Monday, February 22, 2016

How a Software Can Predict Financial Performance by Detecting a CEO's Emotions

Computer programs have been created to scan facial expressions that help determine how people respond to commercials, or whether or not hospital patients are in pain. James Cicon, a finance professor at the University of Central Missouri, constructed a software that he used to help analyze the faces of Fortune 500 executives. Cicon and his software analyze the executives for signs of different emotions such as fear, anger, disgust, and surprise which he found had a correlation to critical data. According to The Wall Street Journal, “The emotions related to the companies’ profit margins, return on assets, stock price moves, and other measures of performance” (Dwoskin).
            Dr. Cicon and his colleagues scanned the faces of more than 200 CEOs and former CEOs of companies including Citigroup, ConAgra Foods, FedEx, and Macys. They determined that emotions such as fear, anger, disgust—even though they are negative—correlate positively with financial performance. The results show, “Signals of disgust as shown by lowered eyebrows and eyes, and a closed, tightened mouth, were associated with a 9.3% boost in overall profits in the ensuing quarter” (Dwoskin). Additionally, “Dr. Cicon found that CEO’s who expressed signs of fear with raised eyebrows, widened eyes and mouth, and lips that were pulled in at the corners saw their companies’ stock price rise .4% in the following week” (Dwoskin).  Cicon related his finding to psychological research done by Steve Ferris and Ali Akansu that states, “Fear is widely recognized as a powerful motivator.” Cicon finds that CEO’s that appear fearful under interrogation will work harder to increase their firm’s value.
            A software such as the one that Dr. Cicon created allows investors to better predict the market and the future of their investments. This software has the potential to change the way people analyze firms in the market and assess their future earnings and growth. Financial and market analysts are constantly looking to improve their predictions and enable a higher percent of accuracy to limit the amount of volatility and instability of the market place.  A software such as this also has the potential to more accurately understand what CEOs are actually thinking which will have an effect of the market value of a firm. For instance, Dr. Cicon is hoping to teach the software to measure degrees of sincerity and truthfulness to help determine the validity of executive’s statements.
            Conversely, there are many faults and future problems that can come as a result of people utilizing a software such as the one Dr. Cicon created. Dr. Ekman, the psychologist whose research to analyze emotions based on facial expressions, that Dr. Cicon traced his software back to claimed 90% accuracy for his emotion-coding system. This makes accuracy a concern for Cicon’s software and the potential for the results to be misused. Dr. Ekman suggested, “A misinterpreted emotion on a CEO’s face, for instance, might contribute to market swings that cause a company to lose or gain value based on spurious information.” Personally, I think that facial expressions could be misleading and if technology like this emerged, CEOs would find ways to hide their true feelings in a way that can go undetected by the software.
Dwoskin, Elizabeth. "Software Detects CEO Emotions, Predicts Financial Performance." Digits    RSS. Wall Street Journal, 17 Feb. 2016. Web. 22 Feb. 2016.       <>.


  1. As a finance major I found this post particularly interesting. It would seem an attractive idea to try to predict stock value by facial analytics of CEOs, but there are many notable potential downfalls. The first and most important was mentioned in the original post. If this technology does become widespread, CEOs will find away to either trick the technology into thinking their companies are doing better or learn how to mask emotions. I do not think it would be wise to plan on using this tactic in the long term. Also, I question the accuracy of the initial survey. Fear leading to a 9.3% boost in overall profits seems very counterintuitive. Wouldn’t a CEO whose company is on the verge of bankruptcy be showing fear? Different people demonstrate emotion in a variety of ways so I wonder if the sample size of fortune 500 CEOs is large enough to really be conclusive. Also, emotions derived from business performance likely very from culture to culture. It could be that this works for certain American CEO’s, but I doubt it can be applied universally. Another difficulty I find is getting the facial data for analytics. How do they know the fear, anxiety, or whatever emotion comes from business performance? There are many other confounding variables such as family or health issues that could throw off the data. Would they require a new picture to analyze of the CEO on a daily basis? That could be difficult to get, specially a picture where there emotion relates to the company’s performance. It is solid idea to attempt to predict stock value from CEO emotion, but until there is more concrete evidence, at the minimum dollar amounts of how much they profited, I would refrain from using this technology.

  2. I’d be interested in seeing this technology utilized for how truthful executives are being when their activity is under question. For example, had this technology been available before the creation of the Sarbanes-Oxley Act amidst all of the creative accounting scandals at companies such as Enron, World Com, AIG, etc. maybe there would have been more convictions instead of everyone just playing dumb and shifting the blame onto someone else. As far as predicting market prices, I think the technology could be reading someone’s mood and facial expressions based on how good of a night’s sleep they got the previous night rather than deciphering some veil of impending success or failure. There is already enough pressure on management in companies to outperform previous quarters, which is what contributed to many of the afore mentioned scandals. Imagine a world where part of your job description is giving off the right facial expressions while being subjected to facial reading technology or else your company would be devalued? It seems a little too invasive to me, but I think it could be useful in a courtroom setting where the truth is needed and can easily be misconstrued by executives playing the “I didn’t know what was happening” card.


Note: Only a member of this blog may post a comment.