In 1961, at the Green Bank Observatory in West Virginia, Frank Drake wrote seven variables on a chalkboard. Multiplied together, they were supposed to estimate the number of civilizations in the Milky Way with which humans might communicate. The rate of star formation. The fraction of stars with planets. The fraction of planets that could support life. The fraction where life actually arises. And so on, down to the expected lifetime of a broadcasting civilization. Drake was explicit, then and afterward, that he had not discovered a method for counting aliens. He had written down a structured confession of ignorance — a way of organizing the things we would need to know in order to have an answer, as a hedge against pretending we already had one.
Sixty-five years later, a second-year associate at a mid-tier investment bank opens Excel and builds a discounted cash flow model. She projects revenue growth for ten years, margin expansion, working capital assumptions, capex intensity, a tax rate, a terminal growth rate, and a weighted average cost of capital assembled from a risk-free rate, an equity risk premium, a beta estimated from a historical window chosen by convention, and a cost of debt adjusted for a tax shield. Then, because the initial output is too low, she tweaks the beta and nudges the terminal growth rate up by fifty basis points until the output matches the multiple the managing director already promised the client. It appears on page forty-three of a pitch book. Nobody in the room questions whether it is science.
This essay is about which of these two people is doing the more epistemically honest work, and why the answer reveals a fundamental delusion in modern finance.
Consider the structural similarity first. Both the Drake equation and a DCF are multiplicative chains in which the output’s uncertainty compounds through the terms. Drake has seven factors. A standard DCF, once you decompose the cost of capital into its constituents and terminal value into its growth and discount assumptions, has at least that many, and often more. In both cases the terms are not independently measurable in any rigorous sense. In both cases small changes in individual parameters produce large swings in the output. In both cases the result is typically presented as a point estimate, with sensitivity tables offered as a concession to the fact that the point estimate is a fiction.
Now consider where the two models diverge, because this is where the case against the DCF actually gets interesting.
Drake’s parameters, for all their current untestability, point at features of the physical universe. The fraction of stars with planets is a real number that exists independently of our ability to measure it, and in fact the Kepler mission and its successors have collapsed the uncertainty on that particular term by several orders of magnitude since 1961. The astronomical parameters in Drake’s equation are the kind of thing telescopes can, in principle and increasingly in practice, resolve. The equation is a scaffolding that awaits better inputs, and its authors have always said so.
A DCF’s terminal growth rate, by contrast, is not a feature of the universe awaiting discovery. It is a claim about how a particular business, embedded in a particular competitive landscape, in a particular regulatory regime, under particular technological conditions, will perform in perpetuity. There is no instrument that will ever measure it, because the thing it purports to measure does not exist as a stable feature of the world. The same is true, in softer forms, of every other input. Beta is estimated from a window; choose a different window and you get a different beta. The equity risk premium is the subject of a literature that cannot agree on whether the number is three percent or seven. The cost of capital is a weighted average of two numbers, one of which is defensible and the other of which is a convention.
The philosophical situation, then, is roughly this. The Drake equation uses honest guesses about things that exist to estimate a quantity nobody trades on. The DCF uses confident point estimates about things that do not exist as stable quantities to produce valuations on which billions of dollars move every day. If you had to pick which of these intellectual objects is more deserving of the adjective “rigorous,” the astronomers would win on several grounds at once: their parameters have referents, their uncertainty is acknowledged, and their methodology has collapsed in precision over time as measurement has improved. The DCF has done none of these things in sixty years.
There is a standard rebuttal here, and it is worth taking seriously. The rebuttal is that the DCF does not have to be right in any absolute sense; it only has to be less wrong than whatever the market is implicitly assuming. Price provides a forcing function that SETI simply does not have. An analyst whose near-term cash flow assumptions are wrong will be violently corrected by quarterly earnings.
But while Years 1 through 3 face immediate market discipline, the bulk of a company’s valuation lives in the terminal value — a mathematical horizon where accountability goes to die. The epistemology is supposedly rescued by the feedback loop, but the feedback loop in finance is slow, noisy, and almost infinitely ascribable to other factors. Careers are short relative to the time horizons on which terminal-value assumptions resolve. Attribution is sufficiently murky that practitioners can maintain the appearance of rigor for entire careers without ever having a specific input falsified in a way they must personally account for. SETI does not have this luxury. If a probe lands on Europa and finds a fish, Drake’s equation does not get to claim it was directionally correct. Finance, by contrast, almost always gets to claim it was directionally correct, because the definition of “direction” is post-hoc.
So the DCF is not rescued by its track record. It is rescued by the fact that its track record cannot be adjudicated.
This generalizes well beyond the DCF. Much of the apparatus of professional finance consists of instruments that are treated as measurements when they are actually arguments in disguise. A credit rating is an argument. A factor exposure is an argument. An ESG score is an argument. A valuation multiple, stripped of its narrative, is an argument that a particular comparable set is in fact comparable. None of these are wrong to use; they are only wrong to mistake for what they are not. The category error is not in the existence of the models. It is in the cultural insistence that because the models produce numbers, the numbers are findings rather than conclusions of arguments whose premises could be contested.
If we strip away the delusion that we can project a static competitive and regulatory landscape into perpetuity, what are we left with? We are left having to price in physical reality. An epistemically honest framework would not hide behind a terminal growth rate; it would contend with the actual forces that will determine whether a business still exists in twenty years, let alone in perpetuity, and it would accept that those forces are harder to model precisely because they are real. That kind of reality is messy, and it doesn’t fit neatly into a standardized spreadsheet cell.
The honest version of valuation work — and there is one — treats the model as a disciplined way of writing down your assumptions so that someone else can interrogate them. The number at the end is not the deliverable. The defensibility of the inputs is the deliverable. When a model is used this way, it is a tool for thinking. When it is used the other way, it is a tool for not thinking, and specifically for not having to defend in prose the claims that have been smuggled into the inputs.
Frank Drake, if asked whether his equation told us how many alien civilizations exist, would have said: of course not, it tells you what you would need to know in order to find out, and most of what you would need to know, we do not yet know. This is the correct answer, and it is the answer finance should give when asked what a company is worth. The model does not tell you. The model tells you what you would need to believe about the world in order to arrive at a given number, and whether those beliefs are defensible is a separate question that the model itself cannot settle.
The Drake equation asks whether we are alone in the universe and answers honestly that we do not know. The DCF asks what a company is worth and answers to two decimal places. Guess which one gets taken seriously on CNBC.
The difference between the two is not rigor. It is fees.
This material is provided for informational and discussion purposes only and reflects the philosophical views and opinions of the author as of the date indicated. It is not intended as, and should not be construed as, investment advice, a recommendation, or an offer to provide investment advisory services. The views expressed do not constitute a guarantee of any investment approach, result, or outcome.
Any references to valuation models, market practices, or industry conventions are general observations expressed for illustrative and educational purposes. They do not represent factual claims about the practices, rigor, or conduct of any specific firm, adviser, or market participant, nor should they be interpreted as comparative claims.
Green Alpha is a registered trademark of Green Alpha Advisors, LLC. Green Alpha Investments is a registered trade name of Green Alpha Advisors, LLC. Green Alpha also owns the trademarks to “Next Economy,” “Investing in the Next Economy,” “Investing for the Next Economy,” “Next Economy Portfolio Theory,” and “Next Economics.” Green Alpha Advisors, LLC is an investment advisor registered with the U.S. SEC Registration as an investment advisor does not imply any certain level of skill or training. Nothing in this post should be construed to be individual investment, tax, or other personalized financial advice. Please see additional important disclosures here: https://greenalphaadvisors.com/about-us/legal-disclaimers/