The Efficient Market Hypothesis

If prices already reflect everything anyone knows, then trying to beat the market is a bit like trying to pick up a $100 bill that a crowd has already walked past.

In 1900, a French graduate student named Louis Bachelier submitted a thesis on the movement of prices on the Paris Bourse. His examiners were baffled. He was applying the mathematics of heat diffusion — the equations Einstein would later use for Brownian motion — to the price of government bonds. His conclusion was that, on any given day, a price is as likely to go up as down, and its wanderings look like the drunken stagger of a pollen grain in water. He called this a random walk. He was ignored for half a century.

When his thesis was rediscovered in the 1950s, it became the foundation of what is now called the Efficient Market Hypothesis (EMH): the claim that asset prices already incorporate all available information, and therefore future price changes must be, essentially, unpredictable noise. It is one of the most influential and most contested ideas in economics.

The core intuition

The argument is almost embarrassingly simple. Suppose a stock is obviously underpriced. Then somebody notices, buys it, and drives the price up until it is no longer obviously underpriced. The very act of a mispricing being spotted destroys the mispricing. In a market with many alert, self-interested traders, any publicly visible free lunch will be eaten before you can reach for a fork.

What survives, then, are prices that already reflect the collective knowledge of the market. The remaining price changes must be driven by new information — and new information, by definition, is not knowable in advance. So prices should move unpredictably. They should look like a random walk.

This has a strange consequence. The chart of any real stock, over any window, looks purposeful in hindsight — peaks, troughs, trends, reversals. But so do the simulated random walks above. Our pattern-hungry eyes will find a story in noise. The EMH claims that most of what looks like a story in market data is noise.

Fama and the three forms

The economist Eugene Fama gave the hypothesis its modern shape in 1970. He noted that “all available information” is vague, so he split it into three concentric levels. Each level is a stronger claim than the last.

The weak form says prices already reflect all past prices and trading volumes. If true, technical analysis — drawing trendlines, spotting “head and shoulders,” timing moving averages — cannot systematically beat a buy-and-hold. The evidence here is unusually strong. Statistical tests find price changes are close to uncorrelated over short horizons in developed markets.

The semi-strong form adds all public information: earnings, news, filings, macro data. If true, fundamental analysis of publicly reported numbers cannot yield an edge either — because the moment any of that information becomes public, prices swallow it within minutes. Event studies broadly support this. When a firm announces surprise earnings, the stock jumps almost immediately, not gradually.

The strong form adds private information. If true, even insiders cannot systematically profit. Almost nobody actually believes this — and the fact that insider trading is illegal, and prosecuted, is a tacit admission that private information does move prices.

The pushback

The EMH sat comfortably at the centre of finance for decades. Then problems started to accumulate.

Robert Shiller, working in the early 1980s, showed that stock prices are far more volatile than the underlying stream of dividends can justify — a phenomenon he called excess volatility. If prices were only responding to news about future dividends, they should swing less. They swing more. Something other than pure information is moving them.

“Markets can remain irrational longer than you can remain solvent.” — attributed to John Maynard Keynes

Then came the behavioural revolution. Kahneman and Tversky demonstrated that real humans systematically deviate from the calm Bayesian agent the theory assumed. They anchor, they overreact, they follow crowds. Under Richard Thaler and others, this hardened into an alternative view: prices reflect not just information but the psychology of the traders holding it. Bubbles — the 1920s, the dot-com boom, 2008 — became less “anomalies to be explained away” and more “the phenomena that any serious theory has to accommodate.”

The most striking rebuttal, however, was empirical. Consistent, decades-long outperformance by a small handful of investors — Warren Buffett, Renaissance Technologies, a few quant funds — is hard to reconcile with a strong version of EMH. If markets are perfectly efficient, such records should be indistinguishable from luck. They are not obviously so.

What survives

The verdict, forty years on, is nuanced rather than clean. Markets are not perfectly efficient — but they are efficient enough that beating them, after costs, is extraordinarily hard for almost everyone. The empirical record of active fund managers is brutal: over any twenty-year window, the large majority underperform a low-cost index fund tracking the market as a whole. The practical descendant of the EMH is not a philosophical claim but an ordinary piece of financial advice: don't try to pick stocks, just buy the whole market and hold.

Even Fama, on receiving the 2013 Nobel Prize, shared it with Shiller — who had spent his career arguing markets are systematically wrong. The committee's implicit judgement was that both were right about different things: prices are hard to beat in the short run (Fama), and prices can be wildly detached from fundamentals over the long run (Shiller). The Efficient Market Hypothesis, in the end, is less a fact about the world than a discipline: a warning about how much easier it is to think you have an edge than to actually have one.