In-sample data

In-sample data is the slice of history you build and tune a strategy on. You look at it, try ideas, and adjust parameters against it during development.

Because the strategy is shaped to fit this data, its in-sample performance is optimistic by nature. A good in-sample result tells you how well you fit the past, not how the strategy will do on data it has never seen. The more parameters you tune, the better the in-sample numbers look and the less they mean. That is why a strategy’s real test comes from out-of-sample data it was never fitted to.