Take two stocks. Both are in your watchlist, and the market just fell 10%. Would you expect both to fall by the same amount? Beta gives you a way to compare them using their past returns.
A stock’s beta measures how sensitive its returns have been to a benchmark such as the S&P 500 or the DAX. A beta of 1.4 means a 1 percentage point move in the benchmark went with about 1.4 percentage points of movement in the stock, on average.
Put $1,000 next to each number. With the market down 10%, the market-driven part of the move would be about −$140 at a beta of 1.4 and −$70 at a beta of 0.7. That’s a rough scenario based on the historical relationship. The stock’s actual return can be very different.
What does beta mean? High beta vs low beta stocks
Use 1 as your reference point. That’s the beta the benchmark has against itself.
- Above 1 (high beta): bigger moves than the market, in the same direction.
- Exactly 1: in line with the market on average.
- Between 0 and 1 (low beta): same direction, smaller moves.
- Near 0: little link to the market. The stock can still swing hard for its own reasons.
- Below 0 (negative): the stock has tended to move against the market. That pattern may not hold in the next sell-off.
Industries differ a lot. In NYU professor Aswath Damodaran's industry beta data for U.S. companies, as of January 2026, semiconductor firms had a beta of 1.52 and water utilities 0.41. Apply those sector estimates to the same 10% market drop and the market-driven moves work out to about −15% and −4%, respectively. Individual companies can behave differently. Any beta comes from past returns against a named benchmark, so it drifts over time.
The beta formula
Beta divides how much the stock and the market move together by how much the market moves on its own. Statisticians call those two parts covariance and variance:
Beta = covariance of stock and market returns ÷ variance of market returns
You can also picture beta on a chart. Plot one dot per month, with the market's return along the bottom and the stock's return up the side, and draw the straight line that best fits the dots. Beta is the slope of that line, so if it rises 1.4 points for every 1 point to the right, beta is 1.4.
That line-fitting method is called regression. Whichever way you calculate beta, use percentage returns, not prices or dollar changes. A $1 rise is 2% for a $50 stock but only 1% for a $100 stock.
How to calculate beta: a worked example
You can calculate beta in three steps: multiply, square and divide. Here are five months of returns for a stock and a benchmark. They're made-up teaching numbers, chosen so each column averages zero, which saves a step.
Five months, one fitted line
The last two columns show what you need for the calculation.
| Month | Index | Stock | Index × stock | Index² |
|---|---|---|---|---|
| 1 | −2% | −1% | 2 | 4 |
| 2 | −1% | −4% | 4 | 1 |
| 3 | 0% | +1% | 0 | 0 |
| 4 | +1% | 0% | 0 | 1 |
| 5 | +2% | +4% | 8 | 4 |
| Total | 14 | 10 |
- Multiply. For each month, multiply the benchmark's return by the stock's return. In month 1, −2 × −1 = 2. When both move the same way, the result is positive; when they move in opposite directions, it's negative. The total is 14.
- Square. Multiply each benchmark return by itself, for example −2 × −2 = 4. Squaring measures how much the benchmark moved, whatever the direction. The total is 10.
- Divide. 14 ÷ 10 = 1.4.
So the stock's beta is 1.4. The fitted line rises 1.4 percentage points for every 1 percentage point increase in the benchmark.
With real data, use far more than five months, because one unusual month could swing the result. The averages won't be zero either, so before step 1, subtract each column's average from each month's return, then carry on as above. Textbook formulas also divide both totals by the same number (the number of months minus one), which cancels out, so you can skip it.
In the chart above, the dots don't all sit on the line. In month 4, the benchmark gained 1% and the stock returned 0%, while the line points to 1.4%. Beta describes the average relationship, and single months can miss it by a wide margin.
Calculate the same example in Excel
Put the months in column A, benchmark returns in B2:B6 and stock returns in C2:C6, entered as percentages, zeros included. Then type:
=SLOPE(C2:C6,B2:B6)
The result is 1.4. Put the stock column first, because SLOPE takes the dependent values (the stock) before the independent ones (the benchmark). Swapped, the same data gives 0.41. The same argument order works in Google Sheets.
