The Efficient Market Hypothesis (EMH) is one of the most misunderstood ideas in modern economics.
To some, it’s a smug claim that markets are perfect, transactions are frictionless, and investors are robotically rational.
To others, it’s an ivory-tower excuse for speculative financial bubbles and the resulting inequality.
Despite decades of ridicule, EMH remains the single most useful framework we have for understanding how prices form. And how hard it is to outsmart the market. Granted, EMH is not complete – but that doesn’t mean it’s not useful.
A recent essay by Laurent Hynes, published by the decidedly Austrian-school Mises Institute, gives us a textbook case of why the EMH debate is so confusing. Hynes paints EMH as a quasi-religious belief in frictionless, omniscient markets – and then triumphantly declares it dead on arrival.
While I’m sure that logical judo domination of a straw man argument was rhetorically satisfying, it’s logically hollow. Hynes doesn’t defeat Fama’s hypothesis. EMH never even steps into the ring – instead, Hynes shadowboxes a caricature, then declares victory.
Today I’m going to set the record straight.
So what does Efficient Market Hypothesis actually say?
Eugene Fama’s 1970 formulation is disarmingly modest:
Security prices fully reflect all available information.
That’s it.
Not that markets are perfect. Only that prices incorporate what participants currently know and believe.
Because information arrives unpredictably, future price changes must also be unpredictable – hence the idea of a “random walk.” Fama’s theory is less famous than Malkiel’s popularization of it, via his 1973 book A Random Walk Down Wall Street.
Fama and later researchers distinguished three forms of EMH:
- Weak: prices reflect all past price data
Semi-strong: prices reflect all publicly available information - Strong: prices reflect all public as well as all private or insider information
(Most economists accept only the middle form as plausible – “weak” discounts data other than past prices; “strong” reflects Rumsfeld’s “unknown unknowns,” which is an incredibly difficult hypothesis to test.)
EMH offered the first cogent and coherent explanation of a well-observed phenomenon: While some investors or fund managers occasionally beat “the market,” few do so consistently – and persistent outperformance after fees is extraordinarily rare. Many index-fund studies confirmed this finding, much to the distress of active managers everywhere.
You can think of it as Kahneman and Tversky’s “reversion to the mean.” It also explains Barber and Odean’s finding that “the more you trade, the more you lose” – because every transaction comes with costs, and those costs reduce returns.
John Bogle wasn’t wrong – EMH explained why.
Now that we’ve marked the grounds, let’s dig into Hynes’s grotesque misunderstanding of EMH.
Shadowboxing a scarecrow
In Hynes’s essay The Efficient Market Hypothesis Is Fatally Flawed, he recites textbook simplifications of EMH: Identical expectations, costless transactions, perfect rationality – and then treats them as literal beliefs. He then concludes that, because real humans are messy and emotional, EMH must be nonsense – or “fatally flawed.”
To begin with, that’s the straw-man fallacy in purest form! First, build a cardboard caricature of your opponent’s view, then set it on fire and declare yourself the winner.
Hynes debated his own straw man, not EMH itself.
To be fair, we call these theories “models” for a reason. Just like the model airplanes I made as a kid, models simplify reality. A WWII-era Republic P-47 Thunderbolt had some 36,000 parts – and 25,000 rivets. My models had, at most, 200 parts (counting decals).
My model airplanes couldn’t fly – but neither did they consume thousands of worker-hours to construct.
Economists use models to isolate one causal mechanism. Just like we ignored friction in high-school physics when studying “perfect interactions,” economists ignore some factors when testing theories.
Criticizing EMH for setting aside some frictions is like calling Newton wrong because apples sometimes hit branches on their way to the ground.
A companion error – the Nirvana fallacy – runs throughout the essay. Contrasting real-world imperfection with a theoretical ideal and then declaring the theory false for not describing imperfection perfectly?
Come on! No model survives that standard!
Does that mean the concept of models themselves is flawed?
Well, no. It does mean we shouldn’t confuse the map for the terrain, though.
The right question is about comparative usefulness, not perfection.
What economic models can and can’t do
Models are tools for reasoning, not mirrors of reality.
The value of EMH lies in the discipline it imposes: before you can assume alpha, you must show that your information isn’t already factored into an asset’s current price. The inverse is also true: current asset pricing reflects all public information.
This baseline underpins nearly all asset-pricing work! From the Capital Asset Pricing Model to today’s multifactor frameworks.
Even the most famous EMH skeptic, Robert Shiller, built his analysis of market volatility on Fama’s framework.
Andrew Lo’s Adaptive Markets Hypothesis doesn’t replace EMH; rather, it explains when and why efficiency breaks down.
Behavioral finance, likewise, enriches the EMH model rather than abolishing it. Biases and feedback loops are deviations around an efficient core. That’s not the same as smoking-gun evidence that markets are random chaos.
That’s not to say that EMH is perfect, though.
