Backtest
A backtest runs a fixed set of trading rules over past prices and records what would have happened: every trade, the return per year, the largest fall and the time spent in the market. The rules must be written so exactly that anyone following them would make the same trades.
A backtest describes the past under stated assumptions, such as the data source, the dates and whether trading costs are included. It cannot say what will happen next. Its most useful comparison is with buying and holding the same investment over the same dates.
Results can be flattered by survivorship bias, by using information that was not available at the time, and by trying many settings and reporting the best. A careful backtest says how many variations were tried and checks the result on data that was not used to choose the settings.
In one sentence
Running a fixed set of trading rules over past prices to see what would have happened.
This definition is research and education, not financial advice.