How to code Simple returns for multiple columns?-CodePudding

How do I code this formula:

Simple returns = [(Pt / Pt-1) - 1]

I have tried the below, but keep getting the wrong numbers.

stockindices = read.csv('https://raw.githubusercontent.com/bandcar/Examples/main/stockInd.csv')

library(tidyverse)

simple_returns <- stockindices %>%  
  mutate(across(3:ncol(.), ~ ((.x / lag(.x-1))-1)))

CodePudding user response：

You had too many -1's in your expression:

simple_returns <- stockindices %>%  
    mutate(across( 3:ncol(.), ~ .x / lag(.x)-1))


str(simple_returns)
'data.frame':   3978 obs. of  8 variables:
 $ X   : int  1 2 3 4 5 6 7 8 9 10 ...
 $ Date: chr  "1999-04-01" "1999-05-01" "1999-06-01" "1999-07-01" ...
 $ DJX : num  NA 0.01382 0.025107 -0.000755 0.011068 ...
 $ SPX : num  NA 0.01358 0.02214 -0.00205 0.00422 ...
 $ HKX : num  NA 0.00835 0.03465 0.04493 0.00272 ...
 $ NKX : num  NA -0.01365 0.01781 0.00506 -0.01069 ...
 $ DAX : num  NA 0.000295 0.036108 -0.022119 0.01308 ...
 $ UKX : num  NA 0.0134 0.03199 -0.00774 0.00754 ...

You could have bracketed the .x/lag(.x) but it's not necessary here because of operator precedence and R's order of operations rules. The default lag-interval is 1 so it doesn't need to be inside the argument to lag. If you had wanted the semi-monthly returns it would have been

  ~ .x/lag(.x, 2) - 1

And as always it will pay to make sure that you have masked the stats::lag function, which is quite different and doesn't play nicely with the tidyverse.