Log Versus Simple Returns in Reporting
Simple returns are what an investor actually experiences and should be used for reporting compounded performance, while log returns are additive across time and better suited to statistical modeling — mixing them up produces numbers that quietly don't mean what they look like they mean.
A simple return is the one an investor actually feels: start with $100, end with $110, that's a 10% return, full stop. A log return is the natural logarithm of the price ratio — for the same move, , or 9.53%. The two are close for small moves and diverge more the bigger the move is, and the reason both exist is that they're useful for different jobs.
Log returns are additive across time: the log return over ten days is just the sum of the ten daily log returns, which makes them convenient for statistical modeling and for combining return series computationally. Simple returns are the ones that actually compound the way an investor's account balance does, and multi-period simple returns must be multiplied, not added, to get a correct cumulative figure.
Adding log returns across periods gives the right cumulative log return, but reporting that number to an investor as if it were a compounded percentage return overstates losses and understates gains for anything beyond a single small period — always convert back with before reporting.
A common mistake is summing daily log returns over a year and quoting the sum directly as "annual return." It's close for calm markets but can be meaningfully wrong after a volatile year, which is exactly when accurate reporting matters most. The safe rule is to use log returns internally for any calculation, and convert to simple returns only at the final step before a number reaches a client or a performance report.
Related concepts
Practice in interviews
Further reading
- Bacon, 'Practical Portfolio Performance Measurement and Attribution'