Quick Question – looks like in the CFAI answer key in order to calculate weekly var from annual numbers for the return part of the number they are simply taking annual expected return / 52. Would it be ok to do (1 + annual return)^{(1/52)} -1 or is there something int the curriculum saying we should be using arithmetic returns for VAR calculations?

(1+r)^{(1/52)}-1 is mathematically more precise IMO. But people (and CFAI) use usually arithmetic mean method (which is an approximation method) in practice. So, you should use the arithmetic mean method (this method is easier to compute).

Return is taken the the root power (1+return)^{(1/52)}

Risk (volatility) is standard deviation / Sq rt (time period) i.e. Stand. Dev. / sq rt(52)

You can substitute the 52 for any timeframe. If they wanted one day volatility, what would your answer be?

So they would accept either? Bc the answer key is arithmetic.

To be accurate, it’s more _ **accurate** _, but not necessarily more *precise*.

(Yes: it’s been one of those days.)