A trader said this to me on a call recently: “I’m doing 10, 15 percent, but that’s purely because the market is really good. It’s not my skill, it’s just luck.”
He was being harder on himself than he needed to be. But he was asking exactly the right question, and it is the question almost nobody asks about their own returns. Not “what number did I make?” but “what does that number actually prove?”
CAGR is the headline figure on every backtest report, every fund factsheet, and every trading course sales page. It is also the number traders most often fool themselves with. Read on its own, it can make a dangerous system look excellent and a good system look dull.
This guide covers what CAGR is, how to calculate it, what a realistic CAGR looks like for a private systematic trader, and the numbers you have to read alongside it before you risk a cent.
What Is CAGR?
CAGR stands for Compound Annual Growth Rate. It is the single annual rate of return that would have taken your account from its starting value to its ending value, if it had grown at exactly that rate every year with profits reinvested.
In other words, it smooths a bumpy equity curve into one clean number: this account compounded at X% a year.
For a systematic trader, CAGR is a portfolio-level statistic. It describes what happened to your capital over the whole test period. It does not describe any individual trade, any individual year, or the path your account took to get there. That last point is where most of the trouble starts.
How Do You Calculate CAGR?
The formula is:
CAGR = [(Ending Value / Beginning Value) ^ (1 / Number of Years)] – 1
Three inputs: what you started with, what you finished with, and how many years passed. That is all.
Worked example. You start a system with $100,000. Ten years later the account is worth $340,000.
- Ending Value / Beginning Value = 340,000 / 100,000 = 3.4
- 3.4 ^ (1/10) = 1.1315
- 1.1315 – 1 = 0.1315, or 13.15% CAGR
In a spreadsheet: =(340000/100000)^(1/10)-1, formatted as a percentage.
Two things to watch. First, the number of years has to be the actual elapsed time, not the number of calendar years your data touches. A test running from June 2010 to December 2024 is 14.5 years, not 15, and rounding it up quietly flatters the result. Second, if you deposited or withdrew money during the period, plain CAGR will be wrong, because it credits your deposits as trading profit. Backtests do not have this problem. Live accounts do.
What Is The Difference Between CAGR, Total Return And Average Annual Return?
They answer three different questions, and only one of them tells you the honest growth rate.
Total return is the overall gain from start to finish. Turn $100,000 into $200,000 over five years and your total return is 100%. It tells you the size of the result but nothing about the rate, and nothing about whether it took five years or fifteen.
Average annual return is each year’s percentage return added up and divided by the number of years. This is the one that misleads traders, and it does it in the direction that flatters them.
Take two years:
- Year 1: +50%
- Year 2: -50%
The simple average is 0%. Sounds like you broke even. Run the actual dollars:
- Start: $100,000
- After Year 1: $150,000
- After Year 2: $75,000
You lost a quarter of your capital while the average told you that you were flat. CAGR on those same two years is [(75,000 / 100,000) ^ (1/2)] – 1 = -13.4% a year, which is the truth.
The gap between the two grows with volatility. A system that swings between big up years and big down years will always have an average annual return that looks better than its CAGR, and the more violent the swings, the bigger the difference. When you are comparing systems, use CAGR. When someone quotes you an average annual return, ask what the compound figure is.
What Is A Good CAGR For A Trading System?
For a single well-designed system traded by a private trader, 12% to 25% a year is the realistic range. Here is how I read the numbers when a student brings me a backtest:
| Backtest CAGR | How I read it |
|---|---|
| Below 10% | On its own, rarely worth the effort. It can still earn a place in a portfolio, as you’ll see below. |
| 12% to 25% | The realistic target zone for a well-built system. |
| Above 30%, consistently | Start looking hard for curve fitting or hidden risk. It is possible, but it is not the base case. |
| 50%+ a year | Almost certainly unsustainable for a private trader without taking catastrophic risk somewhere. |
That table annoys people. Traders arrive expecting to hear triple digits, because that is what the internet promises them.
But here is the part most traders miss. The CAGR of one system in isolation is not that useful a number.
Systematic traders don’t trade one system. We trade a portfolio of them: different markets, different timeframes, long and short, trend following alongside mean reversion. Then we use capital allocation to decide how much of the account each system gets. The result that matters is the combined portfolio, and the risk-adjusted return it delivers.
That changes how you judge a backtest. A 12% system that makes money while your other systems are struggling can be worth more to you than a 25% system that loses money at the same time as everything else you trade. Systems that don’t move together smooth out the equity curve. A smoother curve means shallower drawdowns. Shallower drawdowns let you size up and still stay inside the risk you are willing to carry. That is how a portfolio of modest systems can beat any one of its parts.
It is also why a system compounding at 20% with a drawdown you can live through will build far more wealth over a decade than a 60% system you abandon in year two. The best system in the world is worthless if you will not follow it.
So when you look at a system’s CAGR, the better question is not “is this number high enough?” It is “what does this system add to my portfolio?” And a number in the right range still tells you almost nothing on its own, because you have not yet asked what it cost.
Why Is CAGR Alone A Misleading Number?
Because it answers one question and stays silent on four others that matter just as much.
