Why is PP’s TTWROR changing when value of cashflow changes?

I was playing little bit with the performance metrics available in PP. I got an unexpected effect with the TTWROR.
I simulated an initial purchase the 01/01 of a stock and an additional purchase during the year. By changing the amount of the purchase, I was expecting to see unchanged TTWROR as it is the point with this metric to get rid of impact of cashflows and just have the performance of the stock, in contrary of the IRR. But for some reason, the TTWROR is changing when i changed the amount of new stocks i buy during the second purchase! I dont understand why. Any hint? is that expected effect? thanks.

  1. The varying amount of money left on the cash account from which you buy will have an effect, dampening the performance of the security.
  2. It is not possible to perfectly separate the effects of buying additional shares and of price changes during the same day.

Can you elaborate a little bit, i dont understand. Why the money on the cash account is impacting the internal performance of the stock or the amount of additional purchases? The TWR here should be (1+r1)*(1+r2) - 1, with r being my subperiod rates, right? During the second period, if i buy 1 or 10 new stocks, it is not impacting the performance of the security, that is the point with TWR metric. Does PP uses a variant of it or do i miss something here? Thanks!