The CAPM Paper Nobody Published: Jack Treynor's Lost Manuscript

This is the strangest entry in the asset pricing canon: a paper that shaped the field without ever appearing in a journal.

Jack Treynor was not an academic. He was a consultant at Arthur D. Little who had studied mathematics and become obsessed, on a vacation, with the question of how risk should affect the value of a company. Working largely alone, he produced a manuscript in the early 1960s deriving what we now call the Capital Asset Pricing Model. He circulated it. Franco Modigliani read it and encouraged him. It went around the finance community as a typescript.

And then he never published it. It finally appeared in print in 1999, in a collection edited by Robert Korajczyk, more than three decades after Sharpe, Lintner and Mossin had published their own versions and long after the model had been named for other people. Sharpe won the Nobel. Treynor did not.

The problem: what is a risky company actually worth?

Treynor came at the question from the direction of corporate valuation, not portfolio management. A company generates uncertain future cash flows. Everyone agrees you should discount them. But at what rate?

The received practice was to pick a discount rate that "felt right" given how risky the business seemed. Treynor found this intolerable, because it made valuation a matter of taste rather than a matter of logic, and because he suspected the intuition behind it was flat wrong.

The key idea, via analogy

Consider two lottery-like businesses. Business A is a chain of umbrella shops: it does well in rainy years, badly in sunny ones. Business B is a chain of ice cream stands: sunny years are good, rainy years are bad. Both are individually volatile. Neither is safer than the other.

Now suppose the country's entire economy happens to boom when it is sunny. Then B's fortunes rise and fall with everyone else's, while A quietly makes money in exactly the years when everybody else is struggling.

Treynor's insight, arrived at independently and before Sharpe published it, is that these two businesses should not be discounted at the same rate, even though they look equally volatile in isolation. A is valuable because it pays when everything else is failing. It should be discounted gently. B piles on to the existing risk in everyone's portfolio and should be discounted harshly.

From this he derives the same conclusion the CAPM is famous for: only the part of a company's risk that moves with the broad market matters for its value. The rest is diversifiable and therefore free. And he derives the linear relationship between that co-movement and required return that every finance student now learns as the security market line.

What is remarkable is not just that he got there first, but that he framed it as a valuation problem from the beginning, which is arguably the most useful framing. Treynor's version speaks directly to the question a CFO actually asks: what discount rate do I use for this project?

Why it mattered

• He was first, or effectively tied for first. Sharpe, Lintner and Mossin all deserve their credit for independent derivation, but Treynor's manuscript predates Sharpe's published paper and Sharpe himself acknowledged its existence. Careful histories of the CAPM always name Treynor.
• The Treynor ratio came from the same thinking. Treynor separately proposed judging a portfolio by its excess return divided by its beta rather than by its volatility, which is the right measure when the portfolio is one piece of a larger diversified holding rather than someone's whole net worth. That measure is still taught and used.
• He kept building. Treynor went on to co-author the market-timing test with Kay Mazuy, to develop the Treynor-Black model for combining active bets with a passive core, and to write extensively and idiosyncratically on market microstructure and on the economics of the dealer function. He was one of the most original thinkers finance has produced and spent his career outside the academy.
• It is a lesson in how credit works. The story of this paper is the standard cautionary tale about publication: an idea that is not written down in a citable place effectively does not exist, no matter how good it is or how many people have read it.

The honest limitations

• It carries all the CAPM's flaws. Single period, shared beliefs, frictionless markets, unlimited borrowing at the risk-free rate. Treynor's version is not immune to any of the criticisms that later broke the model, from Roll's untestability argument to the persistently flat empirical beta-return line.
• The manuscript is rough. It was a working draft, not a polished journal article, and Sharpe's published treatment is cleaner and Mossin's more rigorous about equilibrium. Priority is not the same as best exposition.
• The date itself is contested. Different sources cite the manuscript as 1961 or 1962, and there were multiple versions in circulation. That vagueness is exactly what happens when a paper is never formally published, and it is part of why the attribution question has never been fully settled.
• Being right first is not the same as convincing the field. Whatever the merits, it was Sharpe's paper that changed how finance was taught, because it was in a journal where people could find, read, and cite it.

The one-line takeaway

Treynor derived the CAPM independently and arguably first, framing it as the practical question of what discount rate a risky business deserves, and then, by never publishing, handed the credit to everyone else.