# Excess Daily Returns to Excess Quarterly Returns

I am building a model which predicts the Excess Daily Returns over a time period. How do I convert these excess daily returns to excess quarterly returns? Should I just do an average of all the daily returns in that quarter, or is there another method to do this?

Thanks!

• Transform simple percentage returns into log-returns. They are time additive, so you can just aggregate your daily returns to lower frequencies (monthly, quarterly etc). $r_d=\frac{P_t-P_{t-1}}{P_{t-1}}=e^{r_c}-1$. – Kevin Aug 3 '20 at 23:05
• Ah, got it, that seems pretty neat. Thanks! – jitmanchan Aug 3 '20 at 23:59