Overfitting

In short

Overfitting means a strategy was tuned to historical data until it explains the past perfectly — and fails the future.

The more parameters and the prettier the backtest curve, the bigger the suspicion. The only honest test is data the strategy has never seen.

How does a layperson spot overfitting?

Too-perfect results, many "magic" parameters (why exactly 17 days?), no out-of-sample proof.

Information, not investment advice. More of this? The newsletter.