Where EMH actually fails (and why its failures are just fine)
Critics are right about one thing: real investors are not rational calculating machines.
We herd, we panic, we anchor on past prices, we chase stories. We think Palantir is worth 450x its trailing earnings six months ago, and it’s worth 620x its trailing earnings today. We live in a world where there’s a so-called asset called Fartcoin, for the love of Christ! Which is up over 20% today – why?
Listen: anyone who’s spent more than ten minutes watching Jim Cramer prance around and throw things knows, deeply, that investor behavior is (often deeply, sometimes disturbingly) irrational.
I posit that these human flaws that occur in reality, but not in Fama’s model, don’t invalidate EMH. Rather, they define its boundaries.
Consider:
- Behavioral anomalies like momentum, overreaction, and speculative bubbles show where emotion temporarily overwhelms arbitrage
- Limits to arbitrage (Shleifer & Vishny 1997) explain why even rational traders can’t instantly correct mispricing: capital, leverage, and career risk constrain them
- Information asymmetries (Akerlof, Spence, Stiglitz) show that knowledge is both costly and unevenly distributed – yet prices still aggregate more information more efficiently than any central planner could
In short, EMH describes the gravitational center of market pricing. Those factors EMH doesn’t account for, behavioral and institutional forces, they describe the orbit of market pricing.
Yes, prices oscillate around efficiency – but efficiency remains the anchor.
Revisiting the Austrian “knowledge” critique
Broadly speaking, Austrian economists emphasize that knowledge is dispersed and tacit, not a single dataset shared by all. Ironically, that insight strengthens rather than weakens EMH.
Markets are precisely the mechanism that aggregates those scattered insights into prices.
When a farmer in Iowa or an engineer in Shenzhen acts on private information, their trades move prices infinitesimally closer to balance. In this way, even non-publicly available information becomes factored into prices. Market participants themselves make it happen.
Granted, the aggregation isn’t perfect. But it’s vastly more adaptive than any planner – or philosopher – could deliberately design.
The complaint that “If everyone believed EMH, no one would gather information” overlooks the built-in incentive.
Efficiency is not a fixed, static state! Rather it’s the outcome of countless people trying to prove that prices are inefficient – and, in the process, nudging them closer and closer to efficiency.
Their attempts to find mispricing are what keeps markets approximately efficient in the first place.
The joint-hypothesis problem considered
Hynes calls EMH “tautological” because empirical tests pair it with an asset-pricing model.
True: abnormal returns might reflect either inefficiency or a bad benchmark.
But that’s not circular reasoning – it’s simply a measurement challenge.
Measurement error doesn’t falsify a theory; it demands better instruments. Astronomers adjust their optics, and economists do the same – refining their lenses through successive models, from CAPM to Fama-French.
Why all the hate for EMH?
EMH attracts hostility from all sides – I think because it bruises everyone’s ego.
- Ideologues dislike it – the left, for dethroning central planning; the right, for its critique of heroic entrepreneurship.
- Fund managers hate it for implying that their skill is 1) mostly luck, and 2) not repeatable
- Austrian-school economists dislike its probabilistic view of uncertainty, which clashes with their deterministic conception of human action
Popularity isn’t a measure of truth.
Economists keep EMH and use its insights not out of dogma, but because no rival framework explains price behavior as parsimoniously. Or predicts the difficulty of beating the market as consistently.
Personally, I think of EMH this way:
- It’s the single most useful model for considering prices at any moment in time
- On its own, EMH is fundamentally flawed in its oversight of emotion (clustering, herd behavior, overreaction to news and so on)
Dogma and discipline (or babies and bathwater)
The right posture toward EMH is humility, not idolatry.
Its lesson is not “markets are perfect,” but “markets are hard to beat.” Prices embed more knowledge than any individual can process. Even so, they remain imperfect reflections of an uncertain future.
Understanding both sides of that paradox is what separates the seeker of knowledge from the ideologue.
As Andrew Lo put it:
“Markets are efficient when people are, and people are efficient when markets force them to be.”
That feedback loop – imperfect humans generating surprisingly efficient outcomes – is the true marvel of capitalism.
“It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own interest.”
– Adam Smith, The Wealth of Nations (1776)
Self-interest, channeled through voluntary exchange, produces broadly efficient outcomes without requiring perfect virtue or omniscient planning – just as EMH describes markets that aggregate private incentives into coherent prices.
And, like self-interest, EMH survives the bonfire of the straw men sacrificed to impugn it.
References
- Fama, E. F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance, 25.
- Shiller, R. J. (1981). Do Stock Prices Move Too Much to Be Justified by Subsequent Changes in Dividends? American Economic Review, 71.
- Shleifer, A., & Vishny, R. (1997). The Limits of Arbitrage. Journal of Finance, 52.
- Lo, A. (2004). The Adaptive Markets Hypothesis. Journal of Portfolio Management, 30.
- Thaler, R. H. (2015). Misbehaving: The Making of Behavioral Economics.