The Need for a Distributed Allocator Layer

The Law

The power law tells us that a small number of companies produce nearly all venture returns. The empirical record confirms this. Roughly 65% of financings return less than the capital invested. Around 6% of invested dollars produce 60% of returns.

We accept the shape. Value creation in technology is heavy-tailed, and no amount of training converts that distribution into a normal one.

What the record does not confirm is the conclusion the industry drew from it. The power law describes outcomes among companies that were funded. It says nothing about the companies that could have been. The difference between those populations is determined entirely by the allocator layer.

Research on venture performance persistence found that firms that win early do not necessarily get better at picking. They get better access. Reputation compounds. Judgment does not.

The optimal unit of early-stage allocation is a trained individual with deep domain knowledge and direct proximity to the companies they fund. It is not a firm.

An individual, however, has no reliable deal flow, diligence infrastructure, legal apparatus, or realistic path to a portfolio wide enough for the math to work. They see a handful of companies, chosen by who happens to know them. They have too few repetitions and no feedback loop.

The result is that the people best positioned to allocate early-stage capital are the least active at it. An exited operator who spent eleven years in logistics understands logistics companies better than any generalist partner. They may write two checks in their lifetime, both outside their domain.

A new production economy is arriving. AI is reducing the capital, headcount, and time required to turn expertise into a company. More founders—including operators, researchers, and solopreneurs—can now build consequential businesses with smaller teams and less initial capital.

The population of possible companies is expanding far faster than the population of people equipped to evaluate them. Creation is becoming abundant. Allocation remains scarce.

One side of the market scaled. The other did not. A widening base of founders feeding a fixed and narrowing funnel has two possible outcomes: the existing layer funds only the most legible fraction, or the allocator layer expands.

The Collective

A distributed allocator layer resolves the problem by decoupling two scaling challenges that the venture firm bundles together and individuals cannot solve alone.

Judgment scales through independence. Access and infrastructure scale through pooling.

When members allocate across different domains, sectors, and geographies, the collective produces breadth without asking each person to abandon their domain. One allocator holding three companies in their own field is making a concentrated bet. Four hundred allocators each evaluating and backing three companies in their own fields create a diversified intelligence network: twelve hundred domain-informed positions from which properly constructed portfolios can be assembled.

A collective also solves the hard problems every individual currently rebuilds alone: sourcing, structured diligence, legal formation, portfolio construction, follow-on discipline, and shared learning.

A syndicate pools capital behind one person’s judgment. A collective pools judgment and distributes capital behind many. Here, members are the allocators. The infrastructure exists to make their own judgment executable.

The Prudent Unlock

In 1979, the Department of Labor clarified ERISA’s prudent-man standard. A risky investment was not automatically imprudent in isolation; its role had to be considered within the portfolio, alongside diversification, liquidity, risk, and expected return.

That shift did not weaken fiduciary duty. It changed the frame through which prudence was judged — and helped pension capital begin flowing into venture funds in the decade that followed.

The prudent unlock of 1979 expanded the supply of venture capital. The next prudent unlock must expand the supply of people capable of allocating it.

Today, the door to private markets is opening again. But access without competence is not democratization. Equity crowdfunding, blank-check vehicles, token offerings, and passive syndicates repeatedly lowered the barrier to entry while leaving the barrier to judgment untouched.

The School

The School for the Advancement of Abundance trains capital allocators. The program is an apprenticeship rather than a course because the returns evidence is about behavior, not information.

Students conduct real diligence on real companies, against an hours floor, with their work reviewed. Deals are routed to domains where students have actually operated. Investment committees happen live and in public. Capital is deployed into companies students bring. Portfolios are built toward the range the data supports rather than the range an individual can reach unaided.

The prototype has been run. An earlier version of the curriculum taught roughly 1,300 MBAs, put 360 companies through real diligence, and those companies subsequently raised more than $300 million.

Do the work. Stay in your domain. Show up. Hold enough of them.

Our immediate goal is simple: the same dollar allocated by a graduate reaches a better company than it otherwise would have, and reaches it faster.

The Million

Every expansion of access leaves the same question behind: who does this actually serve?

Our answer is one million trained allocators, and we intend to be measured against it rather than praised for it.

We will publish what happens. No quasi-experimental evidence yet shows that investor training improves investor returns. The correlation is documented; the causation is not. We will track trained cohorts against comparable untrained allocators and publish the results whether or not they favor us.

The training will not be gated by the wealth threshold it is designed to make obsolete. Deploying capital into private companies may require accreditation. Learning to allocate it does not.

The next decade will produce more companies than any decade in history. Most will die without meeting a person qualified to evaluate them.

That is not a law of nature. It is a staffing problem, and staffing problems are solvable.