How text fingerprints can predict the collapse of a company


DATA’s ability to spot deceptions faster than even highly focused investors and regulators could be due to the fact that 86.5% of financial information is expressed in text rather than numbers. While the vast majority of the market may believe that numbers in financial statements accurately price in all relevant information, the reality is that badly behaved companies have limited ability to maintain the appearance of good financial performance by flattering their numbers .

Voss highlighted the AIG and Lehman Brothers cases, hidden from the public for the longest period of all the scandals in the analysis. Both financial firms, the two ill-fated corporations, published annual reports that ran into hundreds of pages, with the money flowing on their balance sheets and income statements at tremendous rates. This helped ensure that their bad behaviors and choices weren’t reflected in the numbers for more than 10 years.

Voss went beyond the 10 scandals and pointed out recent controversies like the Wirecard affair. He pointed out five text fingerprints that differentiated misleading companies from their more truthful counterparts.

  • Words Indicating Friendship – Based on the analysis, misleading companies use terms that imply friendship that is 56.1% above the average expected from annual reports.
  • Risky Words – While companies generally proactively cut words like “unwilling,” “avoid,” “worried” and “difficult” from their annual reports as they tend to flag red flags with securities researchers, scandal companies prefer these types of words with a higher rate;
  • Impersonal Pronouns – Compared to truthful colleagues, scandal companies use impersonal pronouns such as “another”, “everyone”, “someone” and “depending” 54.1% more often, presumably to create emotional distance between them and the readers who use them want to mislead ;;
  • Words That Point Out Differences – Because of the cognitive burden of lying, those who do this are less able to distinguish between competing viewpoints, so they can make less than average comparisons. Also, having a preferred narrative to accomplish their goals, scammers are reluctant to emphasize distinctions with others;
  • Words that negate a statement – Research has also shown that liars are 50.4% more likely to use words like “not”, “never”, “shouldn’t”, “not” and “must not” than truthful companies .

Aside from the five fingerprints, Voss has highlighted one indicator that is by far the strongest indicator of corporate fraud:


Please enter your comment!
Please enter your name here