I care a lot about how fast humanity progresses: how quickly we get cures, abundant energy, cheap access to space, and tools that make the next round of progress easier. When I ask where that pace is decided, I keep arriving at the same place. A small number of organizations hold a large share of the world’s specialized talent and capital, and I keep noticing what they are pointed at.
They are pointed at the price of a residual claim. Shareholders receive what is left after everyone else is paid, and because those claims trade freely, their price is reset continuously by people who may not hold them for long. Economists have spent decades on the consequences. Jeremy Stein’s work on managerial myopia showed that even when markets are rational, managers who answer to the current price will trade away long-run value to protect it. Others have traced the same pressure through investment and R&D. The evidence is contested: Steven Kaplan and others argue the popular story of corporate short-termism is overstated, while other work finds that public firms respond less to investment opportunities than comparable private ones. I’m not claiming to have discovered short-termism, and I don’t think the debate is closed. This essay stands on that literature.
What I want to add is a question about scale. Most of that work asks whether firms invest efficiently given their objective. I want to ask what the objective itself costs. What do the best-resourced organizations in history leave undone, or do years late, because the thing they are rewarded for is the price of a residual claim? I’ll argue that this opportunity cost is large and underestimated, largely because it is invisible. Nobody sees the program that was never staffed at full intensity, or the capability that arrived a decade late.
The essay has three parts. First, how the objective inside these firms discounts work with distant or diffuse payoffs, and what that costs. Second, why add-on fixes like ESG and mission metrics can’t change it. Third, what a different primary institution might look like.
The design behind public ownership is elegant. Because equity holders are paid last, their wealth depends on the firm’s ability to keep producing surplus, which should give them every reason to protect and expand its long-term productive capacity. Boards answer to shareholders and employees are paid in equity for exactly this reason: the residual claim is meant to keep everyone pointed at durable health. The question is how reliably it does so once the claim can be sold.
The people inside a large firm can want almost anything: to cure a disease, to build something durable, to keep the lights on safely. But the terms on which they work are set by a board that answers to shareholders, and pay, promotion, and capital budgets follow from that. Whatever the stated mission, the objective that is actually rewarded is the one that shows up in the price. Nobody has to be cynical for this to matter.
It might seem that the price only binds when a firm raises money, and these firms mostly fund themselves. It reaches managers anyway. Their compensation is tied to it, activists and acquirers appear when it lags, and a board’s patience tracks it. Behind the price are holders who can sell at any time, so it reflects what people will pay for outcomes that arrive within the horizon the current owners expect to stay. That horizon is the window the firm is effectively managed to.
Part of this is an old problem. Since Arrow and Nelson, economists have known that profit-seeking firms underinvest in work whose benefits are diffuse and hard to capture, and that would be true of a private firm with a patient owner. What the liquid residual claim adds is a shorter and less forgiving window. A patient private owner can choose to wait. A public firm is managed to the horizon of its marginal holder, and its pay and governance are tied to a price that updates daily. Liquidity amplifies the older problem, and I think it amplifies it substantially, though I can’t say by exactly how much.
Where a firm’s long-run interests and that window point the same way, nothing is lost, and plenty of long-term investment happens for exactly that reason. The cost appears where they diverge: spending whose payoff comes after today’s owners have gone, or work whose value is enormous but spread across many people and hard for any one firm to capture.
The pattern is general. Take space. Cheap access to orbit is infrastructure whose benefits spread to everyone who later builds on it, and reaching it took years of heavy spending with no revenue. The step change, reusable rockets, came from a founder-controlled private company that nearly failed more than once, not from the public aerospace primes, whose contracts and margins reward predictability. Or take neurotechnology. Interfaces that restore movement or speech, or treat disorders we barely understand, have decades-long paybacks, an approval path not designed for them, and benefits that are large and hard to capture. Again the leading efforts are mostly private and founder-funded, not inside the largest firms with the capital to scale them. These are the cases where someone loosened the window. The question is how much else never gets that protection.
