Calculating Performance (%)

[Aaron]

Quote from Babak:

11-Apr-01 -$100 starting capital
15-May-01 $55 trading profit
15-Jun-01 -$75 withdrawal
15-Jul-01 $5 trading profit
15-Aug-01 -$75 trading losses
15-Sep-01 $35 addition to acct
15-Oct-01 $20 addition to acct
15-Nov-01 -$12 withdrawal
15-Dec-01 $15 trading profit
31-Dec-02 $120 ending capital

You can use the XIRR function to get the answer, but you aren't going to be able to have confidence in your answer unless you confirm the result by calculating the return longhand. Here's how to calculate the return...

You need to calculate the return for each subperiod between additions and withdrawals and then compound all these subperiod returns to get your return for the full period. Here's the calculation...

You started with $100 and made $55 in the first subperiod. Return=55%. After withdrawing $75, you started the second subperiod with $80. You lost $70 in the second subperiod for an 87.5% loss. Then, to your $10 account, you added $55 and withdrew $12 --> you start the third subperiod with $53. You made $15 or 28.3% and should have ended with $68.

Since these are made up numbers we won't try to reconcile the sum of the cash flows with your ending capital. (As Biog mentioned, $68 doesn't equal $120.)

You can now compound your subperiod returns to get your period return... (1+0.550) x (1-0.875) x (1+0.283) -1 = -75.1% for the period from 4/11/01 through 12/31/02.

Fortunately this result confirms Biog's NAV calculation.

 

[Babak]

Aaron,

thanks for the explanation. Phew! That can be rather cumbersome if you have a lot of add/withdrawals!! I think XIRR does provide a short cut.

For the numbers we are discussing (with the correction to ending balance of $68 instead of my original $120), the XIRR function returns -32.37%. At first, this may seem to disagree with the -75.1% number you and Biog found.

But this is an annualized average return (I assume geometric and not arithmetic) while your calculation as well as Biog's is for the whole period covering the data mentioned (April'01 to Dec'02). So I think the two (or three, considering Biog's NAV) methods do provide the same answer.

Now, if only I could remember how to flip between annualized average returns and total returns so that I can state the validity of XIRR with no qualms....

Anyone remember the way to do that?

 

[NinjaTrader_Dierk]

Quote from Babak:

Aaron,

thanks for the explanation. Phew! That can be rather cumbersome if you have a lot of add/withdrawals!! I think XIRR does provide a short cut.

For the numbers we are discussing (with the correction to ending balance of $68 instead of my original $120), the XIRR function returns -32.37%. At first, this may seem to disagree with the -75.1% number you and Biog found.

But this is an annualized average return (I assume geometric and not arithmetic) while your calculation as well as Biog's is for the whole period covering the data mentioned (April'01 to Dec'02). So I think the two (or three, considering Biog's NAV) methods do provide the same answer.

Now, if only I could remember how to flip between annualized average returns and total returns so that I can state the validity of XIRR with no qualms....

Anyone remember the way to do that?

Quite a bit late, but anyhow ...

I would calculate annual profit a follows:
2002-12-31 - 2001-04-11 = 629 days
daily return = power(1- 0,751; 1 / 629) -1 = -0,0022 (-0,22%)
annual return = power(1-0,0022; 365) - 1 = -0,5537 (-55,37%)

Still, there is a gap to -32,37%

Hope this helps

Dierk