Behavioral Finance vs Efficient Markets - Who Actually Wins?
So after all the chapters, all the experiments, all the arguments back and forth, Burton and Shah finally ask the big question. Who wins? Is the market efficient or not? Time for a new theory entirely?
And the answer is… complicated. But also kind of satisfying.
The Scorecard
Here is where things stand. Over the past few decades, the behavioral finance crowd has gained serious ground. Even the strongest defenders of the efficient market hypothesis (EMH) have softened their positions a bit. Malkiel now says nice things about mean reversion. Fama admits there are open issues. And Kahneman, who did more damage to the EMH than almost anyone, actually tells people not to throw it out completely. It still has useful truths.
So nobody has declared total victory. But if you look at the scorecard issue by issue, the behavioralists are winning more rounds.
The Semi-Strong Hypothesis
Remember, the EMH comes in three flavors. The weak version says past prices can’t predict future prices. The strong version says prices reflect all information, including insider stuff. And the semi-strong version says prices reflect all publicly known information.
The weak version is mostly accepted. The strong version is basically untestable because insider traders don’t exactly publish their results. So the real fight is about the semi-strong version.
Malkiel’s defense is straightforward. If markets are inefficient and public information is not fully in prices, then smart fund managers should be able to beat simple index funds. But they can’t. Study after study shows that professional money managers, as a group, fail to beat basic market indices. Even when you ignore their fees. And the managers who did well last year are no more likely to do well next year than the managers who did badly.
That is a strong argument for the EMH.
But here is the thing. There is a pile of evidence on the other side too. Mean reversion, momentum, calendar effects, the book-to-market effect found by Fama and French. All of these are patterns that allow some predictability in stock returns. And they are simple strategies. Not complicated hedging schemes. Just basic rules that keep working, year after year, even after academics published papers about them.
EMH defenders say there must be some hidden risk that explains the extra returns. But decades have passed and nobody has identified what that risk is. If it was that risky, it should be easy to spot. It hasn’t been spotted.
The critics have the upper hand here.
Can Prices Move Without New Information?
If the EMH is right, prices should only move when new information arrives. So what happened on October 19, 1987? The stock market fell 22.6% in a single day. What information changed? Nobody has been able to answer that.
The whole year of 1987 was strange. Stocks rose 30% in the first half, crashed in October, then recovered to roughly where they started. If you only looked at January 1 and December 31, you would think nothing happened. But something wild happened in between, and no information story explains it.
Some blame program trading or portfolio insurance. But those mechanisms were known to the market. If markets were efficient, other traders would have been ready to cushion the fall. They didn’t.
Shiller’s “excess volatility” argument makes it worse. Stock prices are generally much more volatile than the fundamentals that are supposed to drive them. If information determines prices, prices should be about as jumpy as information. But they are not.
This is a serious problem for the EMH.
The Law of One Price
Can identical assets trade at different prices? Yes. The Royal Dutch and Shell case is the most famous example. Same company, same earnings, two different stocks that should have traded at a fixed ratio. They almost never did.
Shleifer and colleagues showed why. Limits to arbitrage mean that non-fungible assets can trade at different prices for a long time. Nobody can profitably force them together because the risks of arbitrage are too high.
Clear loss for the EMH. And if identical assets can be mispriced, imagine how far apart prices can get for things that are merely similar. That opens the door to bubbles like tech in the late 1990s and housing in the 2000s.
Three Research Programs, Three Problems for the EMH
The book organized behavioral finance research into three big buckets. And all three are bad news for the EMH.
Noise trader research established that there are real limits to arbitrage. Prices can stay wrong for a long time. Bubbles keep happening across centuries and nobody has a complete theory of why, but they clearly don’t fit the EMH story.
The Kahneman-Tversky psychology research showed that people don’t make decisions the way traditional economics assumes. Fama argues these biases cancel each other out in aggregate. And he has a point. Just because individuals are biased doesn’t automatically mean market prices are wrong. But loss aversion is different. It explains the equity premium puzzle, the disposition effect (selling winners and keeping losers), and several other things that traditional finance cannot explain at all.
Contrarian investing and calendar effects research found persistent patterns in stock returns. Mean reversion, short-term momentum, weekend effects, holiday effects. These patterns exist in stock markets around the world, which makes data-mining accusations harder to sustain. And they line up with what actual traders have believed for decades.
The Critics Hold the High Ground
Burton and Shah are blunt about it. At the current state of research, the EMH is in serious trouble.
Warren Buffett once said that in the short run, the market is a voting machine, but in the long run, it is a weighing machine. Sounds wise but doesn’t actually help, because “short run” and “long run” are vague. What matters is whether the deviations from efficiency are big enough and last long enough to matter for real decisions and real policy.
And the evidence says yes, they are.
But there is a catch. Behavioral finance does not yet have a replacement theory. It has punched a lot of holes in the EMH, but it hasn’t built something comprehensive to put in its place. Even the critics seem reluctant to fully abandon the EMH. So right now, it is more of a fight in progress than a settled outcome.
What We Actually Learned
After all this research, here is what is clear:
People are systematically biased. Human decision-making is inconsistent, easy to manipulate, and nothing like what utility theory assumes. Kahneman and Tversky proved this over decades. Thaler spread the word. Even pro baseball teams were making bad decisions until the Moneyball approach came along.
Loss aversion explains real puzzles. Why do people sell winners but hold losers? Loss aversion. Why do stocks need such high returns compared to bonds? Loss aversion. Why do people buy flood insurance only after a flood? Saliency. These are concrete explanations for things traditional finance could not explain.
Bubbles still have no good theory. Behavioralists are very interested in bubbles. They can explain why prices diverge from fundamentals. They can explain why arbitrage doesn’t fix things. But they still can’t fully explain why a bubble starts spiraling upward, or what exactly triggers the crash. The math is modern and clean, but the economics isn’t much deeper than what experienced traders have always said.
Markets are more predictable than they should be. Contrarian strategies and momentum effects keep showing up in the data, across countries, across decades. EMH defenders say these patterns will disappear once exposed. So far, that hasn’t happened.
Where Does This Go From Here?
Burton and Shah lay out the open questions. Understanding bubbles better, because heavy regulation after crashes might do more harm than good. Rebuilding decision theory to properly incorporate loss aversion, not just leave it as an ad hoc patch. Figuring out whether our biases hurt our welfare or sometimes help. And studying whether financial firms themselves are subject to the same biases as individuals.
Big questions. Far from settled.
The Final Thought
Here is where Burton and Shah land, and I think it is the most honest conclusion possible.
Use whichever theory fits the question you are asking.
Want to know why people sell winners and keep losers? Behavioral finance has the answer. Want to know why a stock with higher earnings growth has a higher price-to-earnings ratio? The EMH framework handles that better.
The right tool depends on the question. That might be unsatisfying if you want one grand theory that explains everything. But sometimes the honest answer is: we are not there yet.
After reading this whole book, that feels right to me. Not everything in markets is efficient. Not everything is irrational. The best investors probably know when to use which framework without getting religious about either one.
Previous: Experimental Economics and Market Bubbles