Overfitting

Building a model that fits historical data too closely, capturing noise rather than genuine patterns, resulting in strong backtested performance that fails in live trading.