We examine earnings forecasts by sell-side analysts employed by a bank with a lending relationship with the covered firms. We find that lender-affiliated analysts’ forecasts are more accurate than forecasts by their unaffiliated peers after establishment of the lending relationship. Evidence from exogenous variation suggests that the relationship is causal. Lender-affiliated analysts are also more likely to issue pessimistic forecasts below their peers’ consensus. These forecasts are likely to be followed by below-consensus earnings. The results suggest that lender-affiliated analysts enjoy an informational advantage that spills over from lending activities of banks.