Manuscript Revision

1. Direct Mapping: Every task in the "Guía de Prompts para Revisión" must be mapped to a specific change in the LaTeX and/or Notebook.
2. Traceability: Figures and tables in the paper MUST match the outputs of the notebook exactly.
3. Academic Prose: Maintain a formal, descriptive, and neutral tone. Use standard Elsevier-style structures.
4. Causal Caution: Avoid strong causal verbs (e.g., "causes", "drives") and prefer associations (e.g., "is associated with", "predicts", "acts as a leading indicator").

🛠️ Specialized Procedures

Influence Analysis (Tarea 1)

• Generate Cook's D and DFBETAS via statsmodels.stats.outliers_influence.
• Plot influence measures against time or observation index.
• Document regression results excluding periods with High Influence (e.g., COVID peak).

Bootstrap Inference (Tarea 2)

• Implement Wild Bootstrap for small-sample regression inference (N=25).
• Report confidence intervals and contrast them with standard OLS t-stats.

Horse-race Comparisons (Tarea 3)

• Use VIX and Baker et al. (EPU) as benchmarks.
• Perform joint regressions: $\sigma^{\text{rev}}_t = \beta_0 + \beta_1 CEUI_t + \beta_2 VIX_t + \beta_3 EPU_t + \epsilon_t$.

📜 LaTeX Standards

• Use threeparttable for all tables.
• Use \footnotesize or \small instead of resizebox for fitting tables.
• Ensure all labels are unique and consistent with sections.

🔄 Workflow Loop

1. Identify Tarea from Guide.
2. Execute Python code to generate table/figure.
3. Insert LaTeX in .tex file.
4. Verify PDF compilation.
5. Update revision_tracking.md.