with your .033(sp)/6937(S) numbers to get the rate per second, I notice you use simple interest. For those who might gripe, complain or just want to see what compound interest would do:
If your base amount was 8,023.485 sp. in basically your first payout of .033, there was a return of about .0004112926...% or a flat interest= .000004112926....
Assuming that there are 31536000 seconds in a year, and each payout cycle is 6937 seconds there are about 4546 cycles per year. ((1+interest)^4546-1)*100% gives about a 1.887321...% return each year. So it really doesn't matter much which of the two methods was chosen.
Thank you @firstamendment!
In the beginning of that calculation I mention that I ignore compounding interests. Even though the interest is referred to as over a year, we are applying it here for just a day. If you want to make it precise for periods of months and years, you must account for A LOT of variables. I've done that too in a running excel-sheet ;)
First you would have to do some assumptions about the calculation and payout cycles, like you did. I can tell you it's shorter than the 6,937. I can refresh in between every 0.001. You would also have to account for the continuously decreasing dilution/interest per tick.
An in the end the effect is neglectable. Like you showed yourself. when you add in the changes in STEEM value and rewards, the effect of compounding interest becomes practically impossible to predict
That’s why I decided to skipped it. It was not worth adding to the complexity for something that neglectable in ym mind.
I appreciate that you bring it up, so we can point the size of its effect though. I'm sure there were more than yourself thinking of the compounding interest.