– Computer-aided text-based analysis can help uncover fraud and deception in company communications.
– The average lead time between identifying deception in company communications and public recognition of a scandal is more than six years.
– Language analysis can be used to identify potential scandal indicators in company communications.
– Five textual fingerprints that differ from truthful companies by more than 50% are: words indicating friendship, words indicating risk, impersonal pronouns, words that indicate differences, and words that negate a statement.
– Scandal companies also tend to use more swear words in their annual reports.
Last month, the concept of using computer-aided text-based analysis to detect fraud and deception in company communications was discussed. However, there is more to be learned from this research in the context of scandal companies.
Deception And Truth Analysis (D.A.T.A.) was used to examine ten of the largest corporate scandals in recent history. It was found that on average, there was a lead time of more than six years between the textual identification of deception and the public recognition of a potential scandal.
The question that arises is why it takes regulators and markets so long to recognize these scandals. Additionally, what insights can be gained from text-based analysis to identify these scandals earlier?
A theory was developed, suggesting that the behavior of corporate management teams, as revealed in text communications, eventually leads to the numerical performance expressed in annual reports. Essentially, poorly behaving firms are able to deceive for an average of 6.6 years until they can no longer manipulate the numbers.
Interestingly, the two scandals that took over a decade to recognize were both related to financial firms: AIG and Lehman Brothers. These firms had extensive annual reports and high money circulation, resulting in a longer time for their poor behaviors to be reflected in the numeric outputs.
If this theory holds true, language fingerprints can be used as early warning signs or additional indicators in investment research. Five textual fingerprints were identified that were significantly different in scandal companies compared to truthful companies: words indicating friendship, words indicating risk, impersonal pronouns, words that indicate differences, and words that negate statements.
Lie detection researchers have found that deceivers often use words indicating friendship more frequently than the norm in business communications. Scandal companies were found to employ such terms 56.1% more often than the average. However, there is a distinction between words indicating friendship and words that are friendly.
Scandal companies also favored words that indicate risk at a much higher proportion than the average company. These include terms such as “averse,” “avoid,” “concern,” and “difficulty.” These words are already considered red flags by securities researchers, and companies are actively removing them from their annual reports.
Impersonal pronouns, such as “another,” “everybody,” “someone,” and “whichever,” were used by dishonest firms 54.1% more often than truthful firms. This may be an attempt to create emotional distance between the deceivers and their targets.
Deceivers often struggle to make distinctions among competing points of view and are less likely to draw comparisons. Hence, the use of words indicating differences is seen as an indication of truthfulness. In contrast, deceivers tend to focus on convincing their target to believe their preferred narrative.
Research has shown that liars often use more negative terms than truth tellers. Scandal companies were found to utilize words like “not,” “never,” and “must not” at a 50.4% greater proportion than the average.
Additionally, it was found that the presence of swear words in annual reports is a strong indicator of deception. Scandal companies had swear words occurring 277.1% more frequently than the mean.
Overall, text-based analysis and language fingerprints can provide valuable insights into detecting potential fraud and deception in corporate communications. By analyzing the language used, investors and regulators can potentially identify early warning signs and assess the truthfulness of company statements.