Aging is the case I care about most. Roughly two-thirds of the people who die each day, on the order of a hundred thousand, die of causes tied to age. A therapy that slowed the underlying biology, even partially, would be among the largest gains in human welfare available. Large pharmaceutical and technology firms do invest in aging-adjacent work, and I’m not claiming otherwise. My claim concerns the margin: how much more would happen if the objective did not discount distant, diffuse payoffs. Universities, public funders, and startups generate many of the frontier ideas, but turning an idea into a therapy for millions takes the assets that sit inside the largest institutions: capital, compute, clinical trial machinery, manufacturing, and regulatory experience. An intervention that delays many diseases at once is hard to protect with a patent, pays back over decades, and fits no clear regulatory category, since aging itself is not an approvable indication. Each of those costs lands inside the window while the benefit lands outside it. So the work proceeds largely as a by-product of disease-by-disease programs, where the payoff is near and capturable, rather than as a goal in its own right.
There is some evidence of what happens when the window is loosened. Alphabet’s founders used their control of the company to start Calico, a company devoted to aging biology, which a conventionally governed firm might have found hard to justify. Meta’s founder-controlled structure has let it fund years of heavy spending on VR and computing infrastructure, although many people think the VR bet destroyed value, which is a reminder that insulation cuts both ways: it can protect a visionary or an entrenched mistake. Foundation-owned firms such as Novo Nordisk and Bosch show similar horizons outside the stock market, though they are profitable enough that their results are suggestive, not conclusive. In every case the ordinary market stays open next door, and the exceptions depend on one founder’s willingness to look past the window. That doesn’t scale, and nothing outside the firm tells us whether the long bet is working.
That is why I think the opportunity cost is large and underestimated. I can’t prove the counterfactual, and it is possible that the binding constraint on aging, or on any of these fields, is science and not capital. If it turned out that substantially more effort from these organizations would have changed little, I would have to revise this argument a great deal. But the question is rarely asked, because the loss is a capability that arrives later than it could have, produced by capacity that was never pointed at it with full force. No one records that, and it never appears on anyone’s books. Consider a rough calculation. Around 36 million people a year die of causes tied to age. If a therapy averted even three percent of those deaths, each year of delay would cost on the order of a million lives.
The mechanisms that police managers all do their job well, and that is the difficulty. Equity compensation, board oversight, and activist investors exist to carry the price of the residual claim into the decisions of the people running the firm. Sharpening them makes the transmission faster and tighter. It does nothing to lengthen the window the price reflects, because the price is the thing being transmitted.
The more telling case is what happens when firms try to add other goals on top. ESG targets, stakeholder commitments, and internal mission metrics are often offered as proof that a company can be steered toward broader aims without disturbing the primacy of the share price. The record is not encouraging. ESG-labelled funds have tended to hold much the same stocks as ordinary ones, and when margins tightened and the political mood turned in 2021 and 2022, many firms quietly softened or dropped targets they had announced a few years earlier. None of this required bad faith. A manager facing a hard, tradable price on one side and a softer target on the other will, reasonably, resolve the conflict in favor of the measure that determines their own wealth and the firm’s access to capital. The softer goal then gets met through redefinition, selective reporting, or programs too small to displace the real allocation of people and money.
There is a general principle underneath this. A measure steers an organization at scale only if capital can enter and leave on it and managerial wealth depends on it. The share price has both properties. Mission metrics have neither. They can restrain the worst behavior at the margin, and some bite harder when tied to regulation, litigation, or large pay components. But as long as they remain weaker than the price, they cannot make full-intensity pursuit of long-horizon work the sensible default inside a firm whose shares trade freely. The opportunity cost survives every overlay, because the overlay never touches what is rewarded.
That points toward the answer. The problem is not that other goals lack advocates. It is that they lack a measure as hard, as external, and as widely trusted as the price. Until they have one, we will keep adding constraints on top of the measure that already dominates, when the thing missing is a second one.
The share price steers firms for a reason that has little to do with markets being wise. It is hard to fake. Exchanges, auditors, disclosure rules, and the threat of prosecution mean outsiders can watch it continuously, and nobody has to take a firm’s word for anything. That is why pay, board evaluation, and access to capital can all be tied to it. Other objectives have no equivalent, and the reason is not that they matter less. Nobody has built the institutions that would make them equally hard to fake.
