How to backtest a trading strategy for free — and read it honestly
A backtest replays a set of trading rules over real past prices and asks a single question: would this have done better than just buying the asset and holding it? It takes about thirty seconds here, no code and no sign-up. The hard part isn't running it — it's reading the answer without fooling yourself. This walks through both.
Run one in 30 seconds (no code)
- 1. Pick a famous asset — a stock like Apple or Tesla, an ETF, or a coin like Bitcoin.
- 2. Pick a well-known strategy — a moving-average crossover, RSI mean-reversion, a breakout. Each one on the site comes with a plain-English description of exactly what it does.
- 3. It runs on real history and shows you the result against buy & hold. Change the settings and the window, and run it again.
This link runs RSI mean-reversion on Apple over the last five years — a real backtest, live, right now. Change the asset or the rule once it opens.
How to read what you got — without fooling yourself
A backtest is easy to misread in your own favour. Five things to look at, in order:
1. Read the excess, not the return
“Up 120%” means nothing on its own — the asset itself may have tripled. The number that matters is the excess over buy & hold: did the rule beat simply holding? If a strategy made money only because its market went up, it didn't earn its complexity. Every result here leads with that comparison.
2. Look at the drawdown you'd have sat through
The worst peak-to-trough fall is the loss you'd have had to hold through without quitting. A higher return bought with a much deeper drawdown is not obviously better — most people sell at the bottom of it.
3. Check the number of trades
A spectacular result built on five trades is a coin that came up heads five times, not an edge. Few trades = a small sample = treat it with suspicion. (There's a whole lesson on telling a real signal from luck.)
4. Make sure the costs were on
Every fill is charged fees and slippage, and no signal can act on a price it couldn't have known — the rules only see bars that have already closed. That's why some “profitable” ideas quietly turn negative — the friction was always there; most backtests just hide it. The full recipe is on the methodology page.
5. Remember what a backtest can't tell you
It shows what would have happened on one slice of the past. It can't tell you it will work next year, that the settings weren't quietly fitted to this exact history, or that the market won't change regime. A good backtest is a hypothesis to test forward — not a promise. Running one is just one stage of preparing to risk real money — not the finish line.
The sobering part most sites skip
When you honestly test the famous strategies on the famous assets — costs on, losses shown — most of them lose to simply buying and holding. That isn't a bug in the strategies; it's the base rate, and it's exactly why testing before you trade is worth doing. You can see the whole scoreboard — every winner and every loser — on the honest results page. Finding the rare rule that genuinely beats holding is the point; assuming yours is one of them is the trap.
Hypothetical / simulated results — past performance is not a reliable indicator of future results. This is an educational, analytical tool, not investment advice and not a recommendation to buy or sell anything. You make all decisions and execute on your own broker.