# Beta

> Beta measures how much your portfolio tends to move when the benchmark moves. It is the first of the benchmark-relative measures, and the one the others build on.

> These figures are derived from a return series that the app currently flags as being reworked.
> Values and holdings are unaffected. See [about these numbers](/metrics/about-these-numbers).

Beta measures how sensitive your portfolio is to the benchmark you selected. A
beta of 1 means the two have historically moved together, one for one. A beta
of 0.6 means your portfolio typically moved about six tenths as far in the same
direction. A beta above 1 means it amplified the benchmark's moves.

This is the first of the benchmark-relative measures on the Performance tab,
and [alpha](/metrics/alpha) is built directly on top of it.

## How it is calculated

```
beta = covariance(portfolio daily returns, benchmark daily returns) / variance(benchmark daily returns)
```

Gylder resamples the benchmark onto your portfolio's own dates first, so the
two return streams are paired day for day. Without that step the calculation
would be comparing returns from different days.

## The benchmark matters

Beta has no meaning on its own. It is always beta _against something_, and
changing the benchmark changes the number.

You can select from the S&P 500, MSCI World, EURO STOXX 50, and STOXX Europe 600. A European portfolio measured against the S&P 500 will show a lower beta
than the same portfolio measured against STOXX Europe 600, and neither figure
is more correct than the other. They answer different questions.

When the benchmark has no usable price history for your window, the whole
benchmark-relative group shows a dash rather than a number. That usually means
its price cache is not warm yet, not that something is wrong with your
portfolio.

## How to read it

Beta describes a historical tendency, not a rule. It is an average relationship
fitted across every day in the window, so it can be badly wrong about any
particular day, and it is least reliable in exactly the conditions people most
want it for: sharp market moves, where correlations across holdings tend to
converge.

A low beta does not mean low risk. A portfolio can have almost no relationship
to the benchmark and still be extremely volatile on its own account. Beta only
measures the part of your movement that tracks the benchmark. For total
movement, see [volatility](/metrics/volatility); for the part that does not
track the benchmark, see [tracking error](/metrics/tracking-error).

Gylder does not suggest a target beta.