I want to describe what building them would look like. This is the good version of the idea, not the version that could pass a legislature next year. The point is to see what we would want if feasibility were set aside, and then to be honest about where it is hard.
It also has neighbors, and I should name them. Prizes and advance market commitments, which Michael Kremer and others have championed, pay for results instead of effort. Impact certificates and retroactive funding reward outcomes after the fact. Robin Hanson’s decision markets use trading to surface information about what will work. Social impact bonds tried to tie capital to measured outcomes and have mostly underperformed, often because the measurement was weak or too easy to dispute. What I am proposing is a variation on this family: continuous independent certification, a liquid secondary market in claims on certified results, and compensation inside organizations that vests against the same certifications. Whether that avoids the failures of its relatives is the thing to test.
Take aging. Suppose an independent certifying body, with disclosure and audit powers modeled on securities regulation, publishes standards for the field: candidate biomarkers of biological age, replicated lifespan and healthspan results in aged mammals, and eventually human trial endpoints. A research organization, whether a division of a large firm or a standalone lab, commits to a program and discloses its methods, data, and the resources it actually deploys. The certifier verifies both continuously. What gets certified is verified inputs and replicated results, not a vague notion of progress. The team’s compensation vests against those certifications, not against a parent company’s share price. The money comes from an outcome fund that pays out as milestones are certified, and investors hold tradable claims on those payouts. When a result replicates, the claims rise. When one fails to, they fall. That price is the second scoreboard.
The same machinery applies well beyond aging. Launch cost per kilogram to orbit, or replicated clinical results for a neural interface, can be certified in the same way: published standards, disclosed data, independent audit, and claims that rise and fall with verified results.
This is how the proposal reaches the organizations I have been describing. It creates a second path: a large firm can point a division at certified outcomes and be paid for them, with pay inside that division tied to the same certifications. It does not require the parent’s shareholders to change what they want. It gives the firm a second thing to be rewarded for, which is exactly what the ESG overlays could not do.
The obvious objection is that certifying contested, long-dated science is much harder than recording a stock trade, and that any measure will be gamed. Four safeguards address this. Certifiers compete, and each stakes its reputation and carries personal liability for what it signs. Paid adversarial teams are tasked with breaking claims, so replication is an income source and not an act of charity. Settlement is delayed, so a claim pays out in full only after a later and harder check. And the standards rest on things that can be checked, such as results that others reproduce and resources that were actually spent. None of this eliminates gaming. It makes gaming costly, which is all the stock market’s own safeguards do.
The capital comes from the people who already want these outcomes and currently have no liquid way to fund them: pension funds and sovereign funds with decades-long liabilities, philanthropists, and governments that would rather buy results than fund effort. Everyone else can stay in ordinary equities. The new market does not need to win everyone, only enough capital to staff the work at the intensity it can absorb.
Liquidity needs care here. The tempting fix for short horizons is to make capital stickier, but illiquid structures buy horizon at the cost of discipline and honest pricing, as the recent troubles of private credit and non-traded funds show. The aim is different: a claim that stays liquid but is tied to a long-horizon measure, so that selling no longer means leaving the objective. That carries its own risk. A liquid market in certified milestones could develop short-termism of its own, with traders speculating on the next announcement. The delayed settlement helps, but I wouldn’t claim it solves the problem.
Whether this can be built is an open question. Certifiers would have to stay independent of the people they certify, and enough capital would have to leave ordinary equities for the new market to staff real work. If certification proves as easy to game as a mission statement, or the money never comes, the case for a second measure weakens a great deal. I would want to know that.
I began with a question about what sets the pace of human progress. My tentative answer is that part of it is set by what the largest concentrations of talent and capital are rewarded for, and that we rarely examine this because we treat the objective as given. It is a human institution, and institutions can be redesigned.
Aging is why I care. Around a hundred thousand people die each day of causes tied to age. No one will ever be able to point to those lost because a program was staffed below what it could have been, or because a cure arrived a few years late. That absence is the cost, and it is the reason the cost is so easy to leave unexamined.