Beatrice Michaeli’s research suggests managers, forced to inform a broader audience, choose not to gather information even for themselves
In 2000, the Securities and Exchange Commission issued Regulation Fair Disclosure, which required public companies to make material information available to all investors at the same time. In adopting the rule, known as Reg FD, the SEC signaled that equal disclosure was better for the investing public than the older practice of allowing managers to selectively disseminate information to favored institutional investors and analysts before they released it to the general public.
Critics of the rule have been saying for years that it could have a chilling effect, reducing the quantity and quality of the information companies choose to make available. A recent paper, published in The Accounting Review by UCLA Anderson’s Beatrice Michaeli, lends some theoretical support to those concerns.
Michaeli developed a mathematical model to examine what happens when a manager decides to limit information to a subset of possible recipients. The model suggests that if managers are allowed to limit their audience, recipients as a whole will end up with more and better quality information than if managers are required to release information to everyone. This is especially true when the manager and recipients have different objectives — just the sort of conflict that rules like Reg FD are designed to address.
“Ironically, under this scenario, regulations forcing equal access will promote fairness but at the expense of reduced overall information,” Michaeli asserts.
Reg FD was designed to make the release of market-moving information fairer by restricting companies’ ability to brief select groups of analysts and investors ahead of a public announcement. The practice was seen to be widely abused; companies would reward analysts with private updates for favorable coverage, while analysts would deliver positive projections in hopes of earning business for the investment banking side of the house.
Michaeli’s model isn’t limited to the issues posed by Reg FD and doesn’t advocate abolishing the rule or otherwise altering the disclosure regimen in corporate America. Rather, the model is designed to explore the hypothetical effects of a mandate to disclose on the quality of information collected and released. Certainly, an argument could be made that Reg FD, rather than going too far, didn’t go far enough and should have mandated the kinds of information that companies must gather and disclose. But this, as we know, would require an extensive enforcement regimen.
The model assumes that a manager (a CEO, CFO or other official) has information to share with a group of “users” — which might include shareholders, analysts or venture capitalists. It further assumes that the two sides have different preferences for how the information is used: The users want to know what is actually going on and to act accordingly, while the manager wants users to follow his preferred course of action.
Consider a chief financial officer about to report quarterly results. Analysts want to base their earnings forecasts on a full picture of the company’s operations and financial health. The CFO might prefer a consensus forecast that would be easy to beat, or one that gooses the company’s stock before bonus time.
If the manager is required to release the collected information to everyone, he may choose to not gather essential information at all (e.g., same-store sale comparisons or data that may potentially reveal a decline in monthly customer sales). The more the interests of the two sides are at odds, the less likely the manager will gather and provide information, the model suggests.
However, if the manager can limit the size of the group that gets the information, or if the interests of the manager and his audience are more or less in synch — for instance, when a successful CEO touts her accomplishments before happy shareholders — the manager will gather (and provide) more precise information. By providing information only to a group of selected analysts (in the model, all users are assumed identical; results are even stronger if this is not the case), the manager can “guide” the average/consensus forecast to the desired level.
The result? Those who receive the information are considerably better off, since they get details the manager would otherwise have omitted. Those who aren’t briefed are at a relative disadvantage to those who are, of course. But on an absolute basis, uninformed users are no worse off than if the manager chooses not to disclose, Michaeli notes.