The rule of thumb is that if the annualized Sharpe ratio exceeds 2–3 on a simple strategy, one should first suspect methodological problems (leakage or overfitting) rather than treating it as a success.