Dismantling the Efficient Market Hypothesis

If price emerges from the aggregation of belief how does this affect the claim that markets process information efficiently? The Efficient Market Hypothesis (EMH) asserts that asset prices reflect all available information, implying that consistently outperforming the market is nearly impossible.¹ The hypothesis presents three forms of increasing scope: weak (prices reflect past information), semi-strong (prices reflect all public information), and strong (prices reflect all information, including insider knowledge).² The availability of Information therefore dictates a valuation upon which markets converge, allowing prices to adjust instantaneously to new information. EMH also assumes that all investors consume this information identically and share similar expectations, rendering participants “rational” by default.

However, each form of EMH shares the same structural defect by assuming that broader information absorption guarantees convergence toward a correct price, conflating the availability of information with uniformity of interpretation. The strong form of EMH goes furthest, presuming that even privately held information somehow reaches the market; yet there are myriad examples of insiders profiting from asymmetries the market supposedly prices in. Beyond insiders, proximity to information creates its own advantage, a Cantillon-like effect, that EMH leaves unaddressed.

Policymakers privy to regulatory changes, or executives adjacent to strategic decisions, absorb context that cannot survive public transmission. Information degrades as it travels from source to market, and those nearest to the source act on a more context rich and complete dissemination of information than those who receive it downstream. The effects of informational proximity and latency are phenomena EMH does not account for within any of its three forms. Furthermore, the persistent outperformance of select investors, while often dismissed as survivorship bias or luck, is likely explained by superior interpretation of identical information; a nuance that further undermines claims to an efficient market.

In reality, the distribution of information is asymmetric, not asymptotic, and even when information is shared, its interpretation will diverge. This essential quality of information, and how it transmits through markets, grants nothing to EMH in any form. The existence or availability of information is an insufficient condition for price action; access to information does not guarantee agreement, eliminate the need for subjective interpretation, or lead to the commitment of capital. If the same information can support disparate beliefs, the market is better understood as reflecting a distribution of interpretations, not the result of the distribution of information. EMH would likely classify this distribution as an “inefficiency” or noise in the market, but it is this very heterogeneity of interpretation that explains diverse investment outcomes.

Proponents might argue that EMH requires only a sufficient majority of rational actors to drive prices; however, this defense merely shifts the locus of the problem and still presumes that rational actors can identify a correct price independent of interpretation and the subjective necessity we have already established. Information cannot mechanically dictate valuation any more than the market can independently process information; both cases would require human interpretation to acquire meaning. Yet EMH treats information as an exogenous force that arrives and then compels a response.³ Conceptually, this reduces notions of efficiency to articles of faith, possible only once sentiment has been hand-waved away.

The Efficient Market as Interpretive Arena

The central issue that runs through all three forms of EMH is that each tends to assume sentiment temporarily distorts prices until rational arbitrage converges the price towards an equilibrium and corrects mispricing.⁴ However, the value of assets depends on forward valuations, not present settlement. Markets, efficient or otherwise, only aggregate beliefs, and beliefs differ in intensity, distribution, and capitalization. Arbitrage and repositioning represents a shift in conviction, acting as a counter-bet that a new narrative interpretation will become the dominant view. Some describe this process as a market “correction”, but the word correction still assumes the very point at issue: that there is a “correct” price.

The conclusion here is not that markets are entirely random or that analysis lacks value. Information matters, but it matters through shaping belief rather than automatically establishing value. Skilled investors distinguish themselves less by access to information than by the ability to anticipate how markets will interpret that information, how sentiment will consolidate, and where conviction will eventually concentrate. EMH mistakes a single observable price as evidence of rationality, but the market rewards those who are positioned ahead of the next trend, not those who insist on defining it.

Roll’s Critique & The Joint Hypothesis Problem

In addition to its dismissal of sentiment, limits of information transmission and the distribution of interpretations, the Efficient Market Hypothesis contains a structural flaw that renders it empirically untestable. To perform a test of market efficiency, we must compare expected returns to actual returns, and expected returns require an asset-pricing model, such as the Capital Asset Pricing Model (CAPM), to be defined.⁵ The CAPM is a model used to quantify an asset’s sensitivity relative to the market through a metric known as Beta.⁶ Very simply, an asset with a Beta of 1.5 rises (or falls) 1.5% for every 1% move in the market.

