Oversight using Data Analytics

Transcription

1 Oversight using Data Analytics MACPA Government and Not-for-Profit Conference April 17, 2015 Overview Data Analytics in Government Applications in Grant Oversight Applications in Purchase Card Oversight 1

2 DATA Act Greater Attention to Analytics in Government Promotes data sharing across government agencies Treasury data analytics center for OIGs automated oversight Government-wide structured data standards for financial reporting USASpending data should be standardized and machine-readable OIGs will audit data quality Improper Payments Elimination and Recovery Act (IPERA) Amends the Improper Payments Information Act of 2002 IPERIA strengthens estimations Strengthens detection, prevention, and recovery efforts Pre-award and pre-payment checks with Do Not Pay Annual risk assessments of covered programs Published improper payment estimates with reduction targets 3 Goal to reduce improper payments by $50B and recover $2B in 2 yrs Definition of Forensic Audit Forensic Audit 4 Audit that looks for financial misconduct, abusive or wasteful activity. Close coordination with investigators More than Computer Assisted Audit Techniques (CAATs) Forensic audit is growing in the Federal government GAO s Forensic Audit and Special Investigations (FSI) Recovery Board s Recovery Operations Center (ROC) Federal outlays are $3 trillion annually OMB estimates Federal improper payments at $110B GAGAS requires tests for fraud in audit work 100% review vs statistical sampling there is a place for both Automated Oversight Continuous monitoring Quick response 2

3 Automated Oversight More efficient use of resources 100% transaction review Still use traditional audit techniques to test transactions Improved risk identification Business rules based on risks Focus review on higher risks Recipients can use similar data analytics techniques Monitor grant spending Identify anomalies early Forensic Audit Approach Audit objectives and audit universe Work with investigations Structured brainstorming with subject matter experts Consider an SME or strategic planning conference Identify indicators of potential fraud and ways to find in data Process to identify financial risks capture all ideas Map out the end-to-end process Identify systems and key processes Identify key controls Identify and obtain transaction-level data Record layout 1000 record dump Modern analytics software can read virtually any data format Flat files, Delimited files, Dbase files, access databases, Report files,. No file size limits Build targeted business rules and run against data Examine anomalies 3

4 Different Levels of Knowledge Data Facts, numbers Information Summary Reports Dashboards and financial reports Knowledge Descriptive Analytics Statistics and automated business rules Wisdom Predictive Analytics Neural networks Decision trees Kohonen Networks Classification NSF OIG does not endorse commercial products Data Analytics Tools Data interrogation with automated business rules Data mining and predictive analytics Statistical analysis Link analysis Data conversion utilities Internet, open-source research Access to system query tools (e.g., SQL) 4

5 Framework for Data Analytics Using Government and Publicly Available Data Award level Data Grants, Contracts Transaction level Data Payee, Payment Amount, Date Award Systems Payment Systems Disbursing Systems Federal Reserve System Commercial Bank Data Download Data Analytics Oversight Review by Auditors Investigators Agencies Data Download Data Download Data Download Data Download Data Download Data Download Data Download Data Download Data Download GuideStar (non profits) SAM (includes EPLS) Federal Audit Clearinghouse Master Death File (SSA) Do Not Pay (Treasury) CPARS, FPDS Examples of systems that can help validate payment transactions Data Analytics Help with Oversight Determine reliability data fields Shape of the data (statistics) Completeness of transactions and fields Show anomalies. within a database between databases and changes in behavior over time Develop risk profiles for comparisons Awardee profiles Award-type profiles Program profiles 5

6 Common Data Analytics Tests Join Summarization Computed fields Corrupt data Blank fields (noteworthy if field is mandatory) Invalid dates Bounds testing Completeness Invalid codes Unreliable computed fields Illogical field relationships Trend analysis Duplicates Comparing Data Files (Three Bucket Theory) Vendors Not Paid Yet Vendors Paid and In Vendor Table Vendors Paid but not In Vendor Table Vendor Table (Valid vendors) Linking field allows for comparison Disbursing Transactions 6

