# How to get get weekly returns from daily data

Good day

I would like to get weekly returns data from daily data , I want to use the Wednesday-to-Wednesday approach – the returns (rt) are computed from the Wednesday closing prices Pt , i.e., rt = ln(Pt/Pt-1). In the cases in which Wednesdays were not active trading days, the closing values from the next date with valid prices from the sequence of the nearest days is used: I'm Stata user as well as an excel user

Any hint will be highly appreciated!! thaks!!

• consider the following YouTube video I made on how to change time intervals. You would just have to tweek it a bit to get desired Wednesday to Wednesday dates. youtu.be/SHVN-AyU0jk @ama
– Rime
Feb 8, 2015 at 13:20
• @Rim thank you Rim, It worked great for one company,my issue is that I have thousand of company and I have to repeat that action a thousand of time!!
– Ama
Feb 9, 2015 at 12:26
• Oh I see. I do not know Stata but I believe there are methods to do this using R-code for various iterations @Ama
– Rime
Feb 9, 2015 at 20:56
• Ama, Is your issue getting a data set with all the symbols/history you require? If you are familiar with PQ/PP, I'll work up an Excel solution for you. Let me know. Thanks. Jason, A simple but elegant approach. Thanks. Jun 10, 2016 at 9:45
• @Nin sute, thank you very much for your help yes sure I would be very grateful if you helped me out with that!!
– Ama
Jun 13, 2016 at 10:40

## 1 Answer

You can compute daily gross returns and then simply multiply them: $R_w=\prod_i \left(1+R_i \right)-1$, where $R_i=P_{close,i}/P_{close,i-1}-1$

• If you do some analytics you will figure out that $R_w=P_{close,the-last-day-of-the-week}/P_{close,the-last-day-before-the-week}$ So, there are much less computations Jun 13, 2016 at 19:37