From a methodological standpoint, CAPM assumes the existence of a market portfolio⁷ that contains every asset in the world. However this assumption is highly problematic and makes testing the validity of CAPM and its claims nearly impossible. Under CAPM, to truly measure an investment’s Beta, we would need to compare it against a portfolio that includes every stock, every asset, all real estate, precious metals, art collectibles, bonds, and perhaps even all potential human capital. Obviously, an all-encompassing market portfolio is an impossible benchmark, which renders the “market portfolio” an unobservable and untestable artifact.

Since we cannot observe the complete market, we use proxies such as the S&P 500; however, doing so inevitably changes the assumptions and results of any analysis we perform. The CAPM tests whether the chosen proxy is efficient, but not whether its assumptions are accurate. As a result, if the CAPM shows mispricing in a particular asset, there would be no way for us to determine if it is the asset that is relatively cheap, or if the S&P 500 is just a poor barometer. Thus, the assessment of whether prices reflect information correctly cannot be proven or disproven because the “market portfolio” assumption required to test it is an abstraction we cannot measure.⁸

This creates a dynamic where the market proxy delivering the best possible return for its level of risk leads us to assume that the CAPM functions as designed; conversely, if the proxy is inefficient, the calculations and results are meaningless. This circular relationship is the core of what came to be known as “Roll’s Critique”.⁹ Fama acknowledged this limitation, stating, “…we can only test whether information is properly reflected in prices in the context of a pricing model that defines the meaning of properly”.¹⁰ Although he presented this as a testing limitation, this highlights a fundamental problem within CAPM: what counts as “proper” is actually defined by the selected model, not by any objective standard.

Given Roll’s Critique and issues with the CAPM, we now see the Joint Hypothesis Problem laid bare. If observed returns deviate from expectations, there is no way to determine whether the market is inefficient or whether the pricing model is simply incomplete.¹¹ In assessing the EMH, when value stocks outperformed in ways the CAPM could not explain, the response was not to reject the claims of market efficiency but to expand the model to include a Three-Factor, and eventually, a Five-Factor asset pricing model.¹² The revisions allowed for dismissing anomalies as errors rather than market failures, thereby insulating the EMH from refutation. This served to absorb anomalous results into a revised definition of risk, preserving the notion of market efficiency by measuring it against adjusted benchmarks. This methodological goalpost shifting illustrates that the EMH cannot function as a feature of markets, given its self-preserving and conjecture-laden nature.¹³

Footnotes

1. Fama, Eugene F. 1970. “Efficient Capital Markets: A Review of Theory and Empirical Work.” The Journal of Finance, Papers and Proceedings of the Twenty-Eighth Annual Meeting of the American Finance Association, vol. 25 (2): 383-417. ↩

2. Fama, “Efficient Capital Markets: A Review of Theory and Empirical Work.” ↩

3. Shiller, Narrative Economics: How Stories Go Viral and Drive Major Economic Events. ↩

4. Pawelzik, Klaus, and Felix Patzelt. “An Inherent Instability of Efficient Markets.” Scientific Reports 3, no. 2784 (2013). https://doi.org/10.1038/srep02784. ↩

5. Fama, Eugene F., and Kenneth R. French. “The Capital Asset Pricing Model: Theory and Evidence.” Journal of Economic Perspectives 18, no. 3 (2004): 25-46. https://doi.org/10.1257/0895330042162430. ↩

6. Investopedia: Understanding the CAPM: Key Formula, Assumptions, and Applications ↩

7. Investopedia: Market Portfolio: Definition, Theory, and Examples ↩

8. Roll, Richard. “A Critique of the Asset Pricing Theory’s Tests Part I: On Past and Potential Testability of the Theory.” Journal of Financial Economics 4, no. 2 (1977): 129-76. ↩

9. Roll, “A Critique of the Asset Pricing Theory’s Tests Part I: On Past and Potential Testability of the Theory.” ↩

10. Fama, Eugene F. “Efficient Capital Markets: II.” The Journal of Finance 46 (1991): 1575-617. https://doi.org/10.1111/j.1540-6261.1991.tb04636.x. ↩

11. Fama, “Efficient Capital Markets: A Review of Theory and Empirical Work.” ↩

12. Fama, Eugene F., and Kenneth R. French. “A Five-Factor Asset Pricing Model.” Journal of Financial Economics 116, no. 1 (2015): 1-22. ↩

13. Schwert, G. William. “Anomalies and Market Efficiency.” University of Rochester - Simon Business School; National Bureau of Economic Research (NBER), 2002. https://ssrn.com/abstract=338080. ↩