Portfolio backtesting guide
How to read a portfolio backtest
A portfolio backtest is a reconstruction of how selected holdings and rules would have behaved over a historical window. Reading it well means checking the setup before focusing on the headline return.
Published · Updated
Check the setup first
Record the start and end dates, instruments, weights, rebalance rule, fees, cash treatment, and price basis. A backtest answers only the question defined by those choices. A different date range or rebalance rule is a different historical experiment, not a correction to the first one.
Read the path, not only the endpoint
Look at cumulative value together with annualized return, drawdown depth, recovery time, and the periods that drove the outcome. Compare the same window and assumptions when using a benchmark. SumWorth’s historical stress and performance lenses are useful for asking where the path was difficult, not for selecting a future winner.
Separate evidence from the model
Market observations, calculated portfolio values, fees, and scenario overlays do not carry the same evidential status. Check coverage and stale or missing observations before treating a smooth line as meaningful. Even a carefully calculated historical result is not a forecast, recommendation, or guarantee of future performance.
Evidence boundary
Backtests are historical reconstructions that depend on chosen dates, data, fees, rebalancing, and other assumptions. They cannot demonstrate that a strategy will repeat or suit a person.