When John Stumpf, Chairman and CEO of Wells Fargo, appeared before the Senate Finance Committee recently, he represented a forlorn and distracted leader some distance from the financial expert who was summonsed to Washington eight years ago by Alan Greenspan to help craft the Government’s response to the 2008 financial crash.

In the period since 2008, he has presided over unprecedented company growth – including the acquisition of Wachovia, in what was the world’s largest ever financial services merger. Wells Fargo became the world’s biggest bank and John was peerless in his sector. The great Warren Buffet commended his leadership, and expressed such confidence that he became a major investor and Board member. Anyone with an aspiration to make it big in banking wanted to work for Wells Fargo, and they became a magnet for talent.

It would seem as though the “too big to fail” mantra gathered credibility in the shadow of this icon of American capitalism. So what went wrong?

According to Senator Elizabeth Warren, it came down to leaders allowing a toxic and corrupt culture to develop. Other Senators concurred, touting words like “greed” and “arrogance”. It is difficult to explain how over 5,500 employees – “team members” in bank parlance – could have made the simultaneous, almost coordinated, bad decisions that led to the whole edifice crumbling down.

There had to be a reason. As Senators screamed of fraud, steadier analysts looked for subtler explanations. Those explanations certainly didn’t come from Stumpf’s lackluster testimony. He was simply awful. Stumbling from bumbling excuse to weasel rationalization, he failed to account for this behavior, never mind explain it. The formerly assured and confident leader was a wreck–an embarrassment.

The central actor, it transpired, was Carrie Tolstedt, head of the Community Bank division, under whose watch this whole thing played out. Since the LA Times 2013 exposé, and the subsequent internal investigation, Tolstedt had been living on borrowed time. In his Senate testimony, Stumpf announced her planned retirement and payoff at the end of this year in an attempt to draw a line under the whole affair.

Far from being the sacrifice he hoped, the terms of her departure simply incensed those who saw her as culpable. The Senators smelled blood, and some saw an opportunity for payback– a settling of past scores. How could Tolstedt be “rewarded” by such outrageously generous terms and allowed to walk free of scrutiny and responsibility?

“But she has given the bank such outstanding service through her long and successful career!” whined Stumpf in her – and his – defense. If ever a leader demonstrated how out of touch with the zeitgeist he was, it was Stumpf in that moment. Loyalty trumped objectivity, and there was to be no second chance, no recovery.

Did Carrie Tolstedt instruct bank employees to set up fictitious accounts for customers in order to defraud them? No. Did the bank set up covert training programs with the intent of stealing their customer’s money? Absolutely not. The story that conspiracists wish to be true clearly isn’t. The likely truth, however, is much more worrying. Rather than a few bad apples behaving badly, here we have principled and well intentioned executives behaving in ways where this outcome became possible, even inevitable. And they were clueless to prevent it.

It all comes down to leadership and how we model the behavior we regard as important.

Consider the following scenario. Let’s say Carrie Tolstedt exhibited strong and purposeful leadership. Wickedly smart, she would know what needed to be done to solve complex problems before most would recognize one existed. She would be confident in her analysis– so much so, that few could outthink or outsmart her. If you wanted to succeed, you would avoid battles with a leader like Tolstedt at all costs, and engage only in those on the periphery– where expressing a strong point of view carried few personal risks.

Tolstedt would likely be oblivious to this. Indeed, if she knew this was how people were behaving, she would have been appalled. She would have wanted the opposite. She would have enjoyed the opposite. But she would have believed the charade was real.

In this environment, you would see the rules of the game very differently than Tolstedt. Experience would show that if you argued against her ideas, you may not find the defense easy to maintain. After all, there is no fun in having your carefully crafted argument ripped to shreds. Her reactions would quickly inform you whether you were enhancing your prospects or not. Over time, you would decide that it wasn’t worth the effort and simply give up. You might decide to follow the path of least resistance, or worse– you could go along with the views of your leader – even if you disagree – because protecting your personal standing and potential for advancement would be more important, and these contentious issues might not be so disagreeable after all.

The result? An echo chamber.

