Attention aux managers trop confiants ! le risque de fraude n'est pas loin.

C'est le thème d'une étude réalisée par Catherine M. Schrand et Sarah L. C. Zechman qui s'intitule "Executive Overconfidence and the Slippery Slope to Fraud".

Par nature, l'homme est optimiste. Witney Tison écrivait en 1999 :

"But humans are not just robustly confident-they are wildly overconfident.

Consider the following:

- 82% of people say they are in the top 30% of safe drivers;
- 86% of my Harvard Business School classmates say they are better looking than their classmates (would you expect anything less from Harvard graduates?);
- 68% of lawyers in civil cases believe that their side will prevail;
- Doctors consistently overestimate their ability to detect certain diseases (think about this one the next time you're wondering whether to get a second opinion);
- 81% of new business owners think their business has at least a 70% chance of success, but only 39% think any business like theirs would be likely to succeed;
- Graduate students were asked to estimate the time it would take them to finish their thesis under three scenarios: best case, expected, and worst case. The average guesses were 27.4 days, 33.9 days, and 48.6 days, respectively. The actual average turned out to be 55.5 days.
- Mutual fund managers, analysts, and business executives at a conference were asked to write down how much money they would have at retirement and how much the average person in the room would have. The average figures were $5 million and $2.6 million, respectively. The professor who asked the question said that, regardless of the audience, the ratio is always approximately 2:1.

Importantly, it turns out that the more difficult the question/task (such as predicting the future of a company or the price of a stock), the greater the degree of overconfidence. And professional investors -- so-called "experts" -- are generally even more prone to overconfidence than novices because they have theories and models that they tend to overweight.

Perhaps more surprising than the degree of overconfidence itself is that overconfidence doesn't seem to decline over time. After all, one would think that experience would lead people to become more realistic about their capabilities, especially in an area such as investing, where results can be calculated precisely. Part of the explanation is that people often forget failures and, even if they don't, tend to focus primarily on the future, not the past. But the main reason is that people generally remember failures very differently from successes. Successes were due to one's own wisdom and ability, while failures were due to forces beyond one's control. Thus, people believe that with a little better luck or fine-tuning, the outcome will
be much better next time."

Certains sont plus optimistes que la moyenne, et c'est le cas des dirigeants d'entreprise. Cet état d'esprit qui est essentiel au dynamisme d'un projet stratégique peut cependant avoir quelques inconvénients. La finance comportementale les a étudié depuis longtemps. Voir par exemple "managerial overconfidence and corporate policies" de ItzhakBen-David, John R. Graham, et Campbell Harvey, (November 2007), ou encore "behavioral corporate finance : a survey " de Baker, Ruback et Wurgler qui font un point très complet sur les biais comportementaux des dirigeants.

L'étude de Schrand et Zechman analyse les circonstances dans lesquelles une fraude apparait à non pas à la suite d'une volonté délibérée et perverse du dirigeant, mais tout simplement parce que ce dernier pense qu'il va pouvoir rétablir une situation difficile avant que le comportement frauduleux qui visait à la masquer ne soit découvert. Quelques semaines après l'affaire de la Société Générale, cette thèse nous semble bien familière ...