Russian banks weren’t held to International Financial Reporting Standards until 2004. What happened before then?
“There was clear manipulation of the numbers going on,” said Steve Swidler, J. Stanley Mackin Professor of Finance at the Harbert College of Business. “Once they adopted the accounting standards, all of the sudden the numbers fell very much in line. You would assume, in the case of banks, they were manipulating numbers to meet capital requirements.”
Swidler and Denis Davydov, Professor of Accounting and Finance at the University of Vaasa (Finland), applied a forensic accounting tool, the Benford distribution, to Russian bank statements in their paper “Reading Russian Tea Leaves: Assessing the Quality of Bank Financial Statements with the Benford distribution” to better understand the banks’ financial reporting practices before and after the 2004 regulations.
“This was literally a lunchtime conversation when I was teaching in Finland (in 2012),” said Swidler. “One of the Ph.D. students (Davydov) was at lunch and I was talking about the Benford distribution. He happened to be Russian, and thought, ‘If I can get some bank data from Russia, we can look at it and see if it follows the Benford distribution.’”
The Benford distribution, also known as Benford’s Law, is a rule concerning the first digits of a set of numbers – for example, dollar amounts and populations. The principle is based the theory of Frank Benford, the late physicist who pointed out the pattern. If a dollar amount on a statement is said to be $1,361, then Benford applies the “1.” If a dollar amount is $437 on a statement, then Benford applies the “4.” According to Benford’s Law, “1” is the first digit of a number roughly 30 percent of the time, with subsequent digits decreasing in frequency.
Forensic accountants use the Benford distribution in screening statement data – searching for potential miscalculations or manipulations.
“If the distribution of the first digits deviates enough from the Benford distribution -- what we would call a statistically significant difference -- then we have reason to believe the accounting numbers were misstated or otherwise manipulated,” Swidler explained.
Swidler and Davydov studied income statements and balance sheets from 1,187 commercial Russian banks from 2001-2014 -- in years before and after the regulated accounting standards. Their findings?
“The Benford distribution results suggest that Russian banks tended to round revenues up, expenses down and thus overstate net income,” the paper states, regarding the pre-2004 statements. “The results also imply a reporting of stronger balance sheets than warranted with equity capital, perhaps, overstated.”
Though forensic accountants use the Benford distribution as a tool to detect fraud, the method isn’t completely foolproof.
“When you see the Benford distribution violated, it’s likely that something funny is going on,” Swidler said. “However, if the Benford distribution is not violated, it could still be possible that data is manipulated. It’s a sufficient condition, but it’s not necessary for there to be some type of fraud.”
The paper concludes the Benford distribution might best “be used with other early warning detection algorithms to discern financial distress in real time.”