The SEC recently issued a long concept release seeking public comment on ways to modernize Regulation S-K, the set of disclosure requirements used both for Securities Act registration statements like Form S-1 and Exchange Act reports like Form 10-K. (As a side note, the term “concept release” invariably brings to my mind concept albums by bands like Pink Floyd.) Given my (self-imposed) limit on the length of my blog posts, I will confine my discussion of the release for now to just one point: the SEC’s solicitation of comment on the suggestion that companies provide, in the Risk Factors section, estimated probabilities of the relevant event occurring and the magnitude of the effect on the company if it does occur.
On one hand, it is important to analyze risks in a probabilistic manner. In the excellent Phillip Tetlock book, Superforecasting, the author notes that those who are best at predicting future events tend to think in probabilistic terms, assigning very specific percentages to the chance of a particular event occurring, based on all available information. Forecasters like Nate Silver get flak if they say a candidate has, say, an 80% chance of winning the election. He’s not “wrong” if the underdog wins, since if you ran the election 100 times (which can’t practically happen), the underdog would win 20 times.
But applying a probabilistic approach to the Risk Factors section is problematic from the company’s perspective. Suppose a company assigns a 1% likelihood of a particular risk coming to pass, and it considers even that figure high, though it has to be mentioned because of the large magnitude of the risk’s potential effect on the company. And then further suppose the event happens, leading to inevitable shareholder litigation. If the judge or jury assessing the issue has trouble thinking in a probabilistic manner – as most people do – they will take the company’s assignment of a 1% chance to be, in essence, a statement that the event will not occur, so the company is (unfairly) found to have misrepresented the risk. There would be little or no downside, at least from a liability perspective, for companies to artificially overstate the chance of the risk, though non-legal personnel from the company dealing with the disclosure will not want to scare off investors with high percentages.