[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: calculating the calmar ration (for the mathematically inclined)



PureBytes Links

Trading Reference Links

What I do is take each month's gain or loss and divide it by some $ constant
to get a monthly percentage return.
This approximates a fixed fractional approach wherein you adjust position
sizes each month, ie if you trade 1 unit (contract in your case...doesn't
have to be...could be a portfolio of markets) for every $20K in account
equity, then take the monthly gain or loss and divide it by $20K to get the
month's ROR.

Scott Hoffman
Red Rock Capital Management, Inc.
www.RedRockCapMgmt.com

----- Original Message -----
From: "John Holton" <jgh1@xxxxxxxxxxxx>
To: <omega-list@xxxxxxxxxx>
Sent: Thursday, October 09, 2003 4:41 AM
Subject: calculating the calmar ration (for the mathematically inclined)


> I recently started using genetic algorithms to optimise systems.
Essentially
> the optimiser will try and find the best solution based on a fitness
> function. I've decided to use the calmar ratio as my fitness function. The
> calmar ratio is defined as annualised compound return / drawdown.
>
> Now my system trades a single futures contract for the entire period. I.e
it
> is not increasing the contract size as the the equity increases. So the
> monthly returns are dimishing in percentage terms as the capital
increases.
> So I need a way of calculating the compounded return.
>
> I calculate the annulised return using the following excel formula:
>
> annualised return =
>
> {=(product(monthly_returns+1)^12/no_of_months)-1)
>
> but here's the question. In order to calculate each monthly return I take
> the balance at the end of each month, subtract the starting balance and
> divide the result by the equity at the 'start of trading'. Is this
correct?
> (If I divide the result by the balance at the beginning of each month I
get
> an ever dimishing monthly percentage gain). Similarly I calculate the
> maximum drawdown by measuring the drop from the highest equity peak
against
> the starting capital. Is this correct?
>
> thanks
>
> john holton
>
>
>
>
>