big risks require big data thinking: how new technologies ... · pdf file top fraud risk...

Click here to load reader

Post on 29-May-2020

0 views

Category:

Documents

0 download

Embed Size (px)

TRANSCRIPT

  • The better the question. The better the answer.� The better the world works.

    Information Security and the Law Seminar

    Big Risks Require Big Data Thinking: How new technologies can detect and prevent fraud incidents.

  • Page 2

    “In [Enter your God Here] We Trust, all others must bring DATA” W.E. Deming

    all rights reserved.

    EY 2015

  • What exactly is Big Data? (and why is this relevant)

  • Page 4

    Big Data, the weird and the wonderful

    NBA teams are using it for improving

    athlete performance.

    Large scale retailers for predicting

    buying habits

    F1 teams for optimizing pit stops and

    evaluating car performance in real

    time.

    Supply Chain and Logistics for

    improving warehouse to shelf JIT

    delivery

    We will be using it for Forensic

    Analysis and Corruption Risk

    mitigation. BUT HOW?

    all rights reserved.

    EY 2015

  • Page 5

    The Three V’s of Big Data

    V e lo

    c it y

    Near Real Time

    Real Time

    Batch

    Periodic

    Frequent / Automated “Big data is high-volume, high-velocity and high-

    variety information assets that demand cost-

    effective, innovative forms of information processing

    for enhanced insight and decision making”, Gartner

    all rights reserved.

    EY 2015

  • Page 6

    Forensic Data Analysis End to End

    Understand and Collect

    Structured and Unstructured data.

    Data about Data is pivotal for using Big data.

    Apply Data Analytics on assumed business understanding and with the right technology

    Interpretation of trend analysis and exception reporting based on harnessed data.

    all rights reserved.

    EY 2015

  • Page 7

    Forensic Data Analysis Using big data

    Useful insight

    1

    How to separate the noise from all that data? What do you need?

    Identify relevant sources of data.

    Databases, devices, social media etc

    2 Extract information using the assistance of Domain experts and data scientists.

    3 Involve domain experts for the industry in question. Define Key Fraud Indicators and Metrics

    4 Deploy the right technology to harness

    the data using Key Fraud indicators.

    5 Present or Visualize the information in front of you.

    all rights reserved.

    EY 2015

  • Page 8

    Forensic Data Analysis Using technology to draw correlations

    Correlate a document to an email to data gathered from an End User Device between them.

    Draw relationships from unstructured data to transactional data entries in typical applications or systems. Process can start from an entry back to other sources.

    In investigations – the primary purpose would be to draw a timeline of relevant events to the case.

    In Proactive Forensic analysis – Use of Fraud indicators or metrics to proactively detect fraud.

    all rights reserved.

    EY 2015

  • Times are changing

  • Page 10

    [FACT] Regulatory scrutiny increases. UK (Bribery Act) and US (Foreign

    Corrupt Practices Act FCPA) provide guidance with respect to an

    effective compliance program. Adopting FDA procedures into the

    monitoring and testing compliance can create a cycle of improved

    adherence to company policy and improved fraud prevention and

    detection.

    Shifting Tectonic Plates of Regulation

    466 Executives Interviewed as part of this survey

    11 Markets From 9 different industries

    Nov 2013 to

    Jan 2014

    Current regulatory landscape creates further impetus for new approaches

    in Forensic Data Analysis.

    Governments around the globe are strengthening anti-corruption and

    enforcement continuous to grow. Some 40 member countries of the

    Organization for Economic Co-Operation and Development (OECD),

    including US, Russia, Brazil, Germany, France and the UK have adopted

    the OECD Anti-Bribery Convention.

    Forensic Data Analysis

    Global Survey

    all rights reserved.

    EY 2015

  • Page 11

    Top Fraud Risk Concerns

    Fraud incidents reported according to the Association

    of Certified Fraud Examiners (ACFE) on their latest

    report for 2013.