1. It ignores drawdown completely. A system making 30% a year sounds excellent. If it produces a 60% drawdown on the way, most traders will abandon it in a panic shortly before the recovery. The return figure tells you nothing about the pain required to collect it. Two systems can post identical CAGRs while one is comfortable to trade and the other is unbearable. Drawdown is the number that decides whether you are still trading the system when it pays.
2. Compounding distorts long backtests. Over a 25 or 30 year test, a small difference in annual rate becomes an enormous difference in final dollar value. That makes side-by-side CAGR comparisons between systems misleading, because the number is being amplified by the length of the test rather than the quality of the edge. It also lets the last few years of a long backtest dominate the result, since that is where the largest dollars are.
3. It ignores how much capital was actually at risk. Two systems both show 30% CAGR. One was fully invested the entire period. The other was in the market half the time and in cash for the rest. On CAGR they look identical. They are nothing alike. If the market gaps down 20% overnight, the fully invested system takes the full hit and the half-invested one takes half. The second system produces the same return with far less exposure, which makes it the better system and the better building block for a portfolio of trading systems.
4. It says nothing about the path. Two equity curves can start and finish at the same points. One climbs steadily. The other does nothing for six years and then triples in eighteen months. Identical CAGR, completely different experience, and completely different odds that you are still there at the end.
What Is The MAR Ratio And Why Does It Matter More Than CAGR?
The MAR ratio is CAGR divided by maximum drawdown. It is my core metric, because it forces the return number to account for the pain that produced it.
MAR = CAGR / Maximum Drawdown
A 25% CAGR with a 50% drawdown gives a MAR of 0.5. A 20% CAGR with a 15% drawdown gives a MAR of 1.33. The second system has the lower headline number and is far and away the better system: more efficient with your capital, and realistic to hold onto when it goes through a bad patch.
Targets I work to:
- Long-only trend following system: aim for a MAR of 0.75 or better. A system returning 20% a year with a 27% maximum drawdown sits at about 0.75.
- Below 0.75: the system can still be valuable to your portfolio. If it is negatively correlated with your other systems, it does not have to be very profitable on its own to add value to the portfolio as a whole. Be careful not to discard a system based purely on its return and drawdown. You also need to weigh the diversification it brings.
- A portfolio of genuinely uncorrelated systems across different markets: target 1.5 or higher. Combining systems whose bad periods do not coincide lifts portfolio MAR above what any single system can reach, because the portfolio drawdown ends up smaller than the sum of the parts.
You will also see the Calmar ratio, which is the same basic idea, conventionally calculated over a rolling 36-month window rather than the full history. Both answer the same question: how much return am I getting per unit of drawdown? Use whichever your platform reports, and be consistent about it.
If you want to work out what a given drawdown does to an account and what it takes to recover, run the numbers through the free drawdown calculator. The recovery maths is worse than most traders expect, and seeing it in dollars changes how people think about the CAGR figure sitting next to it.
Why Will Your Backtest CAGR Not Repeat In Live Trading?
Because a backtest is a best-case indication of how the strategy performed in the past. Several things systematically inflate the number:
Curve fitting. Tune the parameters until the equity curve looks beautiful and you have memorised the past rather than found an edge. Those parameters were fitted to historical noise, and noise does not repeat. This is the single largest cause of the gap between backtest and live results, and it is covered in depth in the guide to trading system optimisation.
Survivorship bias. Most cheap historical data only contains companies that still exist. The ones that went bankrupt or were delisted have quietly vanished from your universe, and with them the worst losses any real trader of that period actually took. Your backtest never buys the company that went to zero. You will. Survivorship bias is why the data vendor you choose matters as much as the rules you write, and why I use Norgate Data for point-in-time index membership and delisted securities.
Underestimated slippage and commission. In a backtest you fill at the exact price you asked for. In live trading you do not, particularly in smaller or less liquid stocks, and particularly as your position sizes grow. Model slippage honestly and watch what it does to a high-turnover system’s CAGR. Plenty of short-term systems that look profitable on paper do not survive realistic costs.
Optimistic execution timing. Backtests often assume you act on the close, or at a price that assumes you were watching at the right moment. Real orders go in the next morning at whatever the market gives you.
Two rules I apply to every backtest as a result:
- Plan for roughly 1.5 times the backtested maximum drawdown. If the test shows 20%, build your position sizing and your expectations around 30% actually happening at some point. Every system eventually posts a new maximum drawdown, because the backtest only ever sampled a limited set of market conditions.
- Expect live CAGR to run below backtest CAGR, especially in the early stages while you are still learning the system’s behaviour and your own reactions to it.
If you want a step-by-step process for building a test you can actually trust, start with the guide to backtesting.
How Should You Set Your Own CAGR Target?
Backwards from the drawdown you can survive, not forwards from the return you would like.
Almost every trader does this in the wrong order. They pick a return target first, usually an ambitious one, then build or buy a system that produces it, and only afterwards discover what drawdown was attached. By the time they find out, it is with real money.
