FICTION · An Unfair Advantage Intervention
Fiction. Real history. Sourced below.
Venture capital abandoned Doriot’s permanent-capital vehicle for the ten-year partnership because the partnership could pay carried interest. It is now rediscovering permanent capital, because the ten-year clock forces firms to sell their winners. The structure Doriot got wrong is being rebuilt, and the reason it failed him has not obviously been solved.
This is a compensation structure review, which is my discipline rather than investing. I am Samantha Reynolds and I built Cartwell’s incentive system, so I read this case the way I read any org design: not as a story about a brilliant investor, which it also is, but as a diagnostic on a firm that had three of the four things it needed and lost everything to the fourth.
Full fiction disclosure and sourcing note
Samantha Reynolds is a fictional character from The Unfair Advantage by Todd Hagopian, analyzing the documented history of American Research and Development Corporation from the present day. She has never met, advised, or corresponded with Georges Doriot, who died in 1987, or anyone else named here, and no dialogue, thought, or motive is invented for any real person. Quoted words come from an internal memo reproduced in academic research. Facts are drawn from the public record and accurate as of 2026; the analysis is invented. One figure is contested: ARD’s lifetime return is reported as 14.7 percent annualised with roughly half attributable to a single investment, and elsewhere as 15.8 percent with that investment accounting for around 80 percent of gains.
- One. Two Inventions, and Only One Survived
- Two. What the Power Law Actually Looked Like
- Three. Three of Four Components
- Four. Doriot Wrote the Diagnosis Himself
- Five. The Clock the Partnership Introduced
- The Contrarian Reappraisal
- About the Stagnation Assassin
One. Two Inventions, and Only One Survived
American Research and Development Corporation was founded in Boston in June 1946 by Georges Doriot, MIT president Karl Compton and Boston Federal Reserve president Ralph Flanders, raising about $3.5 million from insurers, investment trusts and university endowments. It is generally credited as the first institutional venture capital firm, and by 1961 it was the first listed on the New York Stock Exchange.
Doriot produced two things and they are usually discussed as one. The first was an investment thesis: back exceptional founders rather than ideas, expect most positions to fail, allow a rare outlier to carry the whole portfolio, hold patiently, and contribute board-level help rather than only capital. Every element of that is standard practice today.
The second was an organisational structure: a publicly traded closed-end investment company, paying salaries, holding assets indefinitely because it had no fund life to run out. The industry adopted the first invention completely and discarded the second. That gap is the case, and it is worth separating carefully, because “modern venture capital descends from ARD” is accurate about the thesis and precisely backwards about the org chart.
Two. What the Power Law Actually Looked Like
In 1957 ARD put $70,000 of equity into Digital Equipment Corporation for roughly seventy percent of the company. That figure is the one everyone quotes and it understates the commitment: there was also an initial loan of about $30,000 and subsequent lending that reportedly totalled around $2 million. By 1971 the stake was worth about $355 million.
Now the number that should reframe how anyone thinks about a venture portfolio. Across its twenty-five year life ARD returned 14.7 percent annualised, and 7.4 percent without DEC, which trailed the Dow over the same period. Some accounts put the figures higher and the concentration heavier still.
Read plainly, the firm that invented venture capital, run by the man who wrote the thesis, was a below-market investor except for one position. That is not an indictment. It is the clearest documented illustration of the distribution the whole asset class runs on, and it means the operating question for anyone allocating under deep uncertainty is not how to assemble good outcomes but how to ensure you are still holding the exceptional one when it appears. Doriot’s patience was the mechanism: he held DEC for fourteen years and turned down acquisition interest from larger companies along the way.
Three. Three of Four Components
Revenue Responsibility Engineering asks whether the people closest to value creation can see it, own it, benefit from it, and have done so long enough that it is simply how the organisation works. Visibility, accountability, incentive alignment, cultural embedding. Run the four on ARD.
Visibility was excellent, arguably better than at any private firm since. ARD was publicly traded, so its net asset value was reported and every professional could watch the consequences of their work move a published number.
Accountability was clear. Individual officers sourced deals, sat on boards, and were identifiable with specific companies. Nobody was hiding inside a committee.
Cultural embedding was formidable. Doriot taught at Harvard for decades, and the firm’s method, its patience and its founder-first philosophy were institutional habits rather than policies.
Incentive alignment was absent. As a public investment company, ARD could not offer its staff carried interest or meaningful equity in the ways SEC rules permitted private partnerships, and Doriot was not inclined to work around it, holding that his people would lose their drive if they became wealthy quickly. So the professionals who found and built the companies took salaries while the gains accrued to shareholders and to the few who happened to hold stock personally.
