The hypothesis that explains most market behavior
The Efficient Market Hypothesis (EMH) says: market prices already reflect all available information. So you can't consistently earn excess returns by picking stocks based on public information — by the time you act, the price has already adjusted.
Three forms (each stronger than the last):
- Weak form: prices reflect all past price information. Technical analysis (chart patterns, moving averages) shouldn't work consistently.
- Semi-strong: prices reflect all public information. Fundamental analysis (reading reports, news) shouldn't generate consistent alpha.
- Strong form: prices reflect even private information. Insider trading shouldn't work either.
Most economists agree weak form is approximately true and semi-strong is mostly true (for liquid markets). Strong form is widely rejected — insider information clearly works (which is why insider trading is illegal).
Why EMH explains the active vs passive data
Last lesson showed that ~80% of active funds underperform passive after fees. EMH explains why: if markets are mostly efficient, it's hard to identify mispricings. The information edge needed to beat passive is rare and hard to maintain. Most active managers don't actually have edge — they just charge fees as if they did.
This isn't a controversial result among finance academics. It's the consensus view of how liquid public markets behave on average. Active management can sometimes add value (lesson 9-6's exceptions), but for the broad investable market, passive captures most of what's available.
Where EMH cracks — the limits
EMH isn't perfect. Several real-world phenomena suggest markets aren't fully efficient:
- Bubbles and crashes. Tulip mania, dot-com 2000, 2008 housing — each looks like prices wildly disconnected from fundamentals.
- Behavioral biases. Herding, overconfidence, loss aversion (Track 10) demonstrably affect prices.
- Anomalies. Small-cap effect, value effect, momentum effect, post-earnings announcement drift — patterns that consistent with EMH would predict away.
- GameStop January 2021. Reddit's coordinated retail buying drove a stock to obviously irrational valuations. Pure EMH would say markets shouldn't allow this; they did.
- Limits to arbitrage. Even when smart money sees mispricings, they can't always trade them away (constraints on capital, time horizons, prime broker calls).
The behavioral finance counter-argument
Behavioral finance (Kahneman, Tversky, Thaler) shows that investors are systematically biased — overconfident, loss-averse, herding, anchoring. Markets aggregate biased decisions, so prices can deviate from fundamentals for sustained periods. The corrections come (eventually), but in the meantime there are real opportunities — for those who can identify them and stay disciplined.
This doesn't completely overturn EMH; it qualifies it. Markets are mostly efficient most of the time, but with persistent inefficiencies in specific corners and during extreme times. Smart investors can sometimes exploit these — but it's harder than it looks.
Practical takeaway
Treat broad efficient markets as roughly EMH-compliant. Don't try to beat the S&P 500 by picking individual large-cap US stocks; the market is too efficient. But:
- Less-efficient corners (small-cap, emerging markets, distressed) may offer alpha for skilled managers.
- Behavioral patterns (momentum, value) sometimes generate persistent excess returns — but they're known and competed away by quants.
- Don't try to time markets based on "I think we're in a bubble" — markets can stay irrational longer than you can stay solvent.
The takeaway
EMH = market prices reflect available information; consistently beating the market is hard. Strong empirical support (active funds mostly underperform) and theoretical foundation (Sharpe's arithmetic). Real markets show some inefficiencies (bubbles, behavioral patterns, GameStop-style episodes), so EMH isn't absolute. Practical implication: passive default for retail; active only with a real reason. Markets are mostly efficient, mostly of the time, with exceptions.