Revised Risk Statistics

You can expect a more detailed explanation of the what’s and why’s to the following changes to come later this week.

When the term excess is used below, it means that, for a given period, T-bill is subtracted from return.
The changes affect Standard Deviation, Sharpe, Sortino, Correlation, R-Squared, Beta, and Alpha.
Previously, these statistics all used a daily period for returns. They have been changed to use a monthly period.
Standard Deviation and Sharpe now take sample standard deviation instead of population standard deviation.
Sharpe takes the summation of the excesses for the numerator instead of the excess of the summation of the thereof.
Sortino takes the sample semi-deviation for the denominator instead of excluding zeros to compute population standard deviation.
Correlation now takes excesses for both portfolio and benchmark.
Beta is unchanged besides its use of a monthly period.
Alpha has been revised to compute alpha for a monthly period and annualize it. Previously, it was using daily returns to compound the excesses for the whole period then annualizing that value.

Please see ‘Glossary of Risk Statistics’ in the ‘Help’ menu for details on the current methodology.

Aaron,

I see the new statistics already under risk in one of my sims.

You may already know this but the statistics under trade are broken: all zeros.

Edit: looks fixed already: Thanks!

Regards,

Jim

Interesting addition thank you. If I may suggest something, I think it would be helpful to have some small image on the side of each ratio with an i, or an ? sign, and if you mouse over, it would give you an explanation of how the ratio has been calculated.

something on the lines of:

Sharpe Ratio: This ratio measures the excess return of the portfolio in relation to its volatility. Formula: ( Compounded Annual Return - Risk Free rate) / Annualized Standard Deviation

  • Risk free rate used: X.XX%
  • Annualized Standard Deviation = Monthly StDEV * SQRT(12)

This could of course be applied to any other page with statistics/ratios so that it is easy to understand them.