At the June 2019 ACFE conference I spoke about Fraud Data Analytics (FDA): How to Locate Complex Vendor Overbilling Fraud Risk Statements. My sessions were sold out, so for those unable to hear my presentation I am writing three blogs that explain three of the four fraud risk statements I covered: Price inflation scheme: hidden entity scheme and now we will cover the pass-through scheme. The fourth fraud risk statement was explained in a blog posted on August 18, 2018.
Fraud Data Analytics: Uncovering Pass-Through Schemes Involving Real or Shell Companies
In my opinion, the pass-through fraud scheme is likely the most common fraud risk statement in the expenditure cycle. An internal person commits the fraud scenario, either alone or in collusion with an external party. Essentially, there is an illusion of an arm’s-length business transaction using a real company. The goods and services are received, and the internal three-way match is in full compliance. To further complicate the matter, the middle company, which is the pass-through company, may be a real company, but it could also be just a shell.
You can find out more about how to find shell companies in an article I wrote for ACFE Magazine March / April 2019.
There are over 20 variations of the scheme used in different ways depending on the industry and expenditure area. The focus of this blog is the following fraud risk statement:
A sales representative at a real supplier sets up a shell company and convinces the budget owner or senior member of management to purchase from the shell company rather than the real supplier. The budget owner places orders for goods through the shell company. The shell company then places an order with a real supplier. The real supplier ships directly to the budget owner company but invoices the shell company. The shell company invoices the budget owner at an inflated price causing the diversion of company funds. The budget owner receives a kickback from the sales representative.
The Fraud Data Analysis must start with the transactional data because there is no link to an internal person and the master file data does not link to another vendor. A missing data analysis may help - emphasis on “may.”
The starting point is to search for a change in the purchasing of a specific item from one vendor to another. Typically, the original vendor is a publicly known company whereas the new vendor privately-held company. There is usually a good business justification for changing vendors.
In this blog, I will discuss two approaches to searching for the preceding pass through scheme. The first approach is change analysis and the second is invoice number pattern.
First, create a high-level summarization report by general ledger and vendor listing the first invoice date and the last invoice date. The purpose of the report is not necessarily to find the pass-through scheme but to obtain an understanding of vendor change.
Your second attempts to find whether two or more vendors are providing the same item using a commodity code or a line item description. In this report, apply the exclusion theory and eliminate transactions in which the commodity code or line item is associated only with one vendor. Now, you have a database of line item transactions that are associated with two or more vendors.
Summarize the data file by line item and by vendor reporting the number of transactions and the aggregate dollar value for a two-year period. On each line item, report the last invoice date. Then search is for a line item that has significant dollars (relative concept) and for which there has been a change in vendors as evidenced by the last invoice date for both vendors. The pattern that will cause a sample selection is when the last vendor invoice number for one vendor is early in the scope period and the second vendor’s last invoice number is late in the scope period. The assumption is that the company stopped using one vendor and began using the other vendor.
Note that if the change occurred in your scope period, this analysis works. However, if the change occurred before your scope period or at the end of your scope period, the analysis would not reveal the scheme. It is always important to understand the strengths and weaknesses of your fraud data analytics.
Invoice Number Pattern
Next, we look to the key transactional data: invoice number, invoice date, invoice amount, and line item description. Keep in mind that vendors typically send invoices to multiple customers. In my experience with pass-through schemes, there is often an anomaly in the vendor invoice numbers. It could be a low starting number, a sequential pattern of invoice numbers or a limited range of invoice numbers.
Let me explain the limited range of invoice number analysis. Remember the data interpretation strategy? Well, this is a perfect example. The calculation is simple: identify the first invoice number and the last invoice number using the invoice date. Calculate the numeric difference between the two invoice numbers. This difference is the invoice number range, or the number of invoices a company has sent to all vendors. Now compare the number of records on hand to the range. Then next step is judgmental: does the invoice number range seem consistent with the perceived size of vendor or the vendor industry? To illustrate the concept:
The first invoice is dated 01/15/18 and is invoice number 10,100. The last invoice is dated 12/15/18 and is invoice number 10,400. Therefore, the range of invoice numbers for this vendor for the 2018 year is 300 invoices. The The number of invoice records in your database is 55. The company is in the business of renting excavation equipment to the construction industry. Here is the question: Does it seem reasonable that a vendor in the equipment rental business would only issue 300 invoices in a year?
In this blog, I have intentionally showed two different approaches to finding the pass-through fraud risk assessment. Why you ask? In searching for fraud schemes, you never know how the person will commit the scheme or conceal the scheme. But, if you know the scheme permutations, the concealment techniques, the data red flags, and the search routines, then you will find the fraud risk assessment -- if you have an unrelenting desire to find fraud. Do you?