Do it the other way around:
- Decide your maximum tolerable drawdown first. Be honest, and be conservative. Whatever percentage you think you could handle, you will handle less of it in live money than you imagine in a spreadsheet. One trader put it perfectly on a call: “I think that’s my personality, is just that I really hate drawdowns.” That is useful self-knowledge, and it should shape the system you build.
- Set your CAGR target at no less than three-quarters of that number, so your MAR is at least 0.75. Comfortable with a 20% drawdown? Then you need at least 15% CAGR, and ideally 20% or more.
- Remember the 1.5x rule. A 20% live tolerance means designing against a backtested drawdown closer to 13%, because live will be worse than the test.
- If the maths does not work, change the system, not the target. Squeezing a higher CAGR out of a system by adding rules and tuning parameters is how curve fitting happens. Adding a second, genuinely different system across another market or another profit driver is how you actually improve the portfolio’s numbers.
This is the same order of operations I use when setting trading goals and drawdown targets with students, and it is the difference between a plan you can hold and one you abandon at the worst moment.
Which Numbers Should You Read Alongside CAGR?
Never evaluate a system on a single statistic. These are the ones I read together, in order:
| Metric | What it tells you |
|---|---|
| Maximum drawdown | The worst peak-to-trough loss in the test. Check this before the return. If you cannot survive it psychologically in real money, the CAGR is irrelevant because you will quit before you collect it. |
| MAR ratio | CAGR divided by maximum drawdown. Return per unit of pain. Target 0.75+ for a single long-only system, 1.5+ for a portfolio. |
| Exposure and return per unit of exposure | How much of the time your capital was actually deployed. A system making a modest return while invested only a small fraction of the time is highly efficient and combines well with others. |
| Ulcer Index | Accounts for the depth and the duration of drawdowns, not just the single worst dip. A shallow drawdown that lasts three years hurts more than a sharp one that recovers in three months. |
| Expectancy and trade statistics | Percentage winners, average win, average loss, number of trades. These confirm the edge is real and stable rather than the product of a handful of lucky outliers. |
| Sharpe ratio | Risk-adjusted return measured against volatility. Useful, with real limitations for traders, which that guide covers in detail. |
A backtest with 30 trades and a beautiful CAGR is not evidence. A backtest with several hundred trades across multiple market regimes, a MAR of 0.75 or better, and a drawdown you have already decided you can live through is the beginning of something tradeable. For a full walkthrough of reading a system report properly, see measuring the performance of your trading systems.
The Honest Version Of The Question
Stop asking “is the CAGR high enough?” and start asking “is this return worth the risk I am carrying, and can I actually stay with this system through the rough periods?”
A stable 15% to 20% CAGR with a manageable drawdown and a strong MAR ratio is an excellent result. It compounds into serious wealth over a decade or two, it lets you sleep, and it survives contact with real markets. Chasing a bigger number without checking what it cost in drawdown and exposure is the fastest route to either blowing up or quitting a perfectly good system at precisely the wrong moment.
The trader I quoted at the start was right to be suspicious of his own returns. Suspicion of a single flattering number is the correct instinct. The fix is a fuller set of numbers, and a system built around the drawdown you can survive rather than the return you would like to report.
Frequently Asked Questions About CAGR
What does a 10% CAGR mean? It means your capital compounded at 10% a year over the measurement period. $100,000 growing at 10% CAGR becomes about $259,000 after ten years. It does not mean you made 10% in every one of those years. Some were likely much better and some negative.
Is CAGR the same as annual return? No. CAGR is the compound rate, which accounts for gains and losses building on each other. Annual return usually means the simple average of each year’s result, which ignores compounding and overstates performance on any volatile equity curve. The more volatile the system, the wider the gap.
Can CAGR be negative? Yes. If the ending value is lower than the starting value, CAGR is negative and tells you the annual rate at which the account shrank. A losing system produces a negative CAGR in exactly the same way a winning one produces a positive figure.
What is a good MAR ratio for a trading system? Aim for 0.75 or better on a single long-only trend following system, meaning your CAGR is at least three-quarters of your maximum drawdown. Below 0.75, do not discard a system on those numbers alone. If it is negatively correlated with your other systems, it can still add real value to the portfolio as a whole. A portfolio of genuinely uncorrelated systems across different markets should target 1.5 or higher.
Why is my live CAGR lower than my backtest CAGR? Usually some combination of curve fitting, survivorship bias in the historical data, underestimated slippage and commissions, and optimistic execution assumptions. Some gap is normal and expected. A very large gap suggests the system was fitted to the past rather than built on a real edge.
Should I use CAGR or total return to compare two trading systems? CAGR, and only if both are measured over the same period. Total return is distorted by test length, so a system tested over 20 years will show a larger total return than a better system tested over 10. CAGR normalises for time, which is what makes comparison possible.
Where To Go From Here
If you have a backtest with a CAGR you are not sure you believe, you are asking the right question. The next step is learning to read the whole report rather than the headline, and to build systems around a drawdown you can survive.
That is exactly what we work through inside the Trader Success System: how to build, test and validate rules-based systems, how to judge a backtest honestly, and how to combine systems into a portfolio with a better MAR ratio than any one of them could reach alone.
Remember – You are only one trading system away!
