# Sharpe ratio

> The Sharpe ratio measures how much return a portfolio produced for each unit of volatility it took on. Here is how Gylder calculates it and how to read it.

> 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).

The Sharpe ratio answers one question: how much return did this portfolio
produce for each unit of risk it took on?

Two portfolios can post the same annual return while behaving completely
differently along the way. One drifts upwards steadily. The other doubles,
halves, and doubles again. The Sharpe ratio is what separates them, because it
divides the return by how much the portfolio moved around to get there.

You will find it on the **Performance** tab of
[the investments dashboard](/guides/using-the-investments-dashboard), in the
grid of risk measures below the chart.

## How it is calculated

```
Sharpe ratio = (mean daily return − daily risk-free rate) / standard deviation of daily returns × √252
```

Gylder works from daily percentage changes in your portfolio's value. It
subtracts the daily risk-free rate from the average daily return, divides by
the [volatility](/metrics/volatility) of those returns, and annualises the
result with √252, the square root of the number of trading days in a year.

The risk-free rate is subtracted first because a return you could have had
without taking any risk is not compensation for taking risk.

## The risk-free rate is an assumption

Gylder currently uses a fixed **3% annual risk-free rate**. It is a
configuration constant, chosen to sit roughly in the euro short rate
environment. It is not pulled from a live market feed, and it does not change
with the window you select.

This is worth knowing for two reasons. A Sharpe ratio quoted anywhere else was
probably computed against a different assumption, so the two are not directly
comparable. And when your returns are close to 3%, the choice of rate drives a
large share of the result.

## How to read it

A higher number means more return per unit of volatility. That is the whole
interpretation, and it is worth resisting the urge to read more into it.

Gylder does not label a Sharpe ratio as good or bad, and does not compare yours
to a threshold. What counts as a reasonable figure depends on the asset class,
the period, and what you were trying to do, and any number we picked would be
an opinion presented as a fact.

The comparison that does hold is against yourself. The same portfolio measured
over the same length of period, before and after a change in strategy, is a
like-for-like reading.

## What it misses

The Sharpe ratio treats all volatility as risk, including the upward kind. A
portfolio that jumps sharply in your favour is penalised exactly as much as one
that falls just as sharply, which does not match how most people experience
risk.

That is the specific gap the [Sortino ratio](/metrics/sortino-ratio) exists to
close: it uses only downside deviation, so gains no longer count against you.
Both sit side by side in the risk grid for that reason.

The ratio also assumes returns are reasonably well behaved. Portfolios
concentrated in a few holdings, or holding assets that trade rarely, tend to
produce a figure that looks steadier than the position actually is.

For risk measured as the worst actual fall rather than the average spread, see
[maximum drawdown](/metrics/max-drawdown) and the
[Calmar ratio](/metrics/calmar-ratio).

## Period sensitivity

The Sharpe ratio is not stable across time windows. The same portfolio measured
over one year and over five will produce different figures, because both the
return and the volatility are calculated over whatever window you selected.

When you change the date range on the Performance chart, every measure in the
grid below it recalculates for that window. Comparing a one year Sharpe against
a five year one is not a like-for-like reading.