7 Normal Distribution Anomalous Activity Normal Activity Anomalous Activity What is Data Mining? Refers to the use of machine learning and statistical analysis for the purpose of finding patterns in data sets. If you know exactly what you are looking for, use a structured query If you know only vaguely what you are looking for, turn to data mining. Most often used (up until recently) in marketing and customer analysis 7

8 Methods of Data Mining Supervised Modeling Predict patterns in data based on patterns of known information Decision Trees Neural Networks Unsupervised Modeling Identify anomalies or outliers based on grouping like transactions Kohonen Networks K-Means Clusters X 1 X 2 X 3 X 4 y 1 y 2 Building a Forensic Capability Develop Organizational Capability All audit staff should have basic skill with data analysis tools Forensic audit units perform more sophisticated analyses Targeted audits are more efficient in time and cost Phased Development Hardware and software Access to internal and external data Staff: system savvy, analytical, business process knowledge Training, then immediate application to work Very important component is tone at the top 8

9 U.S. Financial Assistance Overview $550 billion in awards 88,000 awardees and 26 Federal grant making agencies Project and research, block, and formula Outcomes are designed to promote public good Challenges Limited visibility of how Federal funds are spent by awardees Support for funding requests much less than for contracts American Recovery and Reinvestment Act (2009) $840 billion of assistance to stimulate the economy Greater accountability and transparency over spending than ever Opportunities to enhance oversight with less Automated oversight Framework for Grant Oversight Data analytics-driven, risk-based methodology to improve oversight Identify institutions that may not use Federal funds properly Techniques to surface questionable expenditures Life cycle approach to oversight Mapping of end-to-end process to identify controls 100% review of key financial and program information Focus attention to award and expenditure anomalies Complements traditional oversight approaches Techniques to review process and transactions are similar Transactions of questionable activities are targeted Recipients and Agency Officials can use data analytics Identify high risk activities through continuous monitoring 9

10 Grants Differ From Contracts GRANTS Promote services for the Public Good Merit review (competitive) Multiple awardees Award budget No government ownership Grant payments Summary drawdowns No invoices for claims Expenditures not easily visible Salary percentages CONTRACTS Specified deliverables (Goods and Services) Competitive process One awardee Contract Price Government ownership Contract payments Itemized payment requests Invoices to support claims Detailed costs Salary hourly rates End to End Process for Grant Oversight 6 PRE-AWARD RISKS Funding Over Time Conflict of Interest False Statements False Certifications Duplicate Funding Inflated Budgets Candidate Suspended/Debarred ACTIVE AWARD RISKS Unallowable, Unallocable, Unreasonable Costs Inadequate Documentation General Ledger Differs from Draw Amount Burn Rate No /Late/Inadequate Reports Sub-awards, Consultants, Contracts Duplicate Payments Excess Cash on Hand/Cost transfers Unreported Program Income AWARD END RISKS No /Late Final Reports Cost Transfers Spend-out Financial Adjustments Unmet Cost Share 10

11 Common Audit Findings 21 Data Analytics Audits Unallowable, unallocable, unreasonable costs Excess salary 2-month rule Indirect Costs Equipment Pre-Data Analytics Audits Unsupported costs Effort reporting Effort reporting (subaward) Pre-award charges Burn Rate Actual vs Expected Actual Expected Award Amount ($K) Expended ($K) %Expend Award Days Days Active % Total Days Delta 1 10,000 9,000 90% % ,000 4,000 80% % ,000 1,500 75% % , % % ,000 12,000 60% % ,000 5,000 50% % 0.57 Awarde e Totals 48,000 32,495 68% 7,302 3,997 55% would be normal 11

12 Risk Identification General risks Certain contract and grant awards tend to be riskier than others Smaller institutions tend to have weaker internal controls Specific risks 23 Something that happens in a process that stands out from normal activity Large drawdown on a single date end of a fiscal year Spending out remaining grant and contract funds at end of the award Challenges General risks can be more obvious Specific risks can be harder to see. Benefits greatly from transaction level data. Focus on Risk Many to the Few 600,000 Grant award drawdowns annually totaling $6.3 billion Each assigned a risk score 40,000 Active awards Each assigned a risk score 2,000 Institutions Each assigned a risk score 20 Audits of higher risk institutions Each audit tests all transactions for all awards with automated risk indicators 12

13 Risk Assessment and Identification of Questionable Transactions Phase I Identify High Risk Institutions Phase II Identify Questionable Expenditures Agency Award Data Award proposals Quarterly expense reports Cash draw downs Agency Award Data Award proposals Quarterly expense reports Cash draw downs Awardee Transaction Data General ledger Subsidiary ledgers Subaward data Data Analytics Continuous monitoring of grant awards and recipients Data Analytics Apply risk indicators to GL data and compare to Agency data Review Questionable Transactions External Data A-133 audits (FAC) SAM (CCR, EPLS) External Data A-133 audits (FAC) SAM (CCR, EPLS) Identification of Higher Risk Institutions and Transactions 13

14 Anomalous Drawdown Patterns $$ Start up costs Drawdown Spike Extinguishing Remaining Grant funds (before expiration) Extinguishing Remaining Grant funds (after expiration) Normal drawdown pattern Grant Award Grant Expiration 27 Early Drawdown 14

15 Spend out Pattern Draw Spike 15

16 A Common Drawdown Pattern Payroll Charges Incurred Subsequent to Grant Expiration 32 NSF OIG does not endorse any commercial software 16

17 Transfer of Payroll Charges 6 Months After the Grant Expired Equipment Charges Incurred Immediately Before Grant Expiration Same day as expiration 57 days before expiration 46 days before expiration 17

18 Cost Transfers after Grant Expiration 35 Travel Related to Award? Just before award expiration Just after award expiration 18

19 Purchase Card Oversight using Data Analytics 37 Government purchase card overview Simplified acquisition Still high risk for abuse without strong oversight Government Credit Card Fraud Prevention Act 2013 DoD Joint Purchase Card Review Current work at NSF Review objective DoD Joint Purchase Card Review 38 Identify purchase card abuses and recommend process improvement Universe under review 15 million purchase card transactions ($9 billion) 200,000 cardholders (CH) and 40,000 authorizing officials (AO) 300 DoDIG and Defense agency auditors/investigators Subject Matter Expert conferences Structured brainstorming with auditors, investigators, GSA officials Developed 115 indicators of potential fraud 46 codable Build targeted business rules and run against data Field research, reporting, and process improvements $122M in recoveries, 100 prosecutions, 275 adverse actions Most important outcome: indicators built into bank systems 19

20 Top Indicator Combinations 97% Adult websites, Weekend/Holidays 67% Purchases from 1 vendor, CH=AO 57% Adult websites 57% Internet transactions, 3rd party billing 53% Interesting vendors, many transactions 43% Even dollars, near limit, same vendor, vendor business w/few CHs 20

21 Current NSF Purchase Card Work Similar approach as DoD Joint Purchase Card Review Universe 3 years of purchase card activity 230 card holders 34,000 transactions $17 million Purchase card transaction data from the bank s website Worked closely with Investigations Developed risk indicators at transaction level Risk-based approach to testing 41 Risk Factor Examples 42 AO Span Of Control >4 Flags transactions for Carholders (CH) whose Approving official has a span of control of 5 or more CHs. (Risk value = 1) Suspect MCC Codes Flags transactions with MCC codes we deemed suspect (Risk value = 2) Blocked MCC Codes Flags transactions with Blocked MCC codes. (Risk value = 3) Holiday Purchases Flags transactions that occurred on holidays (Risk value = 3) Weekend Purchases Flags transactions that occurred on the weekends (i.e., Saturday or Sunday) (Risk value = 3) 21

22 Risk Factor Examples (continued) 43 Suspect Level 3 Data Flags transactions with Level 3 data we deemed suspect based on manual review (Risk value = 3). For example, possible personal purchase, possible split transaction, questionable legitimate business need. One to One Card Holder to Merchant Flags transactions in which the merchant only did business with that particular NSF card holder (Risk value = 2). Possible Split Purchase Flags transactions by a card holder in which more than 1 purchase to the same merchant totaling more than $3,000 occurred on the same day, or within a few days (Risk value = 3). Example of Level 3 Data 22

23 Questions? Dr. Brett M. Baker Assistant Inspector General for Audit National Science Foundation Office of Inspector General Phone: