This is a useful paper as far as it goes. The problem is that it has missed a significant chunk of the “literature” on this rapidly developing topic that has developed in our rapidly developing medium — i.e., the blogs. For example, consider my posts here and here and throughout my executive compensation archive, Josh Wright and Geoff Manne’s comprehensive post, and many many others by Bainbridge, Bodie, Fleischer, etc. This is not merely some kind of procedural problem. By missing this commentary, the article fails to pay any attention to some very important issues, particularly including whether the market looked through any accounting shenanigans. The latter issue alone would seem to be rather critical if you’re trying to explain backdating and its consequences, as Walker is.
Larry’s reaction to the paper has evoked reactions from Vic and Walker in the comments section. Holding aside the issue of whether Manne & Wright should be cited (as readers of this blog know, Geoff and I have elsewhere set forth our own thoughts on backdating, individually and cooperatively), I took Larry’s central criticism to be that the argument raised by some of these bloggers that “stealth compensation” is simply not a good description of backdating if it did not fool the market should be addressed in a paper purporting to explain backdating. It is a fair point and I tend to agree. To be sure, it is an excellent marketing strategy to describe the options this way. Indeed, I could describe backdated options as “alternative in-the-money compensation.” But I digress. Further, Larry offers a second post responding to Walker’s comment, and argues that any substantive explanation of backdating must address whether the market was fooled:
I continue to be puzzled how one can argue that options were “stealth compensation” without discussing whether enough information was available that the compensation was reflected in stock price. If the market knows what the executive is being paid, then I’m not sure how one can argue that it could not make the judgments that Walker is concerned about.
I agree with both Ribstein and Walker that this sort of exchange is precisely what the blogosphere needs more of. In that spirit, let me chime in with a few of my own thoughts here in response to Walker’s article.
First, the empirical contribution of this article should celebrated. In particular, in addition to documenting the fact that a good deal of backdating occurs with rank-and-file employees rather than executives, Walker conducts a descriptive analysis of backdating within the semiconductor industry and highlights differences in executive compensation for firms involved in the backdating scandal (p. 34-35). I think this sort of descriptive analysis is definitely value added and tells us more about the phenomenon which we are attempting to ultimately explain.
Second, I am left somewhat unsatisfied with Part III of the paper (which starts at p.21), which is titled “Explaining Backdating.” To be sure, Walker notes that “the aim of this part is to lay out a range of possible rationales,” and test them against the early empirical evidence. For my tastes, this paper does too much of the former and too little of the latter. Walker discusses a range of rationales including compensation concealment, share dilution limitations, cognitive biases, boosting ISO grants, and the influence of common advisors (which Walker lumps together, somewhat inexplicably, with “herd mentality”). With respect to herd mentality, the “evidence” is that a number of firms adopted the same practice in the Silicon Valley and Larry Sonsini was linked to many firms. I’m not sure what this has to do with “herd mentality,” but I can think of a number of reasons why many firms adopt the same practice which have nothing to do with