18 Conclusion


THIS BOOK TAKES THE READER from the beginning to the end of applying data analytics with practical uses in mind. The first two chapters introduce fraud and fraud detection. The third chapter is a critical one to read before commencing any data analytics, as it details how to obtain good data for use with data analytical software. Chapter 4 provides the reader with a basic understanding of statistics. Statistical theory is the underlying basis of many analytical tests. Chapter 5 and onward details general data analytical tests and specific tests for various fraud scheme areas. Chapter 17 shows the reader how to automate procedures to simplify the data analytical process.

To conclude the book, we will take a look at financial statement fraud and learn why this important area of fraud was omitted from this book. We will also review a listing of features demonstrated with the IDEA software and end this book with a few additional words about data analytics.


Financial statements are summarized transaction records obtained from detailed accounting records that are used to provide information about the financial position of the business entity. The main users of financial statements include:

  • Prospective investors to determine whether to invest.
  • Owners, managers, and shareholders who make business decisions.
  • Lending institutions when deciding to lend or continue to lend funds.
  • Vendors when establishing a business relationship.
  • Government entities such as the tax authorities.

Since financial statements are key documents used in making significant decisions, financial statement fraud impacts users of these statements at great costs. Due to the magnitude of losses from financial statement fraud, this is a hot media topic. Many studies have been made on financial statement fraud and are included in most fraud literature. The studies have found that the frauds are committed by those who have the greatest motive and receive the greatest benefits; that is, senior management. The fraud can be committed by using the accounting system to generate the desired results, to provide falsified information into the accounting system, or merely by creating falsified financial statements. Financial statement fraud schemes include:

  • Enhancing revenue with fictitious sales.
  • Omitting expenses and liabilities.
  • Providing incorrect valuations for assets.
  • Manipulating results by using timing differences.
  • Improper disclosures by addition or exclusion.

At the transactional level, data analytics may detect some of the categories of financial statement schemes by using the tests described in earlier chapters of this book. However, the most significant fraud items impacting financial statements would not likely be in the detailed transactions, but rather in journal entries or outside of the normal business transactions. As such, general tests to detect anomalies within the data may not be the best place to start. Certainly, if there was a specific area targeted for investigation, data analytics can assist with the verification of the details summarized on the financial statements.

A good starting point for testing for financial statement fraud would be from the financial statements themselves. This fundamental analysis can be done by calculating and interpreting both vertical and horizontal ratios. The categories of ratios that can be performed include:

  • Profitability
  • Debt
  • Operating
  • Liquidity
  • Cash flow
  • Investment valuation

Data analysis software is not suitable for these types of calculations and analysis. Spreadsheet software such as Excel would be more appropriate. The formula for the calculations of the ratios within each of the categories can be found in most auditing books and by Internet searches. A website that provides this is www.caclubindia.com/articles/analytical-review-3850.asp.

Before calculating the individual ratios, it would be beneficial to employ the Beneish model test. This test consists of eight financial ratio variables to calculate a score value. The score reflects the degree to which earnings may have been manipulated. If the score is greater than –2.22, there are potential earnings manipulation. The Beneish ratios along with an Excel spreadsheet to perform the calculations for you can be found at investexcel.net/beneish-m-score.


Listings of the IDEA features outlined in this book, along with equations with @Functions, are summarized here. It is the author’s hope that if some of the features, equations, and functions are new to the reader, exposing the reader to them opens up additional potential uses for data analytics. IDEA features demonstrated include: