Tag
#overfitting
4 articles
-
In-Sample vs Out-of-Sample Testing: The Minimum Honest Backtest Check
In-sample data is where a strategy is built and tuned. Out-of-sample data is held back and used once to judge it. Splitting the two is the least you can do to avoid fooling yourself.
Read article -
Walk-Forward Analysis Explained: Rolling Re-Estimation as a Stricter Test
Walk-forward analysis fits a strategy on a past window, applies it to the next unseen window, then rolls forward and repeats. It is a harder test than a single out-of-sample split.
Read article -
What Is Overfitting in Backtesting? Curve-Fitting, Parameters and How to Detect It
Overfitting in backtesting is tuning a strategy until it describes the noise in one sample of history rather than any durable pattern. It looks like a great result and behaves like a coin flip.
Read article -
Why Backtest Results Don't Repeat in Live Investing
Backtest results rarely repeat live because of overfitting, regime change, crowding, and the costs a test omits. Here is what actually causes the gap and what to expect.
Read article