Three of four, executed at a very high standard, with a hole where the fourth belongs. In my experience that configuration is more dangerous than a weak system across the board, because everything visible looks healthy right up to the moment people leave.
Four. Doriot Wrote the Diagnosis Himself
This is the part I find hardest to read, professionally. In an internal memo in 1965, Doriot recorded that ARD’s officers did not feel a heavy responsibility for the inflow of high-quality projects, and that the firm was losing competitive ground. Academic analysis of the firm links that decline in deal discovery to the inability to compensate professionals.
So the diagnosis existed, in writing, from the founder, naming an ownership deficit. The compensation structure did not change.
The same year, Doriot’s senior colleague Bill Elfers left to co-found Greylock, on the limited partnership model that could pay a share of gains to the people generating them. Others followed and founded their own firms. Kleiner Perkins and Sequoia were both established in 1972, both as limited partnerships. ARD, unable to match on compensation and with no succession arrangement in place, was merged into the conglomerate Textron that same year and ceased operating by 1976.
The thesis, meanwhile, propagated through exactly the people who left. Doriot’s fear was that wealth would remove his colleagues’ motivation. What removed them was the absence of it, and in leaving they distributed his method across an industry far more effectively than ARD’s own returns ever could have.
Five. The Clock the Partnership Introduced
Here is why this case is live rather than historical. The limited partnership solved Doriot’s problem and introduced a different one. Carried interest aligned the professionals, and a defined fund life of roughly eight to ten years introduced a clock. A clock means that when a portfolio company goes public, the shares get distributed and the position ends, whether or not the compounding is finished.
Doriot had no clock. That is precisely why he could hold DEC for fourteen years while the value went from a rounding error to $355 million.
The industry has noticed. In 2021 Sequoia restructured around an open-ended vehicle with no fixed term, able to hold public positions indefinitely and feed closed sub-funds by stage, and by a filing in early 2025 that fund had reached $19.6 billion. More broadly, while most private funds still use the traditional closed-end form, perpetual capital structures have been growing across private markets, with venture a small share of evergreen assets and widely expected to expand.
Which sets up the open question, and it is an incentive question rather than an investing one. Carried interest is calculated against realisations. A vehicle designed never to realise has to pay its people against marks instead, on positions the firm itself values, over horizons with no natural end. That is solvable, and several firms are attempting it. It is also the same structural tension that emptied ARD, arriving from a different direction, and I would want to see it working across a full cycle before I called it solved.
The Contrarian Reappraisal
I have argued that incentive design killed this firm, and there is a serious objection that I want to put at full strength.
The objection is that I am dressing up a portfolio outcome as an org-design lesson. ARD took a seventy percent stake in DEC and lent it two million dollars, which is not venture investing as we now understand it; that is a holding company position. Once DEC dominated the balance sheet, ARD was effectively a single-asset vehicle whose remaining activity barely mattered to its returns. On this reading, no compensation scheme would have changed the arithmetic, the firm was over by 1966 in everything but name, and the talent departures were a symptom of a company that had become a DEC proxy rather than a cause of anything. The 7.4 percent figure supports that: the non-DEC portfolio was mediocre while the incentives were still intact.
I find that about half right and the half I accept matters. Concentration did make ARD strange, and I have probably underweighted how much a single dominant position distorts everything downstream, including morale.
What the objection cannot explain is the direction of travel. If ARD’s people were leaving because the firm had become a DEC proxy, they would have gone to established institutions. They went and built limited partnerships, and the specific thing those partnerships changed was compensation. Elfers is documented as leaving to correct the structural inequities. That is not what a symptom looks like. It is what a diagnosis looks like when the person making it has read the same memo the founder wrote and drawn the conclusion he would not.
About the Stagnation Assassin
Todd Hagopian once approved a strategy he believed in and watched two of the three people needed to execute it leave within a year, for the ordinary reason that the plan’s success would not have changed their circumstances. He is VP of Product Strategy and Marketing at JBT Marel, writes as The Stagnation Assassin on incentive design, capital structure, and the organisational arrangements that decide which strategies a company can actually run, and is the author of the Stagnation Assassin series.
Ask who becomes wealthy if the plan works, and whether it is the people you need to keep. If the answer is no, the plan is already leaving the building, however good it is.
Todd speaks to executive teams and conferences on incentive design, ownership structures, and why correct strategies still lose their people. Book a keynote.