This study investigates the impact of inaccurate assumptions on economic forecast precision. We construct a new dataset comprising an unbalanced panel of annual German GDP forecasts from various institutions, taking into account their underlying assumptions. We explicitly control for different forecast horizons to reflect the information available at the time of release. Our analysis reveals that approximately 75% of the variation in squared forecast errors can be attributed to the variation in squared errors of the initial assumptions. This finding emphasizes the importance of accurate assumptions in economic forecasting and suggests that forecasters should transparently disclose their assumptions to enhance the usefulness of their forecasts in shaping effective policy recommendations.

We test whether organizational risk management matters to bondholders of U.S. bank holding companies (BHCs), and find that debt financing costs increase when the BHC has lower-quality risk management. Consistent with bailouts giving rise to moral hazard among bank creditors, we find that bondholders put less emphasis on risk management in large institutions for which bailouts are expected ex-ante. BHCs that maintained strong risk management before the financial crisis had lower debt costs during and after the crisis, compared to other banks. Overall, quality risk management can curtail risk exposures at BHCs and result in lower debt costs.