the problem sound “so much worse than it is” (p. 50). Audit received the same email and “did not follow up” (p. 95). Before the May 2015 Risk Committee meeting, the 1% figure was deleted (p. 105). Internal Investigations had counted 1,229 and 1,293 terminations for 2013 and 2014, and those numbers “went unmentioned” (p. 106).
• Seen, but filed as minor. In the A&E Committee’s materials before 2014, sales integrity matters sat with internal matters “viewed as having minimal, if any, customer impact” (pp. 99–100).
• Supplied from outside. The press, the Los Angeles City Attorney and the regulators forced the issue. The board “learned for the first time” of the 5,300 terminations through the September 2016 settlements (p. 16).
• Retaliation not found, so far. A “limited review completed to date” found no “pattern of retaliation” against employees who complained, and the review was ongoing (p. 87, n. 26).

Claims

id	claim	strength	support
C1	The root cause was the sales culture and performance management system, with aggressive sales management, not individual misconduct alone	moderate	investigators’ judgment from documents, 100 interviews and 1,341 investigation records; key witnesses declined interview
C2	Misconduct tracked the difficulty of the goals	moderate	ICE allegation and termination trends 2007–2016 and funding rates (FTI, on company data); correlational
C3	Employees attributed misconduct to sales pressure more than to pay	moderate	review of investigation records, which the report itself calls possibly “self-serving” (p. 38)
C4	Decentralization and deference kept control functions from seeing or acting on the problem	moderate	narrative of reporting lines and episodes, pp. 11–14, 60–96
C5	The board was not told the aggregate terminations until 2016	strong	board materials reviewed back to 2009, and the deleted 1% figure (pp. 97–110)

Concepts

• sales practice (sales integrity) violation — the report’s umbrella for gaming the sales system, from false referral credits to opening accounts without consent, simulated funding and falsified records (pp. 35–36).
• simulated funding — moving money from one customer account to another, sometimes unauthorized, account so that it looked funded (p. 26).
• Rolling Funding Rate — the share of new accounts funded with more than a de minimis amount, the Community Bank’s own quality metric (p. 21).
• “run it like you own it” — the decentralized model’s mantra, giving lines of business control of their own risk and HR functions (p. 19).

Connections

The report cites no literature. It is a primary document of corporate self-investigation. Its account of structure and culture can be read against French’s corporate decision structure (THEORY-131) and Pettit’s conversable corporation (THEORY-139). It is the source of the documented case CASE-151.

Bearing on the record

The report supports the narrower reading of the “psychopathic corporation” thesis on several points. The harm is traced to the sales model and the system that enforced it, not to bad character. The CEO and the head of the Community Bank are faulted for clinging to that model and for misreading it as a few bad apples. Within the company, the costs to customers were narrowed to fees and seen case by case. The board learned the full scale from regulators.

It also complicates the half of the reading on which nothing in the corporation registers the costs it imposes. The company did register them, in its investigators’ 2004 report, its HR project of 2011–2012, its regional leaders’ objections and its EthicsLine data. What failed was the route from those signals to anyone with authority, and the authority to act on them. That is CLAIM-349’s separation of access, authority and action, documented in one firm. The report does not attribute the harm to the pursuit of profit as such. The leaders it faults were protecting a sales model and its growth figures (p. 5), and the unauthorized accounts made little money (p. 43, nn. 18–19).

Limitations

• A self-report. It is the company’s board reporting on the company, the board included. Its findings on the board are mild. The same committee was weighing shareholders’ derivative demands on the board (p. 2), and the report was filed with the SEC as proxy material.
• No intent established. It infers motive: Tolstedt and her risk officer “likely” kept information from the board “to give themselves freedom to address these issues on their own terms, rather than to encourage improper behavior” (p. 8). But Tolstedt was not interviewed.
• Not a finding of no individual fault. It names leaders, fires five for cause and claws back pay. Thousands of employees were fired, and the report does not say they were blameless.
• Scale is the regulators’. The account and termination counts come from the settlements and from company data, untested by FTI.
• Retaliation review incomplete. Its retaliation review was “ongoing” when the report was published (p. 87, n. 26).

Open questions

Whether the outside investigations confirm this report’s allocation between system and individuals, such as the later regulatory and criminal actions against named executives, none of which is read here. And whether the company’s later conduct bears out the remedies it lists.