    1300

    Months is the median duration of a fraud before

    detected

    18

    Of the fraud incidents committed resulted in over 1

    million USD in losses

    >20%

    FDA Aligned well with perceived risk areas

    74% Of the cases, FDA is used for detecting Bribery and corruption risk

    62% Of the cases, FDA is used for detecting financial statement fraud.

    Top risks

    all rights reserved.

    EY 2015

  • Predictive Fraud Analytics

  • Page 13

    Big Data has Big Potential

    72 %

    7 %

    2 %

    Believe big data can play a key role in fraud prevention and detection

    Are actually using them

    Are aware of any specific big data technologies

     Increasing Data Volumes. IBM suggests that 90% of the worlds data has been generated over the past two years.

     Convergence of unstructured data from social media, email, free- text.

     Merge of unstructured data into traditional accounting analytics which historically rely on numerical information.

    Challenge creates Opportunity

     Model-based Mining and visual analytic tools

     Sophisticated BAU prevention controls based on trend anomalies for stopping improper payments or business transactions

     Integration of statistical analysis and text mining

     Implementation of continuous monitoring detection / prevention framework

     Faster event correlations and investigation of relevant facts.

     Faster resolution of the truth using hard core evidence.

    all rights reserved.

    EY 2015

  • Page 14

    Why use Forensic Data Analysis: key benefits and adoption

    FDA enhances the risk assessment process and improves

    fraud detection

    The top benefits reporting form the use of FDA are the ability

    to “enhance our risk assessment process” followed by the

    ability to detect potential misconduct that we couldn’t detect

    before”. This sentiment is reflected in both C-suite and other

    executives. !

  • Page 15

    Forensic Data Analysis rollout within the organization

    Who can Benefit the most

    Common FDA Testing Area

    Payment Stream, accounts

    payable analysis Altered invoices, inflated process, suspicious payment

    descriptions, requestor/approver conflicts

    Common Fraud

    Vendor master / employee

    analysis and comparisons Fictitious vendors, vendor risk ranking, background

    due diligence

    Bribery & Corruption Bid Rigging, conflicts of interest, contract compliance,

    kickbacks, illegal gratuities

  • Page 16

    Missed Opportunities and biggest Challenges

  • Technology Matters

  • Page 18

    Forensic Data Analysis Technology Maturity Model

    of respondents use MS Excel as

    the main FDA tool

    Traditional focus on the upper left

    quadrant handling structured rule-based

    analytics

    Use database tools such as MS

    SQL Server

    65%

    43%

    M a

    jo ri

    ty

    S o

    p h

    is ti c a te

    d

    Sophisticated Tools like

    Hadoop

    SAS, SPSS and

    Strata

    2%

    11%

    Forensic /Audit

    Analytics like ACL

    and IDEA

    26%

    all rights reserved.

    EY 2015

  • Page 19

    Technology: the right tools for the right job

    Use of Statistical and data-mining techniques for

    predictive modelling, clustering and anomaly

    detection.

    Enable Continuous Monitoring and Real-time

    detection

    Highest Adopters

    Power and Utilities Industry stands out as the highest adoption rate of FDA tools used for continuous monitoring and predictive analytics

    41% Used for Continuous Auditing

    34% Text Analytics

    20% Statistical data Mining

    all rights reserved.

    EY 2015

  • Page 20

    Secure the buy-in, execute the build

    Improved results and recoveries,

    compared to other not using FDA

    Earlier Detection of misconduct,

    compared to other not using FDA

    More Cost-effective results,

    compared to other not using FDA

    Higher visibility to the board,

    compared to other not using FDA

    Technology counts

    11%

    15%

    14%

    12%

    !

    1

    5 Success Factors of FDA Integration

    Focus on the low-hanging fruit: the

    priority of the initial project matters

    2 Integrate more advances FDA technologies

    3 Communicate early and broadly

View more