What is Beta?
Reviewed by Karthikeyan Anbarasan (Founder, TradeLyser) · Updated 19 Oct 2026 · Editorial policy
How beta is calculated
- Collect returns for the same periods for the portfolio and the benchmark: daily, weekly or monthly, usually over one to three years.
- Work out the covariance between the two series: how far they move together around their averages.
- Divide it by the variance of the benchmark. In Figure 1: 7.43 ÷ 6.15 = 1.21.
- Check the result as the slope of the line through the scatter: on average, a 1% Nifty move came with a 1.21% portfolio move.
The same number can be written as beta = correlation × (portfolio volatility ÷ benchmark volatility). In Figure 1: 0.93 × (3.22% ÷ 2.48%) = 1.21. This shows why beta is not the same as correlation: a portfolio can move closely with Nifty (high correlation) yet swing more or less than it. A portfolio’s beta is the value-weighted average of the betas of what it holds: ₹6,00,000 in stocks with a beta of 1.4 and ₹4,00,000 in stocks with a beta of 0.8 gives 0.6 × 1.4 + 0.4 × 0.8 = 1.16. Cash has a beta of zero, so holding part of the account in cash lowers the overall figure.
Reading the number
Worked example: sizing a hedge
Illustrative: a ₹10,00,000 stock portfolio with the beta of 1.21 from Figure 1. Its market exposure behaves like ₹12,10,000 of Nifty. A 5% fall in Nifty would be expected to cost about 6.05%, or roughly ₹60,500. One Nifty futures lot at 25,000 with a lot size of 65 carries ₹16,25,000 of exposure.
Shorting one full lot would hedge about 134% of the exposure, so the hedge would make money in a fall and lose more than the portfolio gains in a rise. The lot size forces a choice between over-hedging and leaving part unhedged, which is worth writing into the plan before the need arises. Beta also matters for sizing without any hedge. If your risk rule is written for an index-like book, a portfolio with a beta of 1.21 will swing about a fifth more than the rule assumes, so a monthly loss limit set for a beta of 1 will be reached sooner. Recording the portfolio beta in the journal each month makes that visible.
Limits and common mistakes
- Beta looks backwards. It describes how the portfolio moved, and sensitivity can change sharply in a sell-off.
- Different windows and benchmarks give different betas. Daily returns over one year and monthly returns over three years rarely agree; always note which you used.
- Low correlation makes beta unstable. If the scatter is a cloud rather than a line, the slope means little.
- For an intraday F&O book, beta to the index’s daily close is close to meaningless; option positions are better described by their delta.
- Stock betas shift after big business changes, such as a merger, a large debt raise or a change in what the company sells. A beta measured before the change can describe a different company.
- Beta measures sensitivity, not total risk. A stock with a beta of 0.7 can still be very volatile if most of its moves come from company news rather than the market.
Reference guide
| Context | Value | Reading |
|---|---|---|
| Beta measured on 1 to 3 years of returns, benchmark stated | Usable estimate | Re-check yearly; betas drift as the business and the market change |
| Beta from a few weeks of data | Unreliable | A handful of points can give almost any slope |
| High beta read as “high quality” | Misread | Beta is sensitivity to the market, not a rating; it cuts both ways |
Questions
How is beta calculated?
Divide the covariance of the portfolio’s returns with the benchmark’s returns by the variance of the benchmark’s returns. It equals the slope of the line through a scatter of the two return series.
What is a good beta for a stock?
There is no good or bad beta, only the right one for your aims. Below 1 means smaller swings than Nifty, above 1 means larger swings in both directions.
What does a beta of 1.5 mean?
On average, the stock or portfolio moved 1.5% for every 1% move in the benchmark. A 4% Nifty fall would be expected to bring about a 6% fall.
Why do different websites show different betas for the same NSE stock?
They use different periods, return intervals (daily, weekly or monthly) and benchmarks (Nifty 50, Nifty 500 or Sensex). Each choice gives a slightly different slope.