How to compute annualized Sharpe Ratio from hourly PnLs?

543 Views Asked by At

I have an hourly time series $\mathrm{pnl}_1, \ldots, \mathrm{pnl}_N$ of profit & loss values (of some trading strategy), spanning a period of only a few days. I would now like to compute an (annualized) Sharpe Ratio from these values. It does not seem sensible to me to compute daily PnL returns and to compute an annualized Sharpe ratio from that since my data only spans a handful of days. How do people typically compute a Sharpe ratio in these cases?

1

There are 1 best solutions below

0
On

In case you believe your returns don't deviate too far away from being IID, it is suitable to take the hourly PnL time series and multiply it by $\sqrt{q}$, where $q$ is the number of trading hours in a day times the number of trading days in a year.

This assumption can be fairly shaky, in which case I recommend consulting the section on temporal aggregation of Sharpe ratios in this paper: The Statistics of Sharpe Ratios.