Calculation of Sharpe's ratio requires keeping the history of what's called the "reference benchmark" which serves as a "risk-free" return, against which the system performance is compared against. Such "risk-free" asset could be a 3-month US treasury bill, for example. These days, the return on the risk-free assets is so low (very close to 0%), that Sharpe's ratio could be reduced to just this: Sharpe = R / Sigma, where R is the return, and Sigma is the standard deviation of returns.
I am not planning to add Sharpe's ratio to JBookTrader. The PI (performance Index) has a shape very similar to Sharpe's: PI = sqrt(trades) * R / Sigma, where "trades" is the number of trades, R is the average profit per trade, and Sigma is the standard deviation of trades. What's important is that in JBT, the performance benchmarks (including PI) are used not for assessing the absolute performance, but rather a relative performance of strategies, as compared to each other during optimization. On Fri, Mar 8, 2013 at 4:09 PM, Simba DIARRA <[email protected]> wrote: > Hi Eugene, > > I have read some history of emails back in 2010(?) btw you and other guys > about Sharpe ratio etc. > And I believe you rather made the choice to keep your PI instead. But > still you mentionned that Sharpe ratio is close to PI. > How would one correctly still calculate Sharpe ratio of JBT strategy > should we want to do so? or convert PI to Sharpe ratio ? > Sorry if my questions don't make eventually much sense. But Sharpe ratio > is very common to be used for strategy evaluation :-) > > Cheers > > > > > > On Thursday, 7 March 2013 04:31:19 UTC+9, Eugene Kononov wrote: >> >> >>> I suppose, with the numbers you are referring to the values achieved on >>> the training set? >>> Or do you include test data (data not seen in optimization). >>> >>> >> Yes, I am referring to PI numbers against the training set. How to >> accomplish the same performance against the out-of-sample data is another >> conversation. I was just giving guidance on the PI numbers. PI is closely >> related to Sharpe's ratio and the SQN (System Quality Number): >> http://www.traderivar.com/**2010/03/sqn-trading-system-** >> quality-number.html<http://www.traderivar.com/2010/03/sqn-trading-system-quality-number.html> >> >> > -- > You received this message because you are subscribed to the Google Groups > "JBookTrader" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/jbooktrader?hl=en. > For more options, visit https://groups.google.com/groups/opt_out. > > > -- You received this message because you are subscribed to the Google Groups "JBookTrader" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/jbooktrader?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
