Hi,
Maybe my query has a very simple and straight forward answer, but I cant find the logic of the multi period calculation of TWRR. Should we take in account the number of days of each period? I suppose is not the same to earn 2% in 11 days than 2% in 5 days for example but I find exercises when we should equally compounding all of the periods, even when they have different length.
Could someone explain me why we should do it in this way?
Thank you in advance
Juan
The whole point of the TWRR calculation is to remove the effects of external cash flows, into and out of the portfolio, and their effects on portfolio return. Investment managers rarely have control over the timing of external cash flows, which is why GIPS requires the use of TWRR. Obviously, if a large cash flow comes in just before a period of strong performance, it can bias the total return higher even though the investment manager had nothing to do with it.
So you determine the length of the period being evaluated, then separate the period by each cash flow and create a sub-period from the start of the sub-period to the date of that cash flow. You then calculate the return for each sub-period, removing the cash flow from the return calculation, then geometrically link those sub-period returns for a total return for the entire period.
Thank you Jaywill but my question is why we geometrically link subperiods with different length?
As per my example above the return earned in teh period of 11 days (2%) will have the same weight than the return earned in 5 days (2%) and will be 1.02*1.02 without consider the length of the period. I do not understand that.
If we subdivide then in other subperiods 8 days each for example, maybe we arrive to different results
Thanks
S
Because the returns are already adjusted for their respective days, i.e. the returns have not been annualized. As you are doing the geometric product of the returns, you are simultaneously adding up the no. of days as well…