Pair each stock return with the benchmark return for the same period. January’s stock return goes with January’s benchmark return.
check the datesBeta vs correlation vs volatility
All three come from the same returns, but each answers a different question:
- Beta: how far does the stock move per 1% benchmark move? In the example, 1.4.
- Correlation: how closely do the two move together? It runs from −1 to +1, and the example’s 0.76 shows a positive relationship, with room for the stock and benchmark to move apart. Beta has no fixed limits.
- Volatility: how widely do the stock's own returns swing? Measured by sample standard deviation, the stock's monthly volatility is about 2.92%, against 1.58% for the benchmark. Neither figure is annualized.
Using the unrounded figures, the stock’s volatility is about 1.84 times the benchmark’s. Its beta is 1.4. The gap comes from correlation, because part of the stock's movement has nothing to do with the market. The three connect like this:
Beta = correlation × (stock volatility ÷ benchmark volatility)
Here, 0.76 × 1.84 ≈ 1.4, using rounded figures.
Low beta leaves plenty of room for company-specific risk. An earnings miss or a debt problem can still send the stock down, even if the benchmark barely moves.
check company risk tooWhy do websites show different betas for the same stock?
Because each one makes its own choices when it calculates beta. Yahoo Finance, for example, labels its figure "Beta (5Y Monthly)", meaning five years of monthly returns. Before comparing two published numbers, check:
- Benchmark: which index does each estimate use? A London-listed stock may be measured against the FTSE 100 on one site and a global index on another.
- Lookback period: do both cover the same years?
- Return interval: are the returns daily, weekly or monthly?
- Return data: are dates and currencies aligned, and how are splits and dividends handled?
- Method: is the figure a raw historical beta or an adjusted one? Many services nudge the raw number toward 1.
Both numbers can be correct for their own settings. A longer history adds data but may include years before the company changed its business or took on more debt. Don't read much into small gaps such as 1.18 versus 1.22.
If you calculate beta yourself, use split-adjusted prices and the same return type for both series: price returns, or total returns with dividends.
How to calculate portfolio beta
Multiply each holding's beta by its share of the portfolio's value, then add the results. That sum is the portfolio's beta, a measure of its overall market exposure. The weighted average works as long as every beta uses the same benchmark and method.
Suppose a hypothetical $10,000 portfolio holds:
- Stock A: $6,000, with an assumed beta of 1.2.
- Stock B: $4,000, with an assumed beta of 0.7.
Same holdings, different weights
60% A / 40% B
Beta 1.00
About −$1,000
30% A / 70% B
Beta 0.85
About −$850
The calculation is (0.60 × 1.2) + (0.40 × 0.7) = 0.72 + 0.28 = 1.0.
A portfolio beta of 1.0 means about the same market sensitivity as the benchmark. The market-driven part of a 10% benchmark drop works out to about −$1,000. In the second scenario, a 30/70 split ($3,000 in A, $7,000 in B) gives (0.30 × 1.2) + (0.70 × 0.7) = 0.36 + 0.49 = 0.85. The same market drop then gives about −$850 for that part of the move.
Two stocks are still far less diversified than an index, because either company could suffer a large loss of its own. Our guide to building a diversified portfolio covers that side. To cut market exposure without selling, see hedging a portfolio with options.
Using beta when trading stocks or stock CFDs
Beta can help compare the market exposure of two hypothetical positions. Multiply the dollar value of each position by its beta. A $1,000 position at a beta of 1.0 and a $714 position at a beta of 1.4 both give about $1,000 of beta-adjusted exposure. For a 10% market drop, the estimate is about −$100 in either case. Any company-specific move can change the actual loss.
Beta won't give you an entry, a stop or a maximum loss, and a monthly beta measured over years says little about the next five-minute candle. For a short-term trade, check company news, the chart on your timeframe, liquidity and the next earnings date. Our guide to analyzing stocks for trading covers both the fundamental and the technical side.
With a stock CFD, use the full position value when comparing market exposure. The margin you put down is only part of that value. Leverage magnifies gains and losses relative to that margin, and financing charges and margin rules can affect the outcome. Our guide to calculating leverage shows how your exposure compares with the margin you put up.