Beta

How it's calculated

Beta is the slope of a regression line plotting an asset's returns against a benchmark's returns over the same periods. Mechanically:

Beta = Covariance(Asset, Benchmark) / Variance(Benchmark)

A daily beta is fitted from daily returns over some window (typically 3 to 5 years). Different windows and different benchmarks produce different betas — the same stock can have different betas vs. the S&P 500 and the Russell 2000, and different betas in 2020 vs. 2025.

What the value means

• Beta = 1.0 — moves with the market on average.
• Beta > 1.0 — amplifies market moves. A beta of 1.5 implies the stock typically moves 15% when the market moves 10%.
• Beta < 1.0 — dampens market moves. A beta of 0.5 implies the stock moves 5% when the market moves 10%.
• Beta < 0 — moves opposite to the market on average. Rare among regular stocks; gold mining stocks and some hedging strategies sometimes hit zero or slightly negative.

Tech stocks frequently have betas in the 1.2-1.8 range. Utilities, consumer staples, and healthcare often run at 0.5-0.9. Highly leveraged or distressed stocks can have betas above 2.

The CAPM and risk-adjusted return

Beta sits at the center of the Capital Asset Pricing Model (CAPM), which says expected return on an asset equals the risk-free rate plus beta times the equity risk premium. The model proposes that investors should only be compensated for systematic risk (beta) — idiosyncratic risk can be diversified away and shouldn't earn a premium.

CAPM has well-known empirical problems (high-beta stocks have historically earned less than CAPM predicts, low-beta stocks more), but the framework still anchors most institutional risk discussions and is the foundation for alpha calculations.

Limitations

Several reasons not to take beta at face value:

• It's backward-looking. A stock's beta is computed from historical returns; it can change as the company's business, leverage, or market conditions change.
• Window dependence. Different time windows produce meaningfully different betas. A growth stock might have beta 1.4 over the last 3 years and beta 1.0 over the last 10.
• Linear assumption. Beta assumes a linear relationship to the benchmark. Real returns have asymmetries (downside more than upside, "beta expansion" during crashes) that one number can't capture.
• Idiosyncratic events. Earnings surprises, M&A, regulatory shocks all produce returns that have nothing to do with the market — they show up as scatter around the regression line.

Crypto betas

Bitcoin tends to have a low beta vs. the S&P 500 (often 0.3 to 0.6 in normal periods). Altcoins tend to have very high beta vs. Bitcoin (often 1.5 to 3.0). This is the formal expression of the everyday observation that altcoins amplify Bitcoin's moves in both directions. The structural reason — altcoins are more speculative, less institutionally held, and more liquidity-constrained — produces consistent beta-amplification through cycles.