Average return vs annualized return

Average return vs annualized return: what is the difference?

The average return (also called the arithmetic mean) adds up each year’s return and divides by the number of years. The annualized return (also called the geometric mean, or the compound annual growth rate) is the single steady rate that would have turned your starting money into your ending money. Only the second one tells you what your money did.

The clearest way to see it is an extreme example.

The 50% loss came out of a bigger balance than the 50% gain went into. Losses and gains of the same size do not cancel, so whenever returns go up and down, a simple average flatters the result.

The two numbers go by several names, which is half the confusion:

• Simple average: arithmetic mean, mean return, average yearly return.
• Annualized return: geometric mean, compound annual growth rate (CAGR), compounded return, and in US fund documents, average annual total return.

How to calculate average and annualized return

If you know only the starting and ending balance, the annualized return is (end ÷ start)^1/n − 1, which is the same number. The rate of return guide covers that version.

Average vs annualized return calculator

Type yearly returns in order, or pick an example. The last two buttons load the S&P 500’s actual yearly returns for 2016–2025 and for 2000–2009.

Why the annualized return is lower: volatility drag

The gap between the two numbers is called volatility drag: the cost of returns bouncing around instead of arriving smoothly. The bigger the swings, the bigger the drag.

You can see it with matched pairs of years. A 10% gain and a 10% loss leave you with 1.10 × 0.90 = 0.99 of your money, down 1%. A 20% gain and a 20% loss leave 1.20 × 0.80 = 0.96, down 4%. A 50% gain and a 50% loss leave 0.75, down 25%. The average of each pair is 0%, and the loss grows with the size of the swing. Mathematicians call the general version the AM–GM inequality: the geometric mean of positive numbers is never above their arithmetic mean, and the two are equal only when all the numbers are the same.

A good rule of thumb links the size of the swings to the size of the gap:

Here is how much drag different amounts of swing cost, starting from a 10% average:

The shortcut is less exact for very large swings. For +50%/−50% it predicts −12.5% against the true −13.40%. For everyday stock returns it is close.

What is the average stock market return? It depends which average

The S&P 500, a stock index of 500 large US companies, is the usual stand-in for “the stock market.” Its average return depends on which average you take, over which years, and whether dividends and inflation are counted. Here are both averages over the longest stretch of good data.

Over the 98 years from 1928 to 2025, the S&P 500’s yearly returns, dividends included, had a simple average of 11.86%. The rate the money actually compounded at was 10.02%. 26 of those years lost money.

That 1.84-point gap looks small, but compounding blows it up. $10,000 invested at the start of 1928 grew to about $116 million by the end of 2025. Growing at the simple average every year would have produced about $587 million: about 5 times as much, all of it money that never existed. Over a plain 30 years, $10,000 at 11.86% becomes about $288,000; at 10.02%, about $175,000.

The same mistake hits savers who add money every month. Someone putting in $500 a month for 30 years who plugs the 11.86% average into a compound interest calculator is told to expect about $1,692,000. At the 10.02% the market actually compounded at, the same deposits come to about $1,135,000. That is a $558,000 difference, made up entirely by using the wrong kind of average. (Both figures are before inflation, fees and taxes, and neither is a forecast.)

The gap and the returns themselves depend heavily on which years you pick:

Look at 2000–2009: the average was a positive 1.16%, yet $10,000 shrank to $9,086, a compounded −0.95% a year. Anyone told that the market “averaged a gain” over that decade still lost money. Returns vary a lot, and the future can be worse than any of these windows.

Averaging returns in Excel and Google Sheets

Put the yearly returns in B2:B11 as percentages. Then:

• Average: =AVERAGE(B2:B11). For 2016–2025 that returns 15.75%.
• Annualized: add a helper column C with =1+B2 copied down, then =GEOMEAN(C2:C11)-1. That returns 14.68%.
• The same without GEOMEAN: =PRODUCT(C2:C11)^(1/10)-1. The product, 3.9354, is also the growth multiple: $1 became about $3.94.
• Volatility: =STDEV.P(B2:B11) returns 14.80%, the σ in the shortcut above.

If you only have a starting and ending balance, skip all of this and use =RRI(years, start, end).

Does the order of returns matter?

Not for a lump sum you leave alone. $10,000 that goes −50% then +50% ends at $7,500, exactly like +50% then −50%; multiplication does not care about order.

It matters a great deal once money moves in or out. A bad year right after you add a lot of money, or right after you start withdrawing in retirement, hurts more than the same bad year at another time. That is why your own return can differ from your fund’s even when you own the same thing.

Why you often see the bigger number

Because the simple average is always at least as high as the compounded return, and the word “average” does not say which one you are looking at.

Some numbers are pinned down. In a US mutual fund prospectus, the “average annual total return” for 1, 5 and 10 years is calculated as a compounded rate, from a hypothetical $1,000 held through the whole period with distributions reinvested, using P(1 + T)ⁿ = ERV (SEC Form N-1A, Item 26). That is the annualized return.

Plenty of other numbers are not bound by that formula: a sales presentation, a newsletter, a retirement-income pitch or a claim that “the market averages 12%.” A figure near 12% is the simple average of yearly S&P 500 returns; the compounded figure is about 10%, before inflation, fees and taxes. Other ways a number gets bigger:

• Picking the window. 2016–2025 compounded at 14.68%; 2000–2009 at −0.95%. Same index.
• Leaving out fees. Index returns have no fees; a real account does.
• Annualizing a short hot streak. A few good months, stretched to a yearly rate, looks spectacular.

Common mistakes when averaging returns

• Adding yearly returns and dividing. It gives the arithmetic average, which overstates growth whenever returns vary. For anything that compounds, multiply the growth factors instead.
• Comparing a simple average with a compounded one. A fund quoting its compounded return will look worse than a pitch quoting a simple average, even if the fund did better. Put both on the same basis first.
• Averaging the returns of different-sized holdings. A 12% return on $6,000 and a 4% return on $3,000 do not average to 8%. Weight each return by the money in it.
• Mixing before-inflation and after-inflation figures. The S&P 500 numbers here are before inflation. Set one against an after-inflation (real) figure and the before-inflation one looks several points better than it is.
• Annualizing a short stretch. One great quarter turned into a yearly rate is not a track record.
• Using either average when you added money. Both the simple and the compounded average describe one lump sum. With regular deposits, your own result depends on timing, which needs a money-weighted return.

Which return should you use for planning?

Use the annualized (compounded) return, and a cautious one. Any time you project a balance forward several years, the money compounds, so the compounding rate is the right one. Plugging a simple average into a compound interest calculator overstates the result, as the examples above show.

The simple average has one fair use: as a guess at a single typical year. It is never the rate your money grows at over many years.

For a stock-heavy portfolio, Bill Router uses 7% a year before inflation, or about 4% after 3% inflation, as a cautious planning figure. That is below the S&P 500’s historical compounded 10.02% on purpose: it leaves room for fees, for a mix that isn’t all stocks, and for a future that may be worse than the past. It is not a promise. You can test your own plan at that rate in the Freedom calculator.