The Price of Permission: Classification Uncertainty in Constrained Capital Markets

Battle Count: 220

The paper is relevant to quantitative trading primarily as a portfolio-monitoring and risk-management framework rather than a direct trading signal generator. It identifies classification-boundary risk as a state variable for constrained portfolio managers, showing where feasible universes are fragmented or vulnerable to change. The event-study evidence (1.76pp over [0,10] for SC Malaysia inclusions) suggests exploitable price patterns around official reclassification events, but the authors explicitly state this does not constitute a validated profitable strategy. The monitoring variables (disagreement, proximity) rank future transitions, which could inform rebalancing timing and liquidity planning. The U.S. evidence shows formal eligibility changes need not reprice, cautioning against naive event-driven strategies. The work is more directly applicable to mandate-constrained fund management and compliance monitoring than to high-frequency or statistical arbitrage.

Implementation requires: (1) constructing seven researcher-emulated Shariah rulebooks from Compustat data with specific denominators, averaging windows, and thresholds; (2) building a 1.3M+ security-month panel with proper lagged accounting data and stale-data guards; (3) computing cross-standard disagreement, nearest-boundary distance, and proximity risk transformations; (4) implementing matched event studies with greedy nearest-neighbor selection on five standardized covariates; (5) performing wild-cluster bootstrap inference with only 24 clusters; (6) parsing 25 official SC Malaysia PDF lists with release-date corrections; (7) integrating LSEG Workspace holder snapshots and ETF data for mechanism diagnostics. The multi-standard screening logic, matching design, and small-cluster inference require careful implementation. Proprietary data access (CRSP, Compustat, LSEG) is essential.

The paper promises a replication package with programs, data dictionary, merge-key documentation, parser logs, and synthetic input files. However, core data sources (CRSP, Compustat, WRDS, LSEG Workspace, index-provider inputs) are proprietary and cannot be redistributed. SC Malaysia list files are publicly available. The workflow is documented but full reproduction requires licensed data access. The Internet Appendix provides extensive diagnostic tables and figures supporting the main claims.