Quant Analytics: from scratch to pro · Quant foundations: returns, risk and clean data · Beginner
Returns: simple, log and compounded
Every quantitative study starts by turning prices into returns. Simple returns add up across a portfolio; log returns add up across time. Knowing which one to use, and when they differ, prevents the first and most common error in market maths.
In this lesson
- Compute a simple return and a log return from two prices
- Chain returns over several periods correctly
- Choose simple returns to combine assets and log returns to combine periods
- Include dividends or interest to get a total return
Read first: Interest and compounding
Prices tell you where a market is. Returns tell you what happened to money held in it, and they are what every quantitative method works on: averages, volatility, correlation, tests and risk limits. There are two ways to write a return, and each has a job.
Two ways to measure the same move
A simple return is the change in price divided by the starting price: R = P_end / P_start - 1. A move from 100 to 110 is a simple return of +10%.
A log return (also called a continuously compounded return) is the natural logarithm of the price ratio: r = ln(P_end / P_start). The same move from 100 to 110 is a log return of 9.53%.
For small moves the two are almost equal. For large moves they part company, as the chart shows. A fall to half the price is a simple return of -50% but a log return of -69.3%; a doubling is +100% simple and +69.3% log. On the log scale, halving and doubling are mirror images.
Worked example
A share closes at 100, then 110, then 99.
Day 1 is a simple return of +10%. Day 2 goes from 110 to 99, a simple return of -10%. The average of the two simple returns is 0%, yet the share is below where it started.
Simple returns chain by multiplying: (1 + 0.10) x (1 - 0.10) - 1, which gives -1.0% over the two days. They do not add.
Log returns chain by adding. Day 1 is 9.53%, day 2 is -10.54%, and the sum is -1.005%. Turning that back into a simple return, exp(-0.01005) - 1, gives -1.0% again. Adding is easier than multiplying, which is why time-series models usually work in log returns.
Compounding works the same way over longer spans. A gain of 1% every month for a year is not 12%; it is 1.01^12 - 1, which is 12.68%.
Returns across assets: use simple returns
Over one period, a portfolio's simple return is the weighted average of its holdings' simple returns. Put 60% in an asset that gains 10% and 40% in one that loses 5%: the portfolio returns 4.0%.
Log returns do not average this way. Weighting the two log returns gives 3.67%, while the portfolio's true log return is ln(1.04), which is 3.92%. The error grows with the size of the moves and the spread between the holdings.
The rule: simple returns add across assets, log returns add across time. Convert between them with r = ln(1 + R) and R = exp(r) - 1.
Total return, not price return
A price series leaves out what the holder was paid. A share bought at 50 that ends the year at 52 has a price return of 4%. If it paid a dividend of 1 during the year, the total return is (52 + 1) / 50 - 1, which is 6%. For bonds the gap is usually larger, because interest is most of the return. Comparisons between assets, and between a trading rule and buy-and-hold, use total returns on both sides.
Common mistakes
- Averaging simple returns over time and calling it the growth rate. The 10% up, 10% down example averages 0% and loses money.
- Adding simple returns across days. They multiply.
- Weighting log returns to get a portfolio return. Use simple returns across assets.
- Comparing a price-only index with a total-return fund. The fund wins by the dividends alone.
- Computing returns across a missing day, a split or a contract roll without adjusting the prices first. The jump is not a return. See clean data.
In practice
Store prices, derive returns, and say which kind you used. Use log returns for time-series work (volatility, autocorrelation, models) and simple returns for portfolio arithmetic and for anything shown to a reader. Analytics on FinUties applies measures like these to the markets it covers, so they are at hand when you read a chart.
Simple return against log return for the same move
- simple return
- log return
Worked numbers
Each result is recomputed from the formula shown.
| What | Formula | Result |
|---|---|---|
| simple return from 100 to 110 | 110/100 - 1 | +10% |
| simple return from 110 to 99 | 99/110 - 1 | -10% |
| two-day simple return, chained | (1 + up) * (1 + down) - 1 | -1.0% |
| log return from 100 to 110 | ln(110/100) | 9.53% |
| log return from 110 to 99 | ln(99/110) | -10.54% |
| two-day log return as a sum | log_up + log_down | -1.005% |
| same total from the end prices | exp(log_total) - 1 | -1.0% |
| 1% a month compounded over 12 months | 1.01**12 - 1 | 12.68% |
| price return 50 to 52 | 52/50 - 1 | 4% |
| total return with a dividend of 1 | (52 + 1)/50 - 1 | 6% |
| portfolio simple return, 60/40 | 0.6*0.10 + 0.4*(-0.05) | 4.0% |
| weighting log returns instead | 0.6*ln(1.10) + 0.4*ln(0.95) | 3.67% |
| the portfolio's true log return | ln(1 + port) | 3.92% |
Check yourself
A price goes from 80 to 100. What are the simple and the log return?
Simple: 100 / 80 - 1 = 25%. Log: ln(100 / 80) = about 22.3%.
A share rises 10% one day and falls 10% the next. Is it back where it started?
No. 1.10 x 0.90 = 0.99, so it is down 1%. Averaging the two simple returns (0%) hides the loss.
You hold 60% in asset A (+10%) and 40% in asset B (-5%). What did the portfolio return?
4%. Simple returns combine across assets as a weighted average: 0.6 x 10% + 0.4 x (-5%). The same weighting of log returns does not give the portfolio's log return.
Why do log returns make a fall of 50% and a rise of 100% look the same size?
Because ln(0.5) = -69.3% and ln(2) = +69.3%. Log returns measure moves on a multiplicative scale, so halving and doubling are mirror images.
Key terms
Sources
- The Econometrics of Financial Markets. Princeton University Press (John Y. Campbell, Andrew W. Lo, A. Craig MacKinlay), 1997. Accessed 9 Nov 2026