Dear All,
As announced here we have been revising our risk statistic algorithms to use more industry standards. During this process we noticed that the following factors/functions needed to be revised:
Sharpe(range[,bars,offset])
Sharpe1Y
Sharpe2Y
Sortino(range[,bars,offset])
Sortino1Y
Sortino2Y
The following changes we made:
-
Sortino used semi-deviation correctly. Previously returns higher than 0 were being thrown out, decreasing the ‘n’ in the equation (see HELP->GLOSSARY of RISK STATS). This change will affect ranks based on Sortino and past simulations could change.
-
Sharpe & Sortino were annualizing both numerator & denominator. It is now done at the end. This change by itself should not affect ranking
THIS CHANGES WILL GO LIVE TONIGHT.
Very sorry for the inconvenience.