Aug

10

There are arguments for and against the log returns in data analysis. Any preferences and why?

Alex Castaldo writes: 

The nice thing about log returns is they are additive across time. The log return for the year is the sum of the log returns for Jan, Feb, Mar… Also the log return is to a first approximation normally distributed.

The nice thing about the percentage return is that the % portfolio return is a weighted average of the individual stock percentage returns, weighted by portfolio weights. So for example if you have 1/2 your money in MSFT and 1/2 in GOOG, the portfolio percent return will be the average of the percent returns on these two stocks.

So when I work with portfolios I always prefer the percentage returns. When I work with indexes I work either with percents or with logs (especially when concerned with options, since option theory is all in terms of log returns).


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search