Miraculously, every idea Tolstedt developed would become the best one– supported and encouraged by those around her and effectively prosecuted and executed. To borrow a well-used analogy, if you show up and your only tool is a hammer, it’s remarkable how so many things resemble nails. Not only had this situation become inevitable, it was eminently predictable.

So how plausible is this scenario?

It is much more common than you might think. Leaders might talk about inviting dissent, but few behave ways that encourage it. One of the most improbable factors in the Wells Fargo situation is that rumor of these false accounts – and the deliberate behavior that created them – didn’t rise up through the ranks. It likely did. It would be nearly impossible for 5,500 employees to behave in undetected ways.

There are two important questions you must ask. First, at what level in the communication chain did awareness of this behavior reach, and why wasn’t it transmitted any further? Wherever in the chain of command that block occurred is the point where the risk of disclosure outweighed the benefit. The leader sitting at this precise point in the pyramid decided it was too risky.

Another important point relates to the location of these 5,500 “bad apples”. We understand that a small number of regional leaders were terminated in this affair, yet the 5,500 employees were spread much more widely across the community bank.

If the chain of communication ended at too low a level in one part of the bank, why didn’t it rise higher in another? Surely, somewhere in the chain of command the information would break through to the level where someone would be able to act. Unless a level of fear governed the entire operation. This might be the most parsimonious explanation and it is a direct result of leadership culture and why Tolstedt had to go.

Stumpf made two mistakes that cost him his job, his career, and his reputation.

First, he failed to terminate Tolstedt upon discovery of the fraud in 2013. Executives who have developed such strong relationships through shared battles and successes are susceptible to this error. Stumpf failed to detach his feelings for the person from his accountability as Chief Executive. Despite describing himself to Senators as a “courageous” leader, he could well have been intimidated by Tolstedt’s standing in the bank – a reputation he undoubtedly helped cultivate. Far better to provide an easy way out – with a huge package to sweeten the pill – than to confront the obvious reality of a failed executive and oversee a difficult termination.

The persistent presence of team members’ fraudulent behavior over such a long period of time told him everything he needed to know in order to act. If any executive can allow this behavior to escape notice or investigation, or to continue unchecked after discovery, then they represent a liability to the reputation of the company. It’s really that simple – and so should have been the decision – yet Stumpf choked.

Second, Stumpf failed to appreciate his own culpability and ultimate responsibility. He appeared tone deaf to the demanding questions of his political interrogators. As they prosecuted their case, Stumpf’s claims of the positive and powerful culture at Wells Fargo were irrelevant. It didn’t matter whether the remaining 260,000 employees were model citizens with impeccable behavior. These Senators weren’t going to allow distracting facts to get in the way of a good story.

Stumpf should have taken pre-emptive action to refuse his own bonus– a bonus that politicians directly linked to the ill-gotten gains of fraudulent behavior. Stunningly, he didn’t. Despite being Chair of the Board, he punted this responsibility in their direction. In doing so, he completely compromised their range of options. When the Board did act, they corrected both errors emphatically. They fired Tolstedt with no severance, and they surgically removed Stumpf’s discretionary remuneration. Both steps that Stumpf should have taken– steps which inflicted his ultimate humiliation. And now, with predictable precision, he too has lost his job. Rightly so.

While hindsight points to what obviously should have been done, few should feel confident that executives of other companies will learn from this lesson.

The current nervousness sweeping boardrooms with anxious discussions and reassuring executives will count for nothing. The problems are far too deep seated and human nature far too imperfect. What passes for rigorous debate in most companies is laughable. CEO’s eager to make their mark and show evidence of their effectiveness cherry-pick their balance sheets. They call for a strong point of view and then allow their insecurities to close down real dissent, quietly decapitating emerging poppies in the meadow. Furthermore, their innate biases lead them to select other executives who think and act just like them– strangling valuable attitudinal divergence at the source. That they don’t see themselves doing it, while demanding the very opposite, provides no evidence that this is actually happening.

It’s Wells Fargo today, but it could well be your company tomorrow.

Download a copy of this article.