Charles de Gaulle said, “The graveyards are filled with indispensable men.” That rings especially true despite the pathetic cries of the ‘Bring Jensen Back’ bros.
Nvidia’s strong earnings release today proves once and for all that Nvidia does not need Jensen Huang to return to flourish as a strong American national champion. Interim CEO Colette Kress deftly raised forecasts moving forward, pointing to the underlying strength of the business, which is far greater than any one individual, while winning plaudits for running the earnings call with command, presence, and impressive candor. And indeed, new board chair Harvey Jones definitely closed the door on Jensen Huang’s return as CEO amidst news that the SEC is now probing Huang and Nvidia.
All this makes the unresolved governance questions and lessons from the Nvidia debacle even more important.
In the aftermath of Jensen Huang’s ouster from Nvidia this week, a tiny band of tech bros, perhaps intoxicated by their own sudden wealth and prominence, have cast good governance principles to the wind by unrealistically advocating for the return of Jensen Huang, decrying the “woke” politically correct management policies which they believe led to Huang’s exit. At the same time, a chorus of finger-pointing good governance advocates has joined the fray, throwing every pet theory of organizational structure and regulatory restraints on risk-seeking entrepreneurs. Both camps ignore the real collapse of board oversight, which is the tragically poor judgement of the CEO matched with the faulty character and faulty process of board oversight and not its clunky bureaucratic mechanics.
The tech bro cheerleaders rooted for a messianic return of Huang, despite the ongoing investigations into his wide array of apparently corrupt actions and character collapses, such as his blatant violations of Nvidia’s code of conduct and unresolved national security concerns. The arrogant, naïve fantasies in the chants of these tech bros is encapsulated by Telegram founder Pavel Durov’s X post declaring, “Jensen made a rookie mistake. He had a board”, apparently forgetting that any executive who is entrusted with other people’s money owes a fiduciary duty to their shareholders first, above their own self-enrichment. They are happy to snag the money but then to dodge accountability.
Then, on the other end of the critics’ spectrum, good governance advocates, self-declared governistas, have rushed to offer a smattering of bad solutions for Nvidia’s governance woes. These proposed solutions are as ludicrous as they are trite, and as irrelevant as they are ineffectual.
For example, Charlie Elson at the University of Delaware somehow baselessly managed to blame a favorite bogeyman of many governistas, dual-class shares, for Nvidia’s woes — even though Nvidia has a single-class stock structure and not dual-class shares at all, a rather inconvenient fact which nevertheless did not detract from Elson launching into his tiring tirades. Elson bewilderingly argued in Directors & Boards that “in the end, this may spark, and I hope it does, a real reconsideration of the whole notion of dual-class stock in publicly traded companies”, as if that would be a panacea for all of Nvidia’s board woes, an argument which The New York Times also haplessly repeated.
Similarly, governance professor Richard LeBlanc of York University proposed an independent chair as the panacea to all Nvidia’s governance woes, even though Nvidia already had an independent lead director who was the former CEO of Cooley Law Firm, Stephen Neal.
All of these proposed governista solutions are irrelevant cliches offering insufficient and misguided recipes.
Between clumsy New York Times op-eds, delusional Silicon Valley social media campaigns, and even an ideologically charged, anti-management Delaware Chancery Court judge, it appears that both anti-oversight tech bros and scolding governistas need to relearn some basic lessons about corporate board diligence.
Here’s what these self-anointed governance experts miss and what the real lessons for boards from the Nvidia crisis are:
- Don’t build your business on other people’s money and then later resent accountability to them
- Don’t fall for mindless metrics which miss the actual process of governance — much like the way the “Goodhart’s Law” of economics warns over confusing the metric with the underlying phenomena
- Don’t confuse capital structure, whether dual-class shares or single class shares, as indicators of board caliber, board culture, and genuine independence
- Don’t fall victim to vague admonitions over “good faith” as genuine definitions of proper board diligence.
