SPB Git

spb/wp17_uqo

Public
History of scripts/r3k_23_construct_validity.py Clear filter
page 1
  1. Thu, Nov 5, 2026 1

    1. FIRC0002 L2–L10 results (run 7): rebuilt panel, core + extended + event-study + survivorship results, construct-validity rows, 33 regenerated tables with notes, 442 macros, 17 figures
      …
      AR-FIRC0002-1: trigger (a) not fired (wins. EW spread +3.9 pp NW t 2.06 / plain 1.81, CI covers 0; wins. FM +1.88 pp/SD t 2.41);
      trigger (b) not fired on the literal criterion (large-vs-small gradient persists in sorts, contrast t 1.83 firm / 1.55 two-way,
      wild p 0.22) but the largest-1,000 1-y CDA t is 3.02 (IWV model) vs 1.12 beta-only / 1.21 market-adjusted / 0.51 FF6;
      trigger (c') not fired (size-tercile attrition; AI-quintile attrition = X1).
      
      Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
      Simon-Pierre Boucher · 1 file +69
Menu

Theme