How many trades do you need? Sample size and start-date tests
Why a few dozen trades prove almost nothing, how to judge whether a result could be luck, and why you should test every start date.
Toss a coin twenty times and you might easily get fourteen heads. That does not make it a winning coin. Trading results work the same way: with only a handful of trades, luck can make a weak strategy look strong and a good one look broken.
More trades, narrower doubt
The uncertainty around a strategy's true average result shrinks with the square root of the number of trades. Four times more trades roughly halves the doubt. A few hundred trades across several years and different market moods is a sensible minimum before drawing conclusions.
Start-date tests
A single backtest starts on one particular day. If that day happened to be just before a great run, the result looks better than a typical user would experience. A start-date test runs the same strategy from many different starting points and looks at the whole spread of outcomes: the typical result, the worst start and how long the worst start took to recover.
We use start-date tests in our own research because they answer the question a new user actually has: what could happen if I start now?
Start the bot, watch your free $10,000 demo trade, and connect your Exness account when you are ready.
Read next: Spread, slippage and commission: the costs that quietly decide a backtest
Risk warning. Trading leveraged products such as CFDs carries a high risk of losing money. Backtest figures are simulations, not real results, and past performance does not guarantee future results. Read the risk disclosure before you connect an account. Nothing here is financial advice.