Discussion of "Digital analysis and the reduction of auditor litigation risk";

              Discussion of "Digital Analysis and the Reduction of Auditor Litigation Risk"
James E. Searing
Partner, Ernst & Young LLP
Mark Nigrini's paper, Digital Analysis and the Reduction of Auditor Litigation Risk, is a leap forward for the audit profession, and I believe that this is an important contribution as we increasingly invest in information systems auditing. So any comments that I have on the paper should not be construed as arguing with the value of digital analysis. In fact, I advocate that future research be directed to finding more ways to analyze data and detect the anomalies which might indicate the existence of material fraud, misuse of corporate funds, errors, or misstatements.
Digital Analysis Overview
The essential point of this paper is that digital analysis, used in conjunction with an audit, will reduce an auditor's risk of litigation. Let's look at an example.
Frank Benford discovered that there are natural and predictable patterns when one looks at how frequently the numbers appear in large data sets. One of the more significant observations that he made involved the first two digits. Take a simple data set as shown below:
$15,987.90 3,894.94 .73 100.00
Taking the first two digits of each number (respectively, 15, 38, 73, 10) we can see that each of these digits are different. But Benford, and others mentioned in the paper, found predictable patterns in large sets of data. They found, for example, that there are more 10's than 99's. And more 11's than 98's; more 12's than 97's, and so on. As they looked at more and more data, they found that, in absence of some human intervention, the data should conform to a given pattern.
The author has taken this work further by proposing 13 methods of examining data, including tests for rounding, estimation, and conformance with known patterns. He has expanded the concept and presented it effectively so that those of us in the business of auditing can apply it.
Observations from Practice
Ernst & Young has used digital analysis and found that it does, at times, lead to some very interesting conclusions. Examples include:
* Frauds, usually low level ones, are sometimes perpetrated through a stream of transactions. Some perpetrators accomplish this by using a recurring pattern of transactions. Such a pattern may be detected through digital analysis and subsequent follow up on the transaction.
* Operational inefficiency: Considering the cost of issuing one check, our clients may not be aware of ineffective procedures which entail the issuance of many thousands of extra checks or other paperwork. (Example: One company issued 20,000 checks for amounts equal to $25, $50 and $75 as they matched employee contributions to charity. Would it have been more efficient to accumulate such checks? Companies might devise a different approach, particularly when the cost of issuing a check can be more than $10.00 per transaction.
83