Fama-French and Predicting Stock Prices
In 1992 two economists published a paper that accidentally shook the foundations of modern finance. They did not mean to. They were actually trying to defend the system. But what they found in the data was so clear and so stubborn that it changed how everyone thought about stock prices.
Their names were Eugene Fama and Kenneth French. And Chapter 14 of Burton and Shah’s book explains what happened.
The CAPM Was Supposed to Explain Everything
Before Fama-French, there was the Capital Asset Pricing Model. CAPM. If you studied finance at any point in the last 50 years, you learned this model. The idea is simple. Every stock has a number called beta. Beta measures how much a stock moves with the market. High beta means the stock swings a lot. Low beta means it stays calm.
CAPM says beta is all you need. If you know the risk-free rate (like government bonds), the expected market return, and a stock’s beta, you can predict what that stock should return. One number. One model. Clean and elegant.
And here’s the thing. Fama and French set out to test whether beta actually works. They wanted to validate CAPM. They were supporters of efficient markets. Fama is basically the guy who defined the efficient market hypothesis in the first place. So this was not some outsider attack. This was an inside job.
What they found: beta did not really matter for predicting stock returns.
That was a problem.
What Actually Predicts Stock Returns
So if beta doesn’t work, what does? Fama and French looked at other variables. They mixed and matched numbers to see what had real predictive power. Two things stood out.
Size. Smaller companies tended to outperform larger companies. This was already known from earlier research. But it was not the big finding.
Book-to-market ratio. This was the big one. Companies with high book-to-market ratios had higher stock returns. Let me explain what this means because it sounds more complicated than it is.
Book value is basically what a company’s accounting books say the company is worth. You take assets, subtract liabilities, and you get net worth. Companies report this every quarter.
Market value is what the stock market says the company is worth. Share price times number of shares.
Book-to-market is just book value divided by market value.
When a stock price drops a lot but the company’s books haven’t changed much, book-to-market goes up. The company looks “cheap” relative to its accounting value. When a stock price shoots up, book-to-market goes down. The company looks “expensive.”
Fama and French found that cheap stocks (high book-to-market) beat expensive stocks (low book-to-market). Consistently. Over long periods of time.
Value Investors Loved This
People who follow value investing, going back to Benjamin Graham and David Dodd in 1934, have always said: buy cheap stocks. Look for companies the market has beaten down but whose fundamentals are still solid. Be patient. Wait for the market to recognize the value.
For decades, academics told these people they were wrong or just lucky. The efficient market hypothesis said you can’t systematically beat the market. All information is already in the price.
But now Fama and French, two of the biggest names in efficient market theory, were basically confirming what value investors had been saying all along. Cheap stocks do better. And you can identify them using simple numbers from public financial statements.
Value investors were thrilled.
The Mean Reversion Connection
Here’s where the story gets more interesting. Almost a decade before Fama-French, two other researchers named Werner De Bondt and Richard Thaler had found something similar. They looked at three-year stock performance and found that stocks that did badly over three years tended to do well over the next three years. And stocks that did great tended to slow down.
They called it overreaction. The market pushes prices too far in both directions, and then prices revert back toward some average. Mean reversion.
Their exact argument: if stocks systematically overshoot, the reversal should be predictable from past return data alone. No accounting data needed. Just look at what went up too much and what went down too much.
But nobody cared in 1985 when they published this. The academic world shrugged. People assumed that even if it was true, once the word got out, traders would exploit the pattern and it would disappear. Also, there were suspicions that the effect only worked with small, illiquid stocks where the data might not be reliable. And maybe transaction costs would eat up any profits.
So De Bondt and Thaler’s research just sat there. Ignored.
Then Fama-French Changed the Conversation
When Fama and French published in 1992, people started connecting the dots. Wait. If high book-to-market stocks outperform, and if stocks that have fallen a lot tend to have high book-to-market ratios… then maybe De Bondt and Thaler were onto something after all?
Some researchers wondered if mean reversion and the book-to-market effect were really the same thing seen from different angles. Stocks fall in price, their book-to-market goes up, and then they outperform. Is it overreaction correcting itself? Or is it some hidden risk factor being compensated?
Fama and French themselves were clear about where they stood. They did not think it was overreaction. They explicitly rejected De Bondt and Thaler’s mean reversion explanation. In their paper, they tested three-year past returns as a predictor and found it had no power. They believed that book-to-market and size were proxies for some unknown risk. If you hold these stocks and get higher returns, it must be because you’re taking on some risk the market is compensating you for. Rational. Efficient. No anomaly.
The problem? Nobody has ever found what that mystery risk actually is. It has been decades. Still no answer.
Why This Paper Mattered So Much
De Bondt and Thaler said stock prices are predictable, and the academic world ignored them. Fama and French said stock prices are predictable, and the academic world had a crisis.
The difference was who was saying it. De Bondt and Thaler were outsiders challenging the system. Fama and French were insiders. They were the system. When they found predictability, it carried enormous weight. You could not just wave it away as some fringe finding.
And the irony is beautiful. Fama and French had no intention of hurting the efficient market hypothesis. They were trying to test and validate CAPM. They genuinely believed their findings could be explained by unknown risk factors. But the effect on the field was the opposite. Their paper gave credibility to the idea that stock prices are not fully efficient. That past data can predict future returns. That simple rules can beat the market.
After Fama-French, there was a flood of new research exploring other types of predictability. The behavioral finance movement got real momentum. Researchers went back to older neglected papers and gave them a second look. The consensus that markets are perfectly efficient started to crack.
What to Take Away
The Fama-French three factor model says three things predict stock returns: the overall market (beta, but weakly), size of the company (smaller is better), and book-to-market ratio (cheaper is better). Two of these three factors were not part of CAPM.
For regular investors, the practical message is this. Value investing is not superstition. Buying companies that look cheap by simple accounting measures has historically produced better returns. Small companies tend to outperform big ones over time, though this effect is less consistent.
But be careful. “Historically produced better returns” does not mean “will always produce better returns.” And the debate about whether this is compensation for hidden risk or market inefficiency is still not settled.
What is settled: the CAPM, in its simplest form, does not work. Beta alone does not predict returns. The world is more complicated than one number. And sometimes the people who crack the system are the ones who were trying hardest to defend it.
Previous: The Illusions That Make Investors Overconfident