- Don’t rely on your accountants and attorneys to protect you – as the buck stops with you.
The checklist fallacy that ISS, Glass Lewis, and many good governance advocates have fallen into
All of the misguided proposed solutions from governistas tend to draw from the same wrongheaded playbook that ISS and Glass Lewis have preached for decades, recited on cue by the governistas: split the chair and CEO roles, encourage more so-called independent board members, tally the committees, and count board diversity. As we have argued in Fortune, this tired and trite playbook has never been able to tell a good board from a bad one.
Consider how Nvidia was an ISS and Glass Lewis box-checker’s showpiece. Every director but Huang was technically independent. It had an independent lead director who was a former CEO of a respected law firm; independent audit, compensation and governance committees and committee chairs; annual elections for every board position without any staggered terms; and the ‘one vote per share’ principle so beloved by governistas. Huang was not even chairman of his own board: Nvidia had no chairman at all until he left, so Nvidia even fit the technical separation of roles box with a separate lead independent director. In June, shareholders re-elected all ten directors, the CEO with 99% of the vote, with ISS supporting the entire director slate this year, as it has each year for many years now.
But the Nvidia crisis reminds us all just how misleading are the metrics of ISS and Glass Lewis’ checklist. This ‘check-the-box’ approach does absolutely nothing to actually weed out genuinely responsible boards from boards asleep at the switch operating on a bedraggled autopilot. Indeed, it is ironic that many of the worst corporate governance crises this century have been at companies which were darlings of the governista community.
For example, Enron had 12 independent directors among its 14 directors, split the chairman and CEO roles in its final year, and had a former dean of Stanford’s business school, an accounting professor, chairing its audit committee, winning widespread praise from governance advocates with one governance publication even naming the Enron board one of America’s five best boards in 2000 – right before Enron blew up, exposed as a fraud. Similarly, Silicon Valley Bank had an independent chair and a board that was entirely independent other than its CEO, and 10 of Lehman Brothers’ 11 directors were independent.
Indeed, when Stanford researchers tested ISS’s then-flagship governance rating against five measures of how companies actually fared, including restatements, shareholder lawsuits and stock performance, they found no significant correlation with any of them.
The first author wrote in Harvard Business Review in 2002, “In the wake of the meltdowns of such once great companies as Adelphia, Enron, Tyco, and WorldCom, enormous attention has been focused on the companies’ boards … And yet a close examination of those boards reveals no broad pattern of incompetence or corruption. In fact, the boards followed most of the accepted standards for board operations: Members showed up for meetings; they had lots of personal money invested in the company; audit committees, compensation committees, and codes of ethics were in place; the boards weren’t too small, too big, too old, or too young.”
Alas, the reliance upon such banal bromides of supposed good governance, lead to the recitation of platitudes in publications but worse yet miss the targeted need entirely – much the way the Goodhart’s Law of economics warns us to not confuse the metrics of the tool with the assessment of the underlying phenomena itself. This is also parallel to intelligence testing, where the test is an indicator of performance against the unidimensional artificial metrics of the intelligence test as an instrument itself and misses the many other untested dimensions of intelligence as Harvard psychologist Howard Gardner has pointed out in his classic Frames of Mind analysis.
The dual-class red herring
Another favorite bogeyman of the governance community tends to be dual-class shares, and this long-held bias is once again rearing its ugly head with governance commentators rushing to blame dual-class shares for the Nvidia crisis – even though, ironically and inconveniently, Nvidia only has one class of stock, and Nvidia has never actually had dual-class shares at all.
Indeed, ISS now generally universally recommends against directors at companies with unequal voting rights, without exception, reflecting how widely held this erroneous strain of thinking is entrenched in some circles of the governance community.
By way of full disclosure, we are sometimes wrongly identified as an ‘advocate for dual-class structures’, in the words of Vice Chancellor Travis Laster of the Delaware Chancery Court, who has been overturned by the Delaware Supreme Court no less than 35 times in his idiosyncratic and tortured readings of the law.
