Your numbers
What it says
Ready to go further
Next: how deep was the worst fall?
Sharpe summarises the ups and downs. The drawdown calculator shows the single worst stretch and the gain needed to recover.
What is included
- The calculator, with your own numbers
- The formula written out
- A worked example you can reproduce
- Where grid, martingale and DCA bots break the maths
Who it is for
- People running or testing a trading bot
- Anyone sizing a position or checking costs
- Readers learning how the numbers work
- Not advice, and not for placing trades
Key facts
- Sharpe ratio = (mean return - cash rate) / standard deviation of returns. It is return per unit of up and down movement.
- To annualise a monthly Sharpe, multiply by the square root of 12. This assumes the returns are independent from one period to the next.
- Sharpe measures how smooth the returns were. It does not see a loss that has not happened yet, so martingale and grid bots can show a very high Sharpe until the day they do not.
- With few periods the estimate is noisy. 12 months of data can give an annual Sharpe that is wrong by 2 or more.
- All example series on this page are made up. A bot's record can be built from many more periods than the page assumes.
How the calculator works
It takes your list of returns and works out the average, then how far returns typically sit from that average. The Sharpe ratio is the first divided by the second.
Standard deviation = sqrt( sum of (return - mean)^2 / (n - 1) )
Sharpe per period = (mean - cash rate) / standard deviation
Annualised Sharpe = Sharpe per period x sqrt(periods per year)
The standard deviation uses n - 1, the usual sample form. The cash rate is the return you could have taken by leaving the money in cash, entered per period in the same unit as the returns.
A worked example
The default series is twelve made-up monthly returns: 2, -1, 3, 1.5, -2, 4, 0.5, 2.5, -1.5, 3, 1 and 2 percent. The mean is 1.25 percent and the sample standard deviation is 1.91 percent. The Sharpe ratio for one month is 1.25 / 1.91 = 0.65, and annualised it is 0.65 x sqrt(12) = 2.26.
The same Sharpe is much less certain than it looks
A common approximation for the standard error of a Sharpe ratio from independent, roughly normal returns is sqrt((1 + 0.5 x Sharpe^2) / n) for the per-period figure. For the series above, scaled to a year, the 95 percent range is about 2.26 plus or minus 2.16, so from 0.10 to 4.42. Twelve months of data cannot separate a Sharpe of 0.1 from one of 4. The range narrows slowly:
| Months of data | Rough 95% range, plus or minus |
|---|---|
| 12 | 2.16 |
| 24 | 1.53 |
| 60 | 0.97 |
| 120 | 0.68 |
Same approximation, assuming a per-month Sharpe of 0.65. Calculated by us on 27 Oct 2026, rounded. Real returns are rarely independent or normal, so the true uncertainty is usually larger.
What Sharpe does not see
Standard deviation counts all movement equally and only counts the movement that has already happened. That is a problem for the bots this site is about.
- A steady bot looks perfect. Take twelve made-up months of +1.0, +1.2, +0.8, +1.0, +1.1, +0.9, +1.0, +1.2, +0.8, +1.0, +1.1 and +0.9. The mean is 1.00 percent, the standard deviation is 0.13 percent and the annualised Sharpe is 25.69. Almost no real strategy looks like that.
- One bad month ends it. Replace the last month with -12 percent, as a martingale ladder hitting its last step might. The mean falls to -0.08 percent and the annualised Sharpe to -0.07. The same strategy went from a number that looks like a miracle to a loss in one month.
- Smoothing is not safety. Many small wins and rare large losses give a high Sharpe over short samples. The ratio rises with the very thing that makes the loss bigger when it comes.
- Costs and slippage must already be in the returns. A Sharpe computed from gross returns is not a Sharpe anyone can earn.
- Drawdown is not the same as volatility. The drawdown calculator shows the worst peak to low fall, which a Sharpe ratio does not report.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
What is a good Sharpe ratio?
There is no single good number, and the answer depends on how many periods it came from. A figure from 12 months can easily be off by 2 or more. Look at the range, the costs in the returns and the worst drawdown beside it.
Should I use monthly, weekly or daily returns?
Use the period your record has most of, and set periods per year to match. Annualising daily returns by square root of 252 assumes the returns are independent day to day, which is often untrue.
Why is Sharpe undefined for steady returns?
If every return is the same the standard deviation is zero and the ratio divides by zero. A record that flat is usually a sign the data is a simulation, a rounding or a bot hiding its open losses.
Can a martingale bot have a high Sharpe?
Yes, until a loss arrives. The wins are small and regular, so the standard deviation is low. The loss is rare, so it may not be in the sample at all.
What cash rate should I use?
The return on cash for the same period length, in the same units. If it is small compared with your returns, leaving it at zero changes little.
Are the numbers on this page real?
Example numbers are made up to show how the maths works, and the page says so. A figure only counts as exact when the page says it was calculated on a stated date. Nothing you enter in a calculator leaves your browser.
Can I use this to decide whether to run a bot?
It can show whether a bot's numbers hold together and how much an account can lose. It cannot say what the market will do, and it is not advice. Use it to ask better questions, then read the terms of your own broker or exchange.
Is this financial advice?
No. It is a calculator and general education. It does not consider your situation, and its output is only as good as the numbers you enter.
Know the Risk Before You Run It
Free calculators and guides, added every week. Leave your details and we will email you when new ones go up.
Trading carries a risk of loss