How the Market Prices Risk: the CAPM

Markowitz taught us that smart investing means balancing return against risk, and that the best portfolios sit on an "efficient frontier." But he left a huge question hanging: if everyone follows this advice, what should any individual stock actually be worth? What return should a company have to offer to convince people to hold its shares?

William Sharpe answered that question in 1964, and the answer was so clean and so useful that it won him a Nobel Prize and became the single most-taught idea in all of finance. It's called the Capital Asset Pricing Model, or CAPM (say it "cap-M"). Here's the whole thing in plain English.

The problem: which risk actually gets rewarded?

Start with a puzzle. You'd expect risky investments to pay more, that's why anyone accepts risk. But which risk?

Imagine a tiny biotech company betting everything on one drug trial. That stock is wildly unpredictable. Now here's the surprising claim Sharpe made: the market will not pay you extra just for holding that wild, company-specific uncertainty. Why not? Because you could have made it disappear for free.

Think back to diversification. If you hold that biotech alongside a hundred other unrelated stocks, its private drama, the drug passes, the drug fails, the CEO quits, gets washed out. Some of your hundred stocks have good surprises, some have bad ones, and in a big basket these cancel out. Risk you can erase simply by diversifying is called idiosyncratic (company-specific) risk. And Sharpe's logic was ruthless: the market doesn't reward risk you could have diversified away for free. If it did, savvy diversified investors would pile in and bid the price up until the extra reward vanished.

So what risk does get rewarded? The kind you can't escape no matter how much you diversify: the risk that the whole market moves. When the economy tanks, almost every stock falls together. No amount of diversification saves you from a general crash. That undiversifiable, everyone's-in-the-same-boat risk is called systematic or market risk, and that is the only risk investors get paid to bear.

This is the heart of the CAPM: not all risk is equal. The only risk that earns a return is your exposure to the market as a whole.

The key idea via analogy: beta is a stock's "sensitivity dial"

If market risk is what matters, we need a way to measure how much market risk a given stock carries. That measure is beta.

Think of beta as a sensitivity dial that tells you how strongly a stock reacts when the whole market moves:

• Beta = 1: the stock moves in step with the market. Market up 10%, this stock tends to go up about 10%. (A big, broad company often looks like this.)
• Beta = 2: the stock is a market amplifier. Market up 10%, this stock lurches up around 20%, and down 20% when the market drops 10%. Twice as jumpy. (Think a high-flying, economically sensitive stock.)
• Beta = 0.5: the stock is a shock absorber. It only moves half as much as the market. (Think a utility or a staple-goods company people buy in any economy.)
• Beta = 0: the stock ignores the market entirely (a Treasury bill is the classic near-zero example).

In one sentence: beta measures how much of the market's roller-coaster a stock will make you ride. A high-beta stock exaggerates the market's swings; a low-beta stock dampens them.

The payoff: a straight-line rule for expected return

Here's where it all comes together into a rule so simple you can say it in a breath.

The CAPM says a stock's expected return equals the return on a totally safe investment, plus a bonus for market risk that is exactly proportional to its beta.

Written out, it looks like this:

Expected return = risk-free rate + beta × (market's extra return over the risk-free rate)

In plain English: start with what you'd earn on something safe like a government bond. Then add a risk bonus, and that bonus is just the market's own risk premium scaled up or down by the stock's beta. A beta-2 stock earns twice the market's risk bonus; a beta-0.5 stock earns half of it. That's it.

Picture it as a straight line. On the horizontal axis, put beta (how much market risk). On the vertical axis, put expected return. The CAPM says every fairly-priced asset sits on one straight line rising from the risk-free rate. This line has a name, the Security Market Line, and its message is beautifully simple: more market risk, proportionally more expected return, in a perfectly straight line. No other feature of the company matters for its return except that one dial, beta.

Why it mattered so much

The CAPM reorganized how the entire financial world thinks and talks.

• It gave everyone a common language for risk. Before Sharpe, "risky" was a vague adjective. After him, you could put a number on how risky a stock was relative to the market. To this day, when a finance website lists a stock's "beta," that's Sharpe's idea, sixty years later.
• It created the benchmark for judging skill. If the CAPM tells you the return a stock should earn given its risk, then anything a fund manager earns above that line is genuine skill (or luck) rather than just being paid for risk. That leftover is called alpha, and the entire active-management industry is, in effect, a hunt for alpha, a word we only have because of this model.
• It's how companies decide what projects are worth. When a corporation calculates whether a new factory is worth building, it needs a "discount rate", the return the project must clear to justify its risk. The CAPM is the standard textbook way to compute that. So Sharpe's model doesn't just price stocks; it quietly shapes real-world investment decisions across the economy.

The CAPM was also the seed of everything that came later. The idea that returns come from exposures to shared risks grew directly into arbitrage pricing, the Fama-French factors, and the whole modern field of "factor investing." CAPM was factor investing with exactly one factor: the market.

The honest limitations

The CAPM is elegant, and elegance came at the price of assumptions that don't survive contact with reality.

• It assumes a fantasy world. Everyone can borrow and lend at the same safe rate, there are no taxes or trading costs, everyone agrees on the same forecasts, and everyone can trade instantly. Reality is messier on every count.
• "The market" isn't a thing you can actually buy. The model needs the return of all risky assets everywhere, every stock, bond, house, and private business on Earth. In practice people substitute something like the S&P 500, which is only a slice. A famous critique (by Richard Roll) pointed out that this makes the theory awkward to truly test.
• The evidence is lukewarm. When researchers checked whether high-beta stocks really do earn proportionally higher returns, the relationship turned out weaker and flatter than the model promises. Worse, they found things beta can't explain: small companies and cheap "value" stocks have historically beaten what their betas predict. Those cracks are exactly what later papers, Fama-French, and the momentum and quality research, were built to address.
• Beta drifts. A stock's sensitivity to the market isn't a fixed constant; it wanders over time as the business changes, so even measuring it is slippery.

None of this means the CAPM is useless, it means it's a first approximation. It captures a genuine and deep truth (only undiversifiable risk gets paid) even if the precise straight-line formula is too tidy for the real world.

The one-line takeaway

The CAPM's enduring lesson is that you only get rewarded for the risk you can't diversify away, the risk of moving with the whole market, and a stock's "beta" tells you exactly how much of that risk you're signing up for.