It is always easier to look back on a story and rewrite it in the light of how it ended. Still, many major financial frauds could have been detected earlier. And with little effort.
Enron: the energy giant that fooled the whole world. Behind the image of a visionary company, the numbers told a very different story. The fraud was everywhere, sophisticated, and yet a simple mathematical look at the accounts could have given the game away.
It all starts with a journalist. In March 2001, Bethany McLean of Fortune magazine asks a question that should have alarmed all of Wall Street: “Is Enron Overpriced?” At the time, analysts are still overwhelmingly rating the stock a buy. Yet McLean points to an opaque business model, massive debt and profits that seem to come out of nowhere. Enron’s executives respond evasively. Too evasively.
What nobody sees yet, or wants to see, is that the empire rests on a mountain of shell companies. More than 2,000 offshore entities, used to hide losses and artificially inflate profits. An accounting game that turns debts into assets and passes bank loans off as revenue. Andrew Fastow, the chief financial officer, orchestrates these structures with breathtaking nerve.
The mechanism?
Another sleight of hand: mark-to-market accounting, which here meant booking today future profits supposedly earned over the life of long-term contracts. Never mind if, in reality, those profits would never materialise. All that mattered was the immediate effect on the share price.
And yet Benford’s law could, and should, have raised the alarm. A forensic analysis of Enron’s figures showed that the reported revenues did not follow the normal first-digit distribution. Too many rounded figures, with a marked tendency for earnings to be adjusted to artificial steps ($0.10, $0.20, etc.), a clear sign of deliberate manipulation. Enron also manipulated its net income, a line that showed a strong deviation from Benford in the run-up to its bankruptcy.
In January 2002, Arthur Andersen, the prestigious audit firm that signed off on Enron, and which had been using Benford for 15 years, hastily destroys 1.7 tonnes of documents.
But the truth is written in the numbers.
When the bubble bursts, it is already too late. The collapse is brutal. The executives had sold their shares by the million before the bankruptcy: Kenneth Lay, Jeffrey Skilling, Fastow himself. They knew. The employees lost everything. Their pensions, their savings, their faith in the system.
The Japanese electronics giant admitted in 2015 to overstating its profits by more than $1.2bn over seven years. The investigation found that the company had been massaging its accounts since 2008, booking future revenue early and hiding costs.
Applying Benford showed that the accounting anomalies were detectable well before the scandal broke13: Toshiba’s financial statements between 2008 and 2014 showed an abnormal over-representation of the digits 5 and 6 in first position, a sign of manipulation. No such distortion was seen at any other major Japanese conglomerate at the time, which could have prompted a closer investigation.
Above all, these gaps appear from 2008, which means an audit using this method could have picked up warning signs seven years before the official investigation was opened. And that Benford made the fraud visible from the very first year.
Yet it was only after Toshiba’s share price collapsed in May 2015 that auditors and the authorities started scrutinising these anomalies.
This massive fraud led to the resignation of chief executive Hisao Tanaka and several senior executives, and forced a complete restructuring of the company, with more than 8,000 job cuts and the sale of some strategic divisions.
Valeant Pharmaceuticals, often compared to Enron for its use of satellite entities to artificially inflate its revenue, saw its share price fall by 90% after the SEC opened an investigation in 2015.
A study14 by Deutsche Bank applied Benford’s law to Valeant’s financial statements for 2013 and 2014, and placed the company among the 5% most likely to have accounting irregularities of all the companies analysed
Yet the model was never used preventively. The SEC only opened its investigation in 2015, and it took until 2020 for Valeant (renamed Bausch Health) to be formally charged over its accounting.
The investigation found that Valeant used specialty pharmacies such as Philidor to book fictitious sales and artificially inflate its revenue. In 2014, more than 10% of Valeant’s revenue came from Philidor, which allowed the company to hide falling sales in other segments15. The company also pushed through spectacular price rises on essential drugs, sometimes more than 500% in a year, to offset the cost of its aggressive acquisitions.
After the SEC’s announcements, Valeant’s market value went from $90bn in August 2015 to under $10bn in 2017. The company was forced to sell several subsidiaries to avoid bankruptcy and had to pay a $45M penalty to the SEC in 202016 for manipulating its financial statements.
The Libor scandal of 2012 revealed how traders at several big banks had manipulated interbank interest rates to their advantage. For years, they submitted artificially adjusted rates to maximise their profits.
Yet as early as 2005, Benford could highlight significant statistical anomalies17, suggesting manipulation had started much earlier than first suspected. A study covering 2005-2008 showed that the submitted rates deviated systematically from Benford’s distribution, confirming a recurring pattern of manipulation.
The main gaps observed:
Seven years earlier, it would have been possible to uncover the fraud18, which led to massive penalties:
Before joining the euro, Greece falsified its public accounts to meet the European Union’s membership criteria. With help from Goldman Sachs, the country used complex currency swaps to hide the true size of its debt. A retrospective analysis19 of Greek budget figures revealed significant deviations from Benford’s law.
Although the fraud was established in 2010, significant gaps could have been detected in 2001 and 200920, the years Greece joined the euro and revised its budget statistics, respectively.
More broadly, a study of the economic data of 27 EU countries reportedly confirmed that Greece showed the largest deviations from Benford’s distribution, which should have raised the alarm21. All the more so as the fraud was visible across 130 macroeconomic indicators a year over 11 years22 and concerned only the Greek figures23.
When the truth comes out in 2010, Greece is on the brink of bankruptcy, triggering a major crisis in the eurozone. Eurostat’s revision of the Greek accounts confirmed that the deficit had been understated by several points of GDP, setting off a loss of market confidence and forcing a succession of bailouts24.
Although the doctoring of Greece’s accounts was an open secret among European institutions, systematically applying Benford to budget monitoring could have made these anomalies more visible and undeniable long before the country collapsed25.