But we are not the universal dual-class share advocate Laster misleadingly portrays us as. Our view is much more nuanced and captured in the conclusion of my Harvard Law Corporate Governance Forum piece: “In evaluating the outperformance of dual-class shares and in evaluating common scenarios where dual-class shares work especially well, balanced against common pitfalls, the evidence is clear that for companies, there is no single right answer between dual-class, multi-class, or single-class shares. It is all situational, dependent on a company’s circumstances, the demonstrated caliber of their leadership, and the environment in which they operate. Just as dual-class shares are not the right structure for all companies; dual-class shares can make sense for certain companies, in certain scenarios. These crucial nuances are often lost in what has increasingly become the single-minded zealotry of ‘one share, one vote’ advocates, seeking to impose single class structures on all companies irrespective of circumstance.”
Indeed, what critics of dual-class shares like Laster and Elson miss is that some of the worst corporate collapses and frauds of the modern era had ‘one share, one vote’ share structures, ranging from Enron, Worldcom, Tyco, Arthur Andersen, Bear Stearns, Lehman Brothers, Wirecard, FTX, Silicon Valley Bank, and First Republic Bank, all of which share a single commonality: they did not have dual-class shares. Thus, critics like Charlie Elson who blame dual-class shares for Nvidia’s woes are committing the sin of governance by non sequitur. It is like prosecuting a car company for making potential getaway cars for a robbery which was actually committed on foot. Clearly, a formula that indiscriminately, automatically scores Berkshire Hathaway’s share structure as a vice and Nvidia’s as a virtue, without human discernment or space for an iota of discretion, is measuring based on the wrong yardstick.
What actually failed at Nvidia, and what it reveals about governista blind spots
What failed at Nvidia was not any of the factors identified above; rather, it was the caliber of the directors and the culture amongst the board that proved problematic.
As the first author wrote in the seminal Harvard Business Review article in 2002, ‘What Makes Great Boards Great’, great boards are social systems before they are governance structures. They build enough trust that management brings bad news early, with open and candid discussions and no secrets kept. Great boards treat dissent as a duty, and promote cultures where questioning authority is not only welcome but expected. They rotate roles so that no director is typecast as a permanent cheerleader or scold.
As my former colleague, the former Yale School of Management Dean and former SEC Chairman Bill Donaldson said at a Yale CEO Summit many years ago, “On the subject of board governance… corporate governance has become a cottage industry with lots of opportunists, and there’s some very superficial thought out there about what a board really is…..The least examined is that it is a social entity. It has human beings on it. Nobody talks much about that. Everybody says, ‘Oh, you ought to have certain people on the audit committee because they’ll be independent, etc., etc.’
“There hasn’t been nearly enough work done on the social contract on the board and the recognition of what is really going on, on that board. And a lot of times, what’s really going on bears the seeds of horror stories. Perhaps there’s a group that’s aligned improperly, and that group dominates either through tenure on the board or whatever and allows bad things to go on because nobody really recognizes what those interpersonal relationships are.”
Donaldson emphasized that great boards hold each director accountable and grade their own performance as hard as the CEO’s, and this gets to the caliber of the board as well as the culture of the board. None of this works without directors of the stature to look a monarch in the eye. As I wrote in The Hero’s Farewell, monarchs are larger than life CEOs who would leave office only by death or palace revolt. With a visionary founder, a board’s job is to retain and contain: keep the genius, check the hubris.
But Nvidia’s board lacked the stature to contain or check Huang’s impulses; they could only retain and worship. Evaluating the board members individually, no Nvidia director has ever run an S&P 500 company and none has national security experience, at a company whose chips are a national security asset. Industry peers Microsoft, IBM, and Dell each have at least four current and former big-company CEOs on their board; Nvidia had none. And apparently, the board was asleep at the switch on succession planning, which is one of the primary functions of a board. Nobody thought to groom even an emergency successor for a CEO with more than 60 direct reports and no president or chief operating officer beneath him, and a system of doing business through handshake deals to centralize all power in himself. And when the crisis came, the board’s first instinct was to vouch for him before its own investigation had begun before finally surrendering to the facts of the situation.
Similarly, it was plainly inadequate and insufficient for Nvidia’s board, to simply defer to the professional judgment of their advisors, such as their external accountant PwC or their external law firm Cooley. PwC signed off on financial reports with what many experts consider to be questionable accounting, obfuscating the geographic origins of Nvidia’s revenues and the end users of sensitive Nvidia chips in ways that could possibly open the board up to legal claims of failure of oversight if not even perhaps fraud in the eyes of many experts.
Even today, there still seems to be a latent expectation amongst some in tech circles, including apparently from Huang himself, that a weak Nvidia board will just collapse and bring Huang back in deference to the founder if enough pressure is brought to bear. Huang himself put out an unusual statement this morning, saying that “If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO”, even though Huang is now the target of no less than 4 ongoing government investigations across congressional oversight hearings and federal probes as well as multiple shareholder lawsuits. Certainly, Marshal Philippe Petain of Vichy France; the deposed Emperor Napoleon, Enron’s disgraced Jeff Skilling, Worldcom’s deposed Bernie Ebbers, and other deposed monarchs dragged out kicking and screaming would have likewise enthusiastically volunteered to return after being thrown out by their constituents. Bewilderingly, even some proxy ratings firms like Egan Jones seem to be lending implicit support to the ‘Bring Jensen Back’ movement, declaring that “It is not entirely clear the board acted in the best interest of shareholders” by removing Huang, looking past the evasion of sanctions, national security, and accounting integrity concerns to say the board should not have removed Jensen on the basis of shareholder returns from prior, possibly misleading financial reports.
Lessons for boards moving forward – especially in working with a strong founder
What Nvidia’s board woes highlight is that contrary to the outdated ideas of some – for example, how Travis Laster would define ‘good faith’ as checking hot tempers at the door – the measure of genuine board diligence is the willingness to foster board cultures of open dissent and genuine partnership with CEOs over blind deference.
The latter challenge is especially salient when it comes to working with larger-than-life founders. The right answer is not for boards to go to war with these creative geniuses by trying to suppress them, or to lose trust by trying to fire them over foot faults. Rather, it is up to the board to develop governance practices that show they know how to partner with the genius they are privileged to have among them.
The boards of Microsoft, Amazon, Dell, Starbucks and, ultimately, Apple figured out how to work with Bill Gates, Jeff Bezos, Michael Dell, Howard Schultz and Steve Jobs, forging alignment between the needs of their founder visionary CEOs and the shareholders. These boards helped their founders appreciate and address their human frailties, whether with non-accusatory accountability (as with Gates through Microsoft anti-trust litigation) or during health crises (Jobs’ battle with cancer and need for credible succession plan).
It is not easy to regulate genius and creative founders do not resemble rational people as they challenge the daunting odds of success that most people wisely avoid. They are also often obsessed and single minded about their visions – which seem to many to be delusions seeking a mirage of unattainable entrepreneurial success. Yet their innovations can disrupt and transform society on the few occasions where they succeed.
Sigmund Freud reminded us that society is changed by its discontents. Painter Robert Motherwell confessed that as an artist there is never satisfaction, because creative joy is mixed with the anguish of what could be. But at the same time, boards must recognize that they are not dealing with an artist painting in a private studio on their own canvas, but rather an artist who is using other people’s resources, painting on a public canvas, and so must listen to other voices.
Only by forging genuine partnerships with their CEO built on trust and mutual candor, can boards strike the right balance between blind deference and overbearing suffocation, particularly when it comes to working with bigger-than-life visionary founders. The lesson for boards, and what both sides of the critics’ spectrum misses, is that the solution is not related to the structure of the board; rather, it is the professional courage to do their jobs fully, and nobody can legislate performance through checklists or rulebooks.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
This story was originally featured on Fortune.com


