A Data Mining Approach To Build AML Indices A Case Study PDF Free Download

1y ago

13 Views

1 Downloads

1.21 MB

27 Pages

Report/dmca

Download PDF

Transcription

A Data Mining Approach to Build AML IndicesA Case StudyClaudio Antonini, Ph.D.Deloitte Financial Advisory Services LLPNew York

MotivationIn 2012, the FSA fined a bank for “failure to take reasonable care to establish andmaintain adequate anti-money laundering (AML) systems and controls [and to]assess the level of money laundering risk posed by its customers.”In particular, 46 of 68 accounts reviewed by the FSA “had been inappropriatelyclassified as normal risk.” [*]Conclusion: The firm has to have a defensible way of assessing risk.[*] 2/055.shtmlDeloitte1.Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Expanding an Existing Index Basel AML Index– Origin Basel Institute on Governance Non-profit Corruption prevention, public/companygovernance– Composition Various sources: indices, reports– Methodology Relies on experts that determine weights– Limitations indicated by reviews of the2012 release Sources — Infrequent, some might bebiased Methodology — no CI,uncertainty/sensitivity analysis– Limitations not indicated by reviews Missing data Non-reproducible, some data difficult tolocate Limited number of countries and regionscovered“The Basel AML Index 2013,” at http://index.baselgovernance.org/index/Project Description.pdf2Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Expert Weightings — Risks5%5%10%Money Laundering — Terrorist FinancingFinancial Transparency and StandardsCorruptionPublic Transparency and AccountabilityPolitical and Legal Risk15%65%“The Basel AML Index 2013,” at http://index.baselgovernance.org/index/Project Description.pdf3Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Areas Covered — Sources“The Basel AML Index 2013,” at http://index.baselgovernance.org/index/Project Description.pdf4Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Availability of Sources — CoverageOrganizationIndicatorBasel Institute on GovernanceBasel AML Index1.) Bertelsmann Stiftung Transformation Index2005200620072008200920102011Rule of Law scores20122013xJun-10x2.) EuromoneyPolitical Risk scores3.) Financial Action Task Force (FATF)Member countries Mutual Evaluation Reports4.) Freedom HouseFreedom in the World & Press Freedom Index5.) International Institute for Democracy and ElectoralAssistance (IDEA)Political Finance Database6.) International Budget PartnershipOpen Budget Index7.) Tax Justice NetworkFinancial Secrecy Index8.) Transparency InternationalCorruption Perception Index9.) US State Dept. - Int. Narcotics Control Strategy ReportMoney Laundering and Financial Crimes10.) World Bank - Doing Business RankingBusiness Extent of Disclosure Index11.) World BankIDA Resource Allocation Index12.) World Economic ForumGlobal CompetitivenessInternational Monetary FundCompliance w AML CFTSources5Data mining approach to build xxxxxxx2x3x2x3x2x3xxx663Copyright 2013 Deloitte Development LLC. All rights reserved.

Missing Data in Most CircumstancesCountry NameAfghanistanAlbaniaAlgeriaAmerican SamoaAndorraAngolaAntigua and BarbudaArab anBahamas, TheBahrainBangladeshIndicator Name1990Patent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidents229Patent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidents 1955Patent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidentsPatent applications, nonresidents 670Patent applications, nonresidentsPatent applications, nonresidents 25Patent applications, nonresidentsPatent applications, nonresidents 815226130178028207943224806524122 23525 21187 5162729318971897727047814278112756276 In most regressions schemes, only data from a few countries orregions would remain6Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Defensible Process Desired to Value Risk It is desired to build an index:– Maintainable in-house– Reproducible– Based on available sources of information– Updated as sources are updated (not once a year)– Informative (not only generating point values)– Valid for new cases (not just previous ones)7Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Process Followed Included more time-series– Basel AML Index, IMF AML CFT Index, WDI, WGI Treatment of missing data– Missingness– Imputation– Create complete cases Modeling– Decision Trees– Random Forest– Linear Models8Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Indicators — Missing DataMissingness vs. IndicatorsLess data(WDI 1 to 31, AML 32 to 45)120%WDIBasel AML 641More data12Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Indicators — Missing DataMissingness vs. Indicators(WDI 1 to 31, AML 32 to 45)Less dataWDI100%80%Basel AML I60%40%20%1982-20120%2002-20121More data1435201279111315Data mining approach to build -20124345Copyright 2013 Deloitte Development LLC. All rights reserved.

