ROUNDING!!!!!!!!!!

[analystdude]

I recall 4 decimal places being the safe number for rounding. I see random problem sets where more/less decimal places will be included. For the exam stick to .0000 ?

 

[JeffO15]

I feel like for derivatives 5 is better. When using discount factors. Just for that section though

 

[chitown84]

I’ve noticed that too…. interested to hear everyone’s thoughts.
I would hope that CFAI is not going to punish anyone for taking the decimals too far out. I can’t recall which question it was, but I have seen a few on the schweser practice exams where going to 5/6 decimals materialy changed the calculation. Where did the 4 decimal rule of thumb come from (I honestly can’t remember?)

 

[analystdude]

chitown84 wrote:
I’ve noticed that too…. interested to hear everyone’s thoughts.
I would hope that CFAI is not going to punish anyone for taking the decimals too far out. I can’t recall which question it was, but I have seen a few on the schweser practice exams where going to 5/6 decimals materialy changed the calculation. Where did the 4 decimal rule of thumb come from (I honestly can’t remember?)

don’t quote me on this but i think i saw it in the schweser L1

 

[JayWill]

I tend to do 4 for most calculations, but I usually use 5 for anything that involves multiplying by a large notional principal.

 

[tau281290]

I usually just store them in the calculator so that’s 9 decimal places.