Beta is a number that shows how sensitive an asset’s returns are to moves in a chosen market index. A beta of 1 means the asset tends to move in line with the market, while numbers above or below 1 point to bigger or smaller swings than the market.
Traders and investors use beta to compare risk, to build hedges and, in models such as the Capital Asset Pricing Model, to estimate a fair return for taking market risk. It does not predict direction on its own. It measures co-movement.
What does a beta of 1, 0, negative or 2 mean in practice?
Beta is unitless. Think of it as a multiplier on market moves, on average, over the period used to calculate it.
- Beta = 1: the asset has moved roughly in step with the index used. If the index rose 1%, the asset typically moved about 1% the same way.
- Beta > 1: more sensitive than the market. A beta near 2 suggests about twice the market’s percentage move, up or down.
- Beta between 0 and 1: less sensitive. The asset still tends to follow the market, just with smaller swings.
- Beta = 0: no consistent relationship with the market over the lookback window. Moves are largely uncorrelated.
- Negative beta: the asset has tended to move opposite to the market. This is uncommon and often unstable through time.
In a broad bear market, high-beta shares often fall harder. In strong risk-on phases they often rally more. That pattern is statistical, not a guarantee for any single day.
How is beta calculated?
Most providers estimate beta from historical returns using ordinary least squares regression. The slope of the line that best fits the asset’s returns against the market’s returns is the beta. You will also see the simpler covariance form:
Beta = Covariance(asset, market) ÷ Variance(market)
Key choices that affect the number:
- Return frequency: daily, weekly or monthly returns.
- Lookback period: for example two or five years of data.
- Market proxy: the index used as “the market”, such as a national equity index or a sector index.
- Currency and dividends: whether returns are in local currency and whether dividends are included.
A quick example. Suppose over many observations the covariance of Stock A with the index is 0.0020 and the index variance is 0.0010. Beta is 0.0020 ÷ 0.0010 = 2.0. Stock A has been about twice as sensitive as the index over that period and sampling choice.
Which market and lookback should you use?
Beta is only meaningful relative to the benchmark chosen. A UK retailer may show a different beta against a global index than against a UK retail index. For crypto, using a broad crypto market index can give a very different reading from using a single large coin as the benchmark.
Frequency and window matter too. Daily data can be noisy and reflect short-term microstructure effects. Monthly data smooths noise but reduces the number of observations. Short windows react quickly to regime changes, yet they can be unstable. Long windows are steadier but may reflect outdated relationships. There is no single right answer, which is why providers often publish their own methodologies and why two screens may show different betas for the same stock.
Levered, unlevered and adjusted beta
Levered beta (the common listing on data screens) reflects the firm’s capital structure. More debt usually amplifies equity sensitivity to the market, which lifts levered beta.
Unlevered beta or asset beta strips out the effect of capital structure to show the business risk of the assets themselves. Analysts often unlever a set of comparable companies to find a peer group average, then reapply a target capital structure to estimate an appropriate beta for valuation.
Adjusted beta is a blend that pulls the raw estimate towards 1.0 to account for the tendency of betas to drift back towards the market over time. The exact formula varies by provider. A common style is to weight the latest beta heavily and add a smaller weight to 1.0.
How investors and traders use beta
- Comparing risk: A high-beta share is expected to be more volatile with the market cycle than a low-beta, defensive name. Portfolio builders can balance exposures by mixing betas.
- Hedging: If your stock portfolio has an estimated beta of 1.2 to an index, you can size an index hedge accordingly. For example, to hedge £500,000 of exposure you would hedge about £600,000 notionally in the index, subject to contract specifics and slippage.
- Setting expected returns: In CAPM, expected return equals the risk-free rate plus beta times the market risk premium. The logic is simple. Take more market risk, demand more return. The residual performance beyond what beta explains is often labelled alpha.
- Portfolio construction: The beta of a portfolio is the weighted average of the betas of its holdings, after accounting for cash or derivatives. Low-beta overlays, market-neutral strategies and long-short funds all manage beta deliberately.
Beta is also used to bucket funds. An equity income fund may target a beta below 1. A small-cap growth fund might run higher beta by design.
Worked example: reading beta on a fact sheet
Imagine a fund factsheet shows a three-year beta of 0.7 to its benchmark. That suggests the fund’s returns have moved 30% less than the market, on average, over that period. If the index rose 10% in a quarter, the fund might be expected to be up about 7%, before fees, assuming the relationship holds. If the index fell 5%, a 3.5% fall would be consistent with that beta. The manager may be holding more defensive sectors or running cash, which would also show up in the fund’s beta.
Limits and common pitfalls
- Beta is backward looking: It reflects the sample used. Structural changes at a company, a new product mix or a shift in leverage can all make the last beta a poor guide to the next one.
- Choice of benchmark matters: A stock can have different betas to different indices. Be clear which benchmark underpins the number you are quoting.
- Non-linearity: Beta assumes a mostly linear relationship. Some assets behave differently in stress, liquidity squeezes or at price limits, which a straight-line fit can miss.
- Low or negative betas can be unstable: Short windows, thin trading and noisy data can produce betas that look precise but move around a lot from month to month.
- Provider differences: Methods vary on data cleaning, return frequency, dividends and adjustments. Expect discrepancies across terminals or websites.
- Beta is not total risk: It captures market-related risk. Company-specific risk shows up as tracking error around the line of best fit. Two stocks can share the same beta yet have very different total volatility.
Quick reference: interpreting typical beta ranges
| Beta range | Typical behaviour vs market |
|---|---|
| Above 1.5 | Very sensitive, tends to amplify market swings |
| 1.0 to 1.5 | More sensitive than market, cyclical tilt |
| 0.5 to 1.0 | Somewhat defensive, smaller swings |
| 0 to 0.5 | Low linkage to market over the window used |
| Below 0 | Tends to move opposite to market, often not persistent |
Used with care, beta is a handy shorthand for market sensitivity. Always pair the number with knowledge of the index, period and method that produced it.