What it means
The Capital Asset Pricing Model says expected returns line up against beta measured on the market portfolio of every asset. Richard Roll's 1977 critique asked a fatal question: which market portfolio?
Roll's argument is mathematical, not rhetorical. The CAPM's testable content reduces to whether the portfolio used is mean-variance efficient, and any efficient portfolio produces the same tidy beta-return line.
The true market portfolio contains everything: stocks, bonds, real estate, human capital, art. Nobody observes it, so every empirical test substitutes an index, and the test verdict attaches to the index, not the theory.
Roll's original paper states the trap plainly: testing the CAPM is equivalent to testing the efficiency of the proxy, and the theory itself is not independently testable until the whole market is measured. The critique rewired empirical finance: anomalies against a stock index might be CAPM failures or merely proof the index is inefficient, and nobody can fully separate the two.
Defenders respond pragmatically: with broad enough proxies the model organises data usefully, even if the deep test remains impossible, and practitioners use beta as a working tool regardless. The critique's afterlife is healthy scepticism: multi-factor models, alternative benchmarks, and humility about what any single index can prove all trace back to Roll's central observation.
For a non-finance reader, Roll's critique is the caution that a theory priced against an unmeasurable benchmark can never be convicted or acquitted, only approximated. The human-capital piece of the true portfolio is the one nobody can even approximate: lifetime earning power dwarfs listed equity for most households, and it trades on no exchange.
Benchmark providers have grown entire businesses from the ambiguity: every new index is another candidate proxy, and every proxy dispute is Roll's point wearing a commercial suit. Pedagogy absorbed the lesson too: finance courses now teach the CAPM with its testability caveat attached, a rare case of one paper permanently changing how a model is introduced to students.
In practice
Real-world examples.
Example
A fund's alpha flips from significant to zero when the benchmark switches from a national index to a global blend. The verdict followed the yardstick, not the manager's skill.
Example
A researcher notes a CAPM anomaly may indict the index's efficiency rather than the model. She reports both interpretations in the paper instead of declaring the model rejected.
Example
A performance report names its benchmark and footnotes Roll's critique as a standing limitation. The footnote stayed in every report, so readers knew the alpha figure depended on the chosen index.
Formula
Calculation
No new formula; Roll showed the CAPM's security market line is mathematically equivalent to the mean-variance efficiency of the chosen benchmark, so testing one is testing the other, and the true all-asset market portfolio is unobservable.
Worked illustration using the standard CAPM alpha calculation, alpha = actual return - (risk-free rate + beta x (benchmark return - risk-free rate)). A fictional fund returns 10% with a 3% risk-free rate. Against a narrow national index returning 8%, with a fund beta of 0.9, expected return is 3% + 0.9 x (8% - 3%) = 7.5%, so alpha is 10% - 7.5% = 2.5% and the fund looks skilful.
Against a broader global index returning 10%, with a fund beta of 1.0, expected return is 3% + 1.0 x (10% - 3%) = 10%, so alpha is 10% - 10% = 0%. The fund, the period and the risk-free rate are identical. Only the yardstick changed, which is exactly Roll's point.Case study
Seen in the real world.
This case study is fictional and illustrative. A made-up asset manager's research head runs a textbook CAPM test on her firm's flagship fund: regress returns against the national stock index, find the alpha positive and significant, and prepare a marketing page declaring skill. Her quant deputy objects with Roll's 1977 paper open on the desk. The objection reshapes the analysis: the positive alpha might mean the manager beat the market, or merely that the index used is an inefficient stand-in for the true portfolio, which would make the alpha a benchmark artefact rather than skill.
They rerun the test against a broader global equity index, then a stock-bond blend, and the alpha migrates from significant to nothing, a verdict as damning of the test as of the fund. The marketing page is rewritten to describe the benchmark and its limits honestly, and the research head adopts a house rule: every performance claim must name its benchmark and carry a footnote acknowledging that Roll's critique applies to it. Her summary in the annual research letter is the critique in one breath: we can measure the fund against any yardstick we choose, but we cannot measure it against the market, because nobody has ever seen it.
Watch out
Common mistakes.
- Reading the critique as CAPM disproof; Roll showed the model is untestable as posed, not that it is false, and beta remains a working tool.
- Treating index betas as market betas; every empirical beta is measured against a proxy, and the proxy choice is a modelling decision, not data.
- Ignoring the benchmark in performance claims; alpha without a named, defended benchmark inherits every ambiguity Roll identified.
Questions
People also ask.
What is Roll's critique?
Richard Roll's 1977 argument that the CAPM cannot be truly tested because the all-asset market portfolio is unobservable, so tests only judge the proxy index.
Why is the market portfolio unobservable?
It includes every asset worldwide, including human capital and private assets, which no dataset measures; every test substitutes a narrower index.
Does it kill the CAPM?
No; it makes the model untestable rather than false, and practitioners still use beta while treating benchmark choice as consequential.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%