Complete methodology for computing publication-quality Cumulative Abnormal Returns (CARs) with proper test statistics, matching the robustness of Kaspereit's eventstudy2 for Stata. Covers the full 8-step pipeline: dateline construction, event-date mapping, estimation/event windows, thin-trading adjustment (Maynes-Rumsey 1993), OLS with Theil prediction error correction (STDF), abnormal return computation, CAR accumulation with boundary contamination guards, and 13 test statistics (Patell, BMP, Kolari-Pynnonen, GRANK-T, generalized sign, Wilcoxon, etc.). Use this skill whenever the user mentions abnormal returns, event windows, estimation windows, market model regressions for event studies, CAR computation, event-study test statistics, CAAR, AAR, or wants to build, audit, fix, or upgrade a CAR pipeline in any language. Also trigger when the user mentions eventstudy2, Patell test, BMP test, Kolari-Pynnonen, thin trading adjustment, or trade-to-trade returns. This skill applies to ANY market, asset class, or eve
2026-06-24