What does beta measure?
Beta compares a stock with a benchmark, usually a broad index such as the S&P 500. It answers one question: when the market went up or down, how far did this stock tend to go with it? Above 1, the stock has amplified the market’s moves; between 0 and 1, it has damped them; near 0, it has moved mostly on its own; below 0, it has tended to move against the market, which is rare for stocks.
Beta measures only market risk, the part of a stock’s ups and downs that comes with the whole market and that diversification cannot remove. The company-specific part, from earnings, lawsuits or new products, is not in beta at all.
- Beta
- A measure of how much a stock’s returns have tended to move with the market’s, estimated from past data as covariance with the market divided by the market’s variance. The market itself has a beta of 1.
- Market (systematic) risk
- The part of an investment’s risk that comes from moves of the whole market, such as recessions or changes in interest rates. Diversification cannot remove it.
- Company-specific (idiosyncratic) risk
- The part of a stock’s risk that comes from the company itself, such as its results, products or legal problems. Holding many stocks reduces it.
How to calculate beta: covariance divided by variance
Beta = covariance(stock, market) ÷ variance(market). Variance measures how widely the market’s returns spread around their average; covariance measures how the stock’s returns and the market’s move together around their averages. Here are five hypothetical monthly returns, in percent: the market 2, 0, 1, −2 and 4, an average of 1; stock XYZ 3, 0, 2, −2 and 7, an average of 2. The drawing below sets out each step:
- Deviations from the average. Market: 1, −1, 0, −3 and 3. Stock: 1, −2, 0, −4 and 5.
- Multiply each month’s pair and add: (1 × 1) + (−1 × −2) + (0 × 0) + (−3 × −4) + (3 × 5) = 1 + 2 + 0 + 12 + 15 = 30.
- Square the market’s deviations and add: 1 + 1 + 0 + 9 + 9 = 20.
- Divide both by the number of months minus one, 4: covariance = 30 ÷ 4 = 7.5 and variance = 20 ÷ 4 = 5.0.
- Beta = 7.5 ÷ 5.0 = 1.5. Because both use the same divisor, 30 ÷ 20 gives the same answer.
- Variance
- The average of the squared deviations of returns from their mean, a measure of how widely they spread. Its square root is the standard deviation.
- Covariance
- The average of the products of two series’ deviations from their means. Positive when they tend to be above or below average at the same time.
Beta vs volatility: not the same thing
Volatility, usually measured as the standard deviation of returns, tells you how much a stock moves, in any direction and for any reason. Beta tells you how much of that movement follows the market. They are linked by correlation: beta = correlation × (the stock’s volatility ÷ the market’s volatility).
In the drawing below, the market’s annual volatility is 16%. Hypothetical stock A, a small mining company, swings a lot, with volatility of 40%, but mostly on its own news, with a correlation of only 0.12 with the market: beta = 0.12 × 40 ÷ 16 = 0.30. Hypothetical stock B, a large industrial company, has volatility of 20% and a correlation of 0.96: beta = 0.96 × 20 ÷ 16 = 1.20. A is far riskier on its own and yet has the lower beta. A low beta does not mean a calm stock.
- Volatility
- How much an investment’s returns vary, usually measured as the annualized standard deviation of returns. It counts all moves, whether or not they follow the market.
Portfolio beta and typical betas by industry
A portfolio’s beta is the weighted average of its holdings’ betas. Hypothetical: 50% in a stock with a beta of 1.5, 30% in an index fund with a beta of 1.0 and 20% in a stock with a beta of 0.5 give 0.5 × 1.5 + 0.3 × 1.0 + 0.2 × 0.5 = 0.75 + 0.30 + 0.10 = 1.15. If the market fell 10%, the market-related part of the portfolio’s move would be about −11.5%, before anything company-specific.
Betas differ by industry. In data compiled by Aswath Damodaran of NYU Stern as of January 2026, the average beta of US system and application software companies was 1.28, of REITs 0.64, of regional banks 0.40 and of general utilities 0.24, with an average across the whole US market of 0.91. A simple average of many companies of different sizes need not equal 1.0, and the figures change from year to year.
- Portfolio beta
- The weighted average of the betas of a portfolio’s holdings, using each holding’s share of the portfolio’s value as its weight.
Where beta comes from, and its limits
Beta is the risk measure of the capital asset pricing model (CAPM) of William Sharpe (1964) and John Lintner (1965), which says investors are paid only for market risk, so a stock’s expected return should rise with its beta. The model is still widely used, for example to estimate a company’s cost of equity in a discounted cash flow valuation. But Fama and French (2004) describe its empirical record as poor: over long periods, high-beta stocks have not earned as much more than low-beta stocks as the model predicts.
Use beta with its limits in mind:
- It looks backward. It describes how a stock moved in the past, and a company that changes its business, debt or size can behave differently.
- It depends on the choices behind it: the index, the period and the frequency of returns. Data services commonly use five years of monthly returns or two years of weekly returns, so two websites can show different betas for the same stock.
- It drifts toward 1. Blume (1975) found that betas estimated in one period tend to be closer to 1 in the next, which is why some services publish an adjusted beta that pulls the raw figure toward 1.
- It is an average. On any single day the actual move can differ a lot from what beta suggests.
- It ignores company-specific risk, the risk that hurts most in a concentrated portfolio, and says nothing about how far a stock can fall on its own news.