I study the relationship between sell-side analysts’ forecast bias and stock returns by comparing three forecast families—price targets (PTG), earnings per share (EPS), and long-term growth (LTG)—to ex-ante machine-learning (ML) benchmarks. Bias in return expectations co-moves positively with bias in cash-flow expectations, suggesting a common behavioral source of belief distortions. Analysts’ deviations from ML benchmarks across all forecast families predict returns in a systematic and nonlinear manner. Across horizons, forecast bias predicts returns positively in the short term but negatively in the long term, with LTG-based return-predictability regressions exhibiting similar patterns. Across the sign of bias, optimistic (pessimistic) forecast bias positively (negatively) predict returns. Finally, I document a common tendency to anchor expectations on consensus forecasts. An asset-pricing model with asymmetric information and positively skewed fundamental shocks generates these dynamics: analysts anchor on consensus and misperceive the informativeness of signals, producing the observed bias–return relation.