Michael Guttentag develops a nice model of disclosure regulation that captures the interactions between two possible types of disclosures: those that improve share price accuracy and those that reduce agency costs.  The unique feature of Guttentag's model is that it allows for the possibility that the former category of disclosures might also reduce agency costs.  The model is generally tractable, though a bit messy in a few places, and pretty straightforward in its implications.  Some of the results of the model are not very surprising.  For example, the model suggests that the gains from accuracy enhancing disclosure regulations are increasing in the size of inter-firm externalities and spillover benefits (the extent to which accuracy enhancing disclosures reduce agency costs).  Perhaps the most important contribution of the model is that it provides a useful way to disentangle some of the tough issues regarding social welfare and disclosure.

    A few comments on the model and its implications...

    One brief modeling note before I move on to some more general comments.  I wonder, and readily admit that I have not taken the time to do the math here, how robust the findings are here to alternative assumptions about the costs of disclosure regulation.  There is a plausible story for assuming linear costs (and a simple model always is best to start with), but one could come up with a story where these costs increased in a non-linear manner.  If you have tried other cost functions, I would like to know if the results changed and how.  I would find persuasive evidence that the policy implications drawn from the model were robust to changes in the costs of disclosure.

   One implication of the model is that increased disclosure results in greater social welfare gains for firms where "accuracy enhancing disclosures are more likely to substantially reduce agencys costs" (p. 19).  From this result, Guttentag concludes that social welfare gains would therefore be greatest from regulation of publicly held firms (or more generally, those with dispersed shareholdings).  I follow the logic here, but have some concerns about social costs of regulation not captured in the model.  For example, one social welfare gain caused by the dispersion of shareholder wealth is reducing firm specific risk.  The concentration of ownership of the firm is the outcome of the process of economizing on both agency costs and the costs of bearing firm specific risk.  Both are real social costs.  Because additional disclosure regulations introduce costs to firms with unconcentrated shareholder wealth, on the margin, such regulation will have an impact on ultimate concentration of ownership of the firm.

    Of course, we dont know the magnitude of the social gains (from reducing agency costs) or costs (increasing firm specific risk, costs of regulation itself, unintended consequences, etc.).  A nice feature of Guttentag's work is that he derives some testable implications (p. 20) with respect to some of these questions.  This is a very clean model with some pretty clear testable implications.  I think that Section 7 does not yet do justice to the implications that could be teased out of this model (and could result in a follow-up paper with some empirical work).

    My final comment is that this model made me think about this JLS paper by Cain, Lowenstein and Moore.  The punchline is some experimental evidence that disclosure can make agency problems worse.  I wonder what implications this finding might have for your paper.  Maybe none since the justification for regulation is not on the conflicts side but the accuracy enhancing side, but I thought I would point it out nonetheless.

    This is an interesting paper that takes an important step in disentangling often convoluted issues with respect to social welfare and different forms of disclosure rules.  By focusing on the possibility that accuracy enhancing regulations might also reduce agency costs, the model gives some interesting answers to questions like: should any firms be regulated? If yes, which ones?  I am hesitant to adopt the policy recommendations that follow from the model because I believe there are some real social costs it does not capture.  Nonetheless, the paper contributes to the discussion of types of disclosure regulation and offers promising testable implications.

Junior Scholars | Bookmark

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345157d569e200d83454fcea53ef

Links to weblogs that reference Guttentag on Disclosure Regulation: