The Three Checks Failed at Once. That Was Not an Accident.

All three were loosened by the same idea, in the same era… and now they hold each other open.

I treat fiduciary duty as a real obligation. Not a clause. Not a liability to be managed down. An obligation… the kind you carry whether or not anyone is watching, whether or not a court would ever make you. Leave the institution stronger than you found it. That is the entire job.

So I want to be precise about what has gone wrong, because precision is the only honest form of anger.

There was a time, within living memory, when that obligation had teeth. After the Savings & Loan crisis, thousands of bankers failed, hundreds of institutions collapsed, and more than a thousand people were convicted of crimes connected to the wreckage. The lesson was not subtle. If you ran an institution into the ground through recklessness or self-dealing, the law could find you personally.

Stewardship was not a sentiment. It was enforced.

Four decades later, that enforcement has gone quiet — and not because we lost the statutes. The duties of care and loyalty are still owed. The words are still in the law. What collapsed was something larger and more deliberate than any single doctrine.

Three checks, one job

Powerful corporate decision-makers are supposed to face discipline from three directions.

• The market disciplines them through competition: do badly, and a rival takes your customers.
• The state disciplines them through regulation and enforcement: cross the line, and an agency with funding and will comes after you.
• The courts discipline them through fiduciary liability: betray the institution, and you can be held personally answerable.

A serious failure of stewardship used to run a gauntlet. To get away with it, you had to escape all three. That redundancy was the point. No single check had to be perfect because the other two stood behind it.

Over the last forty years, all three were loosened. Not by accident, and not independently. They were loosened by the same intellectual project, beginning in the late 1970s, and the result is that none of them now stands behind the others.

The courts: the threshold moved

Start with the one I know in my bones.

Fiduciary duty did not disappear. The threshold of proof rose until much of what feels like betrayal sits safely on the legal side of the line.

Shareholder primacy supplied the philosophy — Milton Friedman’s clean formulation that the business of business is profit¹, which quietly recast the duty from stewardship of an institution into maximization of returns within legal bounds. Same words, different center of gravity. The business judgment rule supplied the shield: courts defer to decisions made in good faith on reasonable information, which is sound, risk requires protection, but it means most bad decisions are not breaches. Losing billions is rarely enough. You generally have to prove self-dealing, fraud, gross negligence, or a conflict.

Then compensation rewired the incentives. Tie a CEO’s wealth to the share price and you have pointed the paycheck at the stock while the duty still points at the corporation. And scale diffused the blame: in a modern enterprise, the destructive decision rarely has a single author. Everyone signed off. No one decided. Diffusion is not a defense in theory. In practice it makes accountability nearly impossible to pin to a name.

The court — the third check — went soft.

The market: competition was redefined away

Now the first check, and here is where the same project shows its hand.

In 1978, Robert Bork’s The Antitrust Paradox² gave courts and enforcers a new lodestar: the “consumer welfare standard.” Monopoly was no longer a structural danger to be prevented; it was acceptable so long as consumer prices did not visibly rise. It was the same Chicago School instinct that produced shareholder primacy, trust the market, distrust the state, applied to competition policy.

The effect compounded over decades. Mergers that once would have been blocked went through. Industry after industry consolidated into a handful of dominant players. And as competition thinned, the market’s ability to discipline a badly run firm thinned with it. You cannot lose your customers to a rival who no longer exists.

The market — the first check — went soft.

The state: the referee was bought and starved

The second check required a government willing and able to enforce. So the same project worked on the government.

Money flowed into politics on an arc that ran from Buckley v. Valeo³ in 1976 through Citizens United in 2010⁴ and McCutcheon in 2014⁵, each decision widening the channel. Lobbying spending climbed for decades. And concentrated industries, the product of the first failure, had exactly the resources to fund it.

The result was capture in two forms. Rules got written to favor incumbents. And the enforcement apparatus that would otherwise raise the cost of breaching duty — the agencies, the antitrust divisions, the examiners — was underfunded, outgunned, and discouraged from acting.

After Enron and WorldCom we got Sarbanes-Oxley⁶; after the financial crisis, Dodd-Frank⁷. Real reforms. But the pattern held: reform follows scandal, attention drifts, enforcement softens, and accountability never returns to its old weight.

The state — the second check — went soft.

The flywheel

This is what makes it more than three sad coincidences.

Concentration produces fewer, larger firms with more money to spend on political influence. That influence captures and starves the enforcement that would discipline them. Weak enforcement plus a high liability threshold means executives face minimal consequence from any direction. Minimal consequence frees them to optimize for the share price — including through the consolidation that produces still more concentration.

Each check that fails makes the others easier to disarm.

It is a flywheel, and its output is an accountability vacuum: a class of decision-makers who answer to no rival, no regulator, and no court. Not because they broke the law. Because the law’s threshold, the market’s competition, and the state’s will were all moved to where they now sit.

What the vacuum looks like with names on it

The abstraction is easy to wave away. The cases are not.

When the financial crisis was finally settled, JPMorgan paid $13 billion⁸… the largest settlement with a single entity in American history to that point. Roughly $7 billion of it was tax-deductible. The shareholders wrote the check, not the executives. No senior banker connected to the crisis was criminally prosecuted. And the year after the settlement, Jamie Dimon’s pay rose 74%⁹. Read that sequence again. This is not the story of a man who broke the law and escaped it. It is the story of a system in which the largest penalty in its history landed on no individual at all and the person at the top was paid more for the year it was levied.

Or take a smaller, cleaner case. Kevin O’Leary was paid roughly $15 million to be the public face of FTX¹⁰, and he pushed it hard to retail investors, even boasting that the exchange had finally solved his crypto compliance problems, months before its own incoming CEO would describe its controls as a complete failure of corporate controls. FTX collapsed. Its founder was convicted of fraud and sentenced to twenty-five years.

Ordinary investors were wiped out. O’Leary faced a class action, and in 2025 a federal judge dismissed most of the claims against him and the other celebrity promoters, finding the plaintiffs could not show the promoters knew of the fraud. He kept his platform and his standing.

Notice what neither story requires. Neither man was convicted of anything. Neither, in the end, was found liable.

That is not the system failing to catch wrongdoers. That is the system operating exactly as it has been rebuilt to operate, producing no individual accountability while every part of it works as designed.

The vacuum is not an absence of rules. It is the result of them.

The honest objection

I owe the other side its strongest case, because anger without rigor is just noise.

The defenders are not wrong about everything. Capital discipline is real… the optimization era did punish empire-building and complacent management. Concentration is not always sinister; some of it is genuine scale economies and network effects, not merely lax enforcement. Consumer prices in many sectors did fall. And decades of equity returns funded the pensions and index funds of ordinary people. A vague duty “to the institution” can itself become a license for managers to entrench and answer to no one.

Grant all of it. The argument still holds, because it is not a claim that one thing mechanically caused another. It is a claim about a common root and mutual reinforcement: the same idea loosened all three checks, and the loosened checks now prop each other open.

You can believe the era produced real efficiency and still see clearly that it produced a real accountability vacuum. Those are separate ledgers. The gains do not erase the cost, and the cost is the one nobody priced.

The duty didn’t fail. Its meaning did.

There is a temptation to call all of this a failure of fiduciary duty. It is more precise, and more damning, to say the reverse.

The duty did not fail. Two definitions of it pulled apart, and the world kept the wrong one.

• The older definition is stewardship: leave the institution stronger than you found it, act with integrity, do not say one thing in public while your balance sheet does the opposite.
• The newer definition is shareholder primacy: maximize the owners’ returns within legal bounds, and owe nothing past that line.

Hold the two side by side and the cases stop being puzzles.

A man who disparages an asset in public while his institution quietly builds the business that profits from it either way is, under the first definition, a disgrace. Under the second, he is doing the job well. A board can preside over the destruction of enormous value and breach nothing. An executive can optimize the share price at the institution’s long-term expense and stay comfortably inside the rules.

None of it is a betrayal of the duty as the duty is now understood. It is the fulfillment of it.

That is the part that should keep you up at night. We did not watch fiduciary duty get broken. We watched its meaning get hollowed out until breaking it became unnecessary. The word survived. The obligation inside it was quietly swapped for a smaller one.

What the duty asks now

So here is where the precision turns into resolve.

The older meaning did not die. The people who still carry it simply got rarer, and that is a different problem, with a different answer. You do not restore a standard by mourning it. You restore it by embodying it until it is normal again.

Start with the failure mode that made the hollowing possible. The opposite of stewardship is not theft; it is diffusion, everyone signed off, no one decided. And diffusion can only be beaten the way it was built: structurally. You refuse to let a consequential decision exist without an owner. You re-attach the name. You make authorship a design constraint, so that every call of consequence traces to a person who answers for it, not because the law demands it, but because the architecture does.

This is no longer a question for the next decade. When the decision is made by an autonomous agent, the name is harder to find than it has ever been, and the same systems can manufacture the appearance of diligence faster than anyone can read it: the reports, the reviews, the sign-offs, generated on demand, with no one standing behind a word of them.

A compliance record that looks accountable and answers to nobody is the hollowing-out finished.

AI did not start this drift. Pointed carelessly, it completes it, diffusion at machine speed, stewardship reduced to a convincing artifact. The institutions worth building are the ones that insist on a human owner exactly where it would be easiest to let the machine absorb the blame.

That is what survives the audit the S&L era once ran… responsibility with a name on it, incentives pointed at the enterprise and not the print, stewardship made structural rather than sworn. It means treating the duty as binding precisely because the law no longer reliably binds it for you. The standard does not live in the statute anymore. It lives in the people who refuse to let it lapse, and in the systems they refuse to let decide without a name attached.

Leave the institution stronger than you found it.

It was never the court’s job to make us mean that. It was ours.