ValueThe one-line idea behind CAPM
Investors will not hold a risky stock for the same return they can get risk-free from a government bond. They demand compensation for risk, and CAPM prices that compensation in the simplest possible way: the risk-free rate, plus a premium scaled by how risky the stock is relative to the market.
Cost of equity = Rf + β × (equity risk premium)
The formula splits the required return into two pieces you can see in the stats above. The risk-free slice pays you for time: money today is worth more than money later, even with zero risk. The risk premium slice pays you for the stock's specific riskiness, measured by beta. A stock with β = 1.2 earns 1.2 times the market's risk premium on top of the safe rate, because it amplifies market moves by about 20%.
The equity risk premium, in plain English
The equity risk premium (ERP) is the answer to: how much extra per year do investors need to choose stocks over safe bonds? It is the single most argued-over number in valuation, so here is how practitioners actually pick it.
- The historical approach. Measure how much US stocks beat Treasury bonds over the last century or so. That excess return lands around 4-6% depending on the exact period and bond maturity used. This is where the common 5% default comes from.
- The survey approach. Ask finance professors and practitioners what they use. Surveys consistently cluster around 4-6% for the US, higher for emerging markets.
- The implied approach. Work backward: given current stock prices and expected growth, what premium must investors be demanding? This moves with the market and tends to be lower when stocks are expensive.
There is no single correct ERP, which is exactly why the calculator shows a sensitivity table. If your valuation only works at a 3% premium but breaks at 6%, you do not have a valuation; you have a bet on the premium.
Beta: what it measures and where it breaks
Beta measures a stock's wiggle relative to the market's wiggle. A β of 1.5 means the stock has historically moved about 50% more than the market in both directions. You can find published betas on Yahoo Finance, Bloomberg, or your brokerage, typically estimated from two to five years of monthly returns against an index like the S&P 500.
Three things to know before trusting a published beta:
- It is backward-looking. A beta estimated during calm years understates risk if the business model or leverage has changed. A company that just took on heavy debt is riskier than its five-year beta says.
- Levered vs unlevered. Published betas usually include the company's debt (levered beta), which is what you want for the cost of equity. If you are comparing business risk across companies with different debt levels, unlever the beta first.
- It only sees market risk. CAPM assumes investors hold diversified portfolios, so only risk that moves with the market gets priced. Company-specific risk (a bad CEO, a factory fire) is assumed diversifiable and earns no premium in the model. Whether you believe that determines whether you believe CAPM.
Where the cost of equity goes next
CAPM is rarely the end of the story. Its main job is feeding the equity half of WACC, the weighted average cost of capital, which blends the cost of equity with the after-tax cost of debt. WACC then becomes the discount rate in a DCF. So a one-point error in your CAPM inputs flows straight through to your intrinsic value estimate: with default inputs, moving the ERP from 5% to 6% raises the cost of equity from 10% to 11.2%, and that higher discount rate can cut a DCF value by 10% or more.
Worked example
Take a stock with β = 1.2, a 4% risk-free rate, and a 5% equity risk premium:
Ke = 4% + 1.2 × 5% = 10.0%
The market itself is expected to return 4% + 5% = 9%, and this stock demands one extra point because it amplifies market moves. If you instead believed the ERP should be 3% (an expensive market), the same stock's cost of equity drops to 7.6%, and every valuation built on it gets more generous. Same company, same beta, different required return: the premium assumption is doing real work.
Limitations
- One factor explains everything. CAPM says market risk is the only priced risk. Multi-factor models (size, value, momentum) and plain common sense suggest other risks get compensated too.
- The risk-free rate is not risk-free. Government bonds carry inflation risk and, outside a few countries, real default risk. Using a nominal Treasury yield also bakes in inflation expectations your cash flows must match.
- Beta is unstable. Betas drift over time and vary by estimation window and index. A cost of equity built on a shaky beta is a shaky hurdle rate.
- It assumes everyone agrees. CAPM is an equilibrium model: it describes what returns should be if all investors are rational and diversified. Markets frequently disagree.
Frequently asked questions
Should the risk-free rate be the 10-year Treasury or the 3-month bill?
Match the rate to your horizon. For valuing a long-lived business in a DCF, the 10-year Treasury yield is the standard choice. The 3-month bill fits short-horizon decisions. Mixing a short rate into a long valuation understates the cost of equity.
Can beta be negative?
In theory yes: an asset that reliably moves against the market (some gold miners, put options) has a negative beta and, per CAPM, a cost of equity below the risk-free rate. In practice, published stock betas are almost always positive.
Why not just use a fixed 10% required return?
You can, and many investors do use a flat hurdle rate. CAPM's advantage is that it adjusts the hurdle to the stock's risk: a sleepy utility should not clear the same bar as a volatile growth stock. Its disadvantage is that the adjustment rests on debatable inputs.
Does CAPM work for private companies?
There is no market beta for a private company, so practitioners borrow one: take the average levered beta of comparable public companies, unlever it to remove their debt, then relever it to the private company's target capital structure. Each step adds estimation error, so widen your sensitivity ranges.
What beta should I use in CAPM?
Use the stock's levered equity beta measured against a broad market index, typically from two to five years of monthly returns. A beta of 1.0 means market-level risk, while a beta above 1.0 means the stock is more volatile than the market. Unlevered betas must be relevered to the company's debt level first.