Few Changes in the Index from 2012 to 2013BASEL AML Index 2013 - 2012Positive difference -- getting better ( more compliant)1.5EcuadorMoldova1GeorgiaLaosSlovak 1.51611 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 101 106 111 116 121 126 131 136 141 146Country Rank 2013, ordered by compliance score (lowest Afghanistan, highest Norway)Deloitte16.Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Forecasting Slow process– y(t) a(t) β(t) * x(t)– y(t 1) a(t 1) β(t 1) * x(t 1)Assume that a(t) a(t 1), β(t) β(t 1)– y(t 1) y(t) β(t) * Δx(t- t 1) Can also forecast the individual time-series. No new series until Jun-1017Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Decision Tree Model (143 countries)Data collection is usually limited in less developed countries modeling bias43% of the rows were deleted due tomissing data. The model was builtwith data from only 82 countries.18Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Decision Tree Model (246 countries)After imputation, all 246 rowsare used to create the model. Amore detailed tree is obtained.19Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Decision Tree Model (246 countries)Predicted vs. ObservedDecision Tree Modelcomplete imp246423Predicted56Linear Fit to PointsPredicted Observed1Pseudo R-square 0.8357234567Basel AML IndexRattle 2013-Jun-13 22:38:32 Patricia20Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Linear Model (246 countries)7Predicted vs. ObservedLinear Modelcomplete imp2464123Predicted56Linear Fit to PointsPredicted ObservedPseudo R-square 0.927612345678Basel AML IndexRattle 2013-Jun-14 09:58:44 Patricia22Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Linear Model (246 countries)Initial (143) and imputed (103) index values (in red), and their (246) estimatesIndex extended toadditional 103countries andregions[1] Original 143 countries (red)[2] Estimates of 143 countriesIndex Value[3] Imputed values (red)[4] Estimates onimputed valuesInitial 143 countriesCountry23Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

Different approachesItemOther IndicesOur ApproachOptionsExpertsUse of non-reproducible,‘arbitrary’ weightsRegression, decisiontrees, Random ForestVarious models,supervised learningIndexN/A (The index isgenerated, not used as areference.)Still need a referencefor modelingWe can select from agrowing number ofindicesSourcesPotentially biasedPublic dataWe can select from agrowing number ofdata sourcesDataSporadic, difficult toobtain, categorical We can select from agrowing number ofimputation methodsCountriesLimited numberUnlimited, andextended to regionsEstimatesOnly point valuest-stats, CI24Data mining approach to build indicesSelect sourcesImputationCopyright 2013 Deloitte Development LLC. All rights reserved.

Conclusions Limitations of the Current Procedure to Create an AML Index– Reproducibility– Reliance on experts– Estimates restricted to point values– Limited modeling options due to missing data– Many data points are deleted, resulting in biased estimates– New points (in our case, countries) cannot be scored Given the amount of public data available– The index can be easily replicated and extended– Other related indices can be used– No need of expert weights– No need to rely on sporadic or potentially biased sources The process can be applied to other indices/scores– Use one index as a reference to determine variables– Identify relevant variables with various models– Impute data (do not delete variables to create model)– Create various models for different purposes25Data mining approach to build indicesCopyright 2013 Deloitte Development LLC. All rights reserved.

About DeloitteAs used in this document, "Deloitte" means Deloitte Financial Advisory Services a subsidiary of Deloitte LLP. Please seewww.deloitte.com/us/about for a detailed description of the legal structure of Deloitte LLP and its subsidiaries. Certain service maynot be available to attest clients under the rules and regulations of public accounting.This presentation contains general information only and is based on the experiences and research of Deloitte Financial AdvisoryServices LLP practitioners. Deloitte Financial Advisory Services LLP is not, by means of this presentation, rendering accounting,auditing, business, financial, investment, legal or other professional advice or services. This presentation is not a substitute for suchprofessional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Beforemaking any decision or taking any action that may affect your business, you should consult a qualified professional advisor.Deloitte Financial Advisory Services LLP, its affiliates, and related entities shall not be responsible for any loss sustained by anyperson who relies on this publication.Copyright 2013 Deloitte Development LLC. All rights reserved. Member of Deloitte Touche Tohmatsu Limited

A Data Mining Approach to Build AML Indices A Case Study Claudio Antonini, Ph.D. Deloitte Financial Advisory Services LLP New York