Risk and Reward: Understanding Beta and CAPM

Chapter 9 of A Random Walk Down Wall Street opens with a quote from George Stigler: “Theories that are right only 50 percent of the time are less economical than coin-flipping.” That’s a warning shot. Malkiel is about to walk us through some fancy academic models. And then he’s going to tell us they don’t quite work the way everyone hoped.

But the core idea in this chapter is still useful. Risk has its rewards. The question is: how do you measure risk properly?

Beta: Putting a Number on Risk

By now you know that diversification reduces risk. But it can’t eliminate all of it. Stocks tend to move together. When the market drops, most stocks drop with it. When it rises, most come along for the ride.

Three academics, William Sharpe, John Lintner, and Fischer Black, wanted to figure out which part of a stock’s risk you can diversify away and which part you’re stuck with. Sharpe got a Nobel Prize for this work in 1990.

Their answer: risk breaks into two pieces.

Systematic risk is the part tied to the overall market. Recessions, interest rate changes, broad economic shifts. You can’t escape it by holding more stocks. It comes with the territory.

Unsystematic risk is the part specific to one company. A product recall, a fraud scandal, a surprise contract win. This risk is random and specific. And you can make it nearly disappear by owning enough different stocks. Around 30 to 60 well-diversified securities does the job.

Beta is the measure of systematic risk. It’s just a number that tells you how much a stock moves relative to the market. The market gets a beta of 1. A stock with a beta of 2 swings twice as much. When the market goes up 10%, that stock tends to rise 20%. A stock with a beta of 0.5 moves half as much.

High-beta stocks are called aggressive. Low-beta stocks are called defensive. Simple enough.

Beta became wildly popular in the early 1970s. Institutional Investor magazine put the Greek letter on its cover. The SEC endorsed it. Brokers who could barely do long division started tossing beta numbers around like they had PhDs in statistics.

The Capital Asset Pricing Model

Here’s where it gets interesting. The Capital Asset Pricing Model (CAPM) says that since unsystematic risk can be diversified away, the market won’t pay you extra for bearing it. You only get rewarded for systematic risk. For beta.

Think of it this way. Say you have two groups of 60 stocks. Both groups have the same beta of 1. But Group I stocks have tons of company-specific risk (weather exposure, currency swings, natural disasters) while Group II stocks have very little.

Old-school thinking said Group I should earn higher returns because those stocks are “riskier.” But CAPM says no. Once you hold all 60 stocks in each group, the company-specific risk washes out. Both portfolios end up with the same actual volatility. Same beta, same returns.

If Group I stocks did somehow offer higher returns, investors would pile in, bidding up their prices until returns equalized. The market self-corrects.

The practical takeaway of CAPM: want higher returns? Just increase your portfolio’s beta. You can do this by buying high-beta stocks or by investing on margin. Want less risk? Lower the beta. Mix stocks with bonds or savings accounts. A portfolio that’s half index fund and half government bonds gives you a beta of about 0.5.

Does It Actually Work?

This is the fun part. In theory, CAPM is elegant. Higher beta, higher returns. Clean line on a graph.

In practice? Not so much.

In 1992, Eugene Fama and Kenneth French published a study covering 1963 to 1990. They sorted all stocks into groups by beta. The result was devastating for CAPM believers. There was essentially no relationship between beta and returns. Stocks with high betas didn’t earn more than stocks with low betas.

Malkiel found the same thing when he looked at mutual funds. The financial press had a field day. “The Death of Beta.” “Bye, Bye Beta.” One anonymous quant wrote to Institutional Investor: “The Capital-Asset Pricing Model is dead.”

Malkiel’s Defense of Beta

Malkiel isn’t ready to write the obituary. He gives four reasons to hold off.

First, people clearly do care about volatility. If a safe government bond and a risky oil well paid the same return, nobody would drill for oil. Beta captures something real about how much a stock bounces around.

Second, UCLA professor Richard Roll pointed out that “the market” is impossible to measure perfectly. The S&P 500 isn’t the entire market. Include international stocks, bonds, real estate, commodities, and even human capital, and beta numbers change. How you define the market changes everything.

Third, over longer time periods (back to 1927), there is some positive relationship between beta and returns.

Fourth, even if low-beta stocks earn the same returns as high-beta stocks, that’s actually useful information. Buy low-beta stocks, get market returns with less volatility. That’s a win.

But Malkiel is honest: beta alone is not a crystal ball for future returns.

Arbitrage Pricing Theory

Stephen Ross came up with an alternative called Arbitrage Pricing Theory (APT). The idea: maybe a stock’s systematic risk is too complicated to capture with a single number.

Think about all the things that move stocks in ways you can’t diversify away. Changes in national income. Interest rate shifts. Inflation. Exchange rates. These all affect stocks systematically, but beta doesn’t capture them individually.

APT says returns depend on sensitivity to multiple risk factors, not just one. A Ford factory worker, for example, shouldn’t hold Ford stock. When Ford does badly, they lose their job and their investment at the same time. That’s a risk beta alone doesn’t address.

Early tests of APT showed some promise. Using multiple factors explained stock returns better than beta alone. But the model had its own problems and measurement difficulties.

The Fama-French Three-Factor Model

Fama and French didn’t just kill beta. They offered a replacement. Their three-factor model keeps beta but adds two more variables.

Size. Smaller companies tend to earn higher returns than larger ones. Maybe because small firms are more fragile during recessions, making them genuinely riskier.

Value. Stocks with low market prices relative to their book values tend to earn higher returns. These companies may be in financial distress. Think of major bank stocks in early 2009, trading at rock-bottom prices relative to book value. People genuinely worried they’d go bankrupt. That’s real risk, and the market priced it in.

Some researchers have added more factors since then. Momentum (stocks that are rising tend to keep rising). Liquidity (investors demand extra return for holding hard-to-sell stocks).

The Honest Conclusion

Malkiel wraps up Chapters 8 and 9 with a refreshingly honest admission. The stock market adjusts quickly to new information. Neither technical analysis nor fundamental analysis consistently beats it. The only reliable way to get higher long-term returns is to accept more risk.

But we still don’t have a perfect way to measure risk. Beta is useful but flawed. APT adds complexity but not certainty. The Fama-French model is better but still incomplete.

No single number captures all the ways the market can hurt you. Future risk models will probably be more sophisticated, not less. But there will never be a magic formula that predicts returns with certainty.

Malkiel closes with a joke about a woman who meets a genie offering three wishes. She gets a golden rocking chair, her youth back, and her cat turned into a handsome prince. The prince then asks, “Now aren’t you sorry you had me fixed?”

The moral: even if we found the perfect risk measure, we’d probably mess it up anyway.

Previous: Modern Portfolio Theory: Your New Best Friend Next: Your Brain Is Bad at Investing: Behavioral Finance Part of the series: A Random Walk Down Wall Street Book by Burton G. Malkiel | ISBN: 978-0-393-08169-5