How is backtested performance different from live performance?
A backtest applies rules to an earlier market period. A live model record begins after the model is published and tracks decisions as new market data arrives. Both can be useful, but they are not equivalent evidence and should not be combined without clear labels.
Rules applied to historical markets
A backtest explores how a defined process would have behaved. Its value depends on data quality, assumptions, validation, and protection against hindsight.
Records rebuilt from saved decisions
A reconstructed record can use rankings or publications saved before outcomes were known. It is stronger than rebuilding with hindsight, but it is still not a live account result.
Results recorded after publication
A forward-tracked model preserves what was published and measures what followed. It tests whether the production process behaves as documented.
An account adds implementation effects
Actual results can differ because of prices, timing, cash flows, taxes, fees, fractional shares, missed updates, and investor decisions.
- When were the strategy rules and parameters finalized?
- Was each decision saved before the return being measured occurred?
- Which dates belong to research, reconstruction, and forward tracking?
- Are dividends, benchmark returns, costs, and timing handled consistently?
- Does the presentation show weak periods as well as strong ones?
The SEC's performance claims bulletin describes backtesting as applying a model to past markets and recommends asking for actual historical performance, calculation methodology, reasonable time periods, and an appropriate benchmark.
Review the backtest and validation evidence, the profile records, and the publication audit trail. The record labels explain which history is reconstructed from saved rankings and when forward tracking begins.