A Unified Financial Crime Risk Index: Joint Measurement of Anti Money Laundering & Sanctions Exposure and its implications for Market Risk during Geopolitical Shocks-Replication package

Replication package-Read Me 1. Purpose This Read Me accompanies the manuscript and provides the necessary information: understand the data sources and transformations used in the study, and replicate the key empirical results reported in the paper, including the reported beta (β) estimates, tables, and figures. The replication package supports two replication paths: Path A: Replicate results directly from the provided Excel panel dataset.This reproduces the regression coefficients and summary statistics reported in the paper without requiring downloads from any external data vendor. Path B : Rebuild the Excel panel from publicly available market prices using MATLAB.This downloads daily sector ETF prices and recomputes returns, Expected Shortfall (ES), event-study ΔES, and the regression outputs. 2. Software Requirements MATLAB (R2022a or later recommended) Microsoft Excel (for reviewing and/or replicating regressions from the derived panel) 3. Files Provided This replication package consists of the following files: FCRI_underlying_data_plus_templates.xlsxContains the final event × sector panel dataset used in the regressions and figures. This file is sufficient to reproduce the reported β estimates. download_etf_prices_stooq.mMATLAB script that downloads daily closing prices for sector ETFs. fcri_case_study_run_from_prices.mMATLAB script that (i) constructs returns, (ii) computes downside tail risk (ES at 5%), (iii) computes event-study changes in ES (ΔES), and (iv) produces regression outputs and figures. 4. Data Sources 4.1 Sector ETF Market Data (Public Market Prices) Sector equity exposure is proxied using publicly traded U.S. sector exchange-traded funds (ETFs) (the specific tickers are referenced in the MATLAB code). The analysis uses daily closing prices for these ETFs to compute returns and risk measures. Daily close prices are accessed programmatically via Stooq, which is used solely as a data access conduit to enable license-free replication. The empirical methodology depends only on standard daily closing prices of liquid, widely traded ETFs; it does not depend on any Stooq-specific analytics, adjustments, or proprietary transformations. 4.2 Geopolitical / Sanctions Event Dates Event dates correspond to major geopolitical escalations and sanctions-relevant shocks, identified from public announcements and widely reported events. The event labels and dates used for replication are listed explicitly in the Excel file’s panel tab(s) and referenced by the MATLAB event table. 5. What Is in the Excel File (and What It Is For) FCRI_underlying_data_plus_templates.xlsx is designed as the regression-ready dataset. The key tab used to reproduce the main results is the event-level panel containing, at minimum, the following fields (column names may vary slightly but these variables are present): event (event identifier) event_date (event date used in the event study) sector (sector label) FCRI (sector-level Financial Crime Risk Index) dES5 (event-specific change in Expected Shortfall at 5%, ΔES) This panel is the dataset used in the paper’s core regressions and summary plots. 6. Financial Crime Risk Index (FCRI) Construction The Financial Crime Risk Index (FCRI) is a constructed sector-level exposure index intended to capture relative sensitivity to financial crime and sanctions-related risk. High-level construction: sector exposure components are normalized to ensure comparability across sectors; AML-related and sanctions-related components are combined using the baseline weighting described in the manuscript; the final index is rescaled to the interval for interpretability. Sector FCRI values are reproducible from the FCRI_Inputs sheet using generate_fcri_index.m, which writes FCRI_Scores and populates the FCRI column in Panel_Data prior to the event-study and regression steps. For convenience, the final sector-level FCRI scores used in the empirical analysis are provided directly in the Excel panel dataset and are treated as fixed regressors in the regression models. 7. Downside Tail Risk Measure: Expected Shortfall (ES, 5%) Downside market risk is measured using Expected Shortfall (ES) at the 5% level. In the full MATLAB pipeline (Path B), the computation is: download daily close prices for each sector ETF; compute daily returns; compute ES(5%) using the rolling estimation procedure described in the manuscript; compute event-window changes in ES (ΔES). For convenience (Path A), the resulting ΔES values are already provided in the Excel panel as dES5. 8. Event Study: Construction of ΔES (dES5) For each event and sector : estimate ES in a pre-event window; estimate ES in a post-event / event window; compute the change: The resulting event-by-sector ΔES values correspond to the dES5 field in the Excel panel. 9. Regression Models and How to Replicate β 9.1 Core cross-sectional model (event-level) For each event , the paper estimates: Replication using Excel (Path A): For each event, regress dES5 on FCRI across sectors. 9.2 Pooled model (with event controls) The pooled specification estimated in the paper is: Replication using Excel (Path A): Run a pooled regression of dES5 on FCRI including event fixed effects (event dummies). Result reproducibility:The β coefficients reported in the manuscript are reproducible directly from the provided Excel panel. 10. Reproducing Figures Figures in the manuscript are generated from the same event-level panel. From Excel (Path A):Figures that plot average ΔES vs FCRI (or high- vs low-FCRI group comparisons) can be reproduced directly from the panel by aggregating dES5 by sector (and by event where applicable). From MATLAB (Path B):Running fcri_case_study_run_from_prices.m reproduces the event-study panel and then generates the plots. 11. Replication Paths (Step-by-Step) Path A: Open FCRI_underlying_data_plus_templates.xlsx Locate the event × sector panel tab containing FCRI and dES5 Run: event-level regressions: dES5 ~ FCRI for each event pooled regression with event dummies: dES5 ~ FCRI + event FE Compare β values to those reported in the manuscript. Path B: Place download_etf_prices_stooq.m and fcri_case_study_run_from_prices.m in the MATLAB working directory Run fcri_case_study_run_from_prices.m The script will: download daily ETF prices (via Stooq) compute returns and ES(5%) compute ΔES around each event produce regression outputs and figures Compare the resulting β and plots to the manuscript. 12. Notes / Reproducibility Considerations Stooq is used only as a public access point for daily close prices; results are not vendor-dependent. Minor numerical differences may occur if a replicator changes: event-window definitions, rolling window parameters, sample endpoints, or ticker mappings. For strict replication of manuscript numbers, use the event definitions and panel as provided in the Excel file. 13. Correspondence For replication questions, please contact the corresponding author listed in the manuscript.