Assembling a database of independently sourced transaction evidence is critical for any Fraudulent Transfer or Financial Litigation. Gregory Hays, CTP, CIRA, Past President NAFER, published some guidance as a chapter of a publication called Fraud and Forensics, Piercing Through The Deception of a Commercial Fraud Case. This blog contains a few excerpts from the book along with VALID8 recommended best practices.
“Building an accurate and complete database of all relevant transactions is critical in complex fraudulent transfer litigation…
…The process of building a comprehensive database of all known transactions begins with collecting pertinent financial and non-financial documents. After the identification of relevant accounts, data from available bank, credit, and investment statements and other pertinent sources is verified and used to develop a database of all transactions uncovered by the forensic accountant.”
– Fraud and Forensics, Piercing Through The Deception of a Commercial Fraud Case, American Institute of Bankruptcy, 2015, Gregory Hays, CTP, CIRA, Past President NAFER, Chapter 5, pg 59
Beyond just getting the data, it has to be 100% accurate, so Greg goes on to describe the criticality of reconciling every transaction for every statement for every account. Often times this work uncovers more leads, questions and information to follow up on.
An accurate transaction database is a must have first step, but because of the possibility of litigation and the standards of evidence for US DoJ, below are five best practice capabilities that any transaction database should have.
- Multiple Methods for Ingesting Transaction Data: Any case involving financial litigation or a fraud investigation involves a subpoena to acquire banking statements. At best, they’ll be native PDFs from the institution, at worst, it’s a bunch of boxes of paper that will need to be scanned, your database solution should be able to handle both. And for those cases associated with operating companies where ongoing monitoring is required, the ability to directly link up to bank accounts from various institutions and get a live account feed is crucial and can save dozens of hours per month.
- Automated Reconciliation and Data Quality Control: Most of the time, professionals will be forced to deal with PDF statements. Today, best practice is to use OCR software to extract the transactions but no OCR program is perfect. Tiny errors like a 6 is extracted as a 9 or one digit in the year is wrong can wreak havoc on the data integrity and require hours of excruciatingly detailed detective work to find the specific errors. Solutions should have auto-reconciliation features and tools to help professionals find and fix the OCR errors as they are inevitable.
- Fully Indexed and Linked Source Documents: Any record in the transaction database should be hyperlinked back to the original source document to avoid the painstaking process of assembling all of the source documents into the final work papers. It requires a programmatic integration between the OCR engine and the transaction database.
- Centralized Audit Trail for Full Custody and Control: Any changes to the data from the first time a transaction record is loaded into the database until the case is closed, need to be documented. Who did what and when needs to be comprehensively documented to ensure chain of custody of all data.
- Use Visualizations to Trace Funds: Oftentimes cases require the tracing of funds from sources into and out of various accounts to some ultimate use. This is a very difficult task to perform with a standard spreadsheet or database. To simplify case exhibits, use interactive visualizations to avoid presenting multiple complex tables, charts and graphs.
Building databases isn’t a core expertise for forensic accountants and professionals providing litigation support services for fraudulent transfer and preference payment work. New fully integrated, end to end solutions are available to simplify this process and make best practice analysis cost effective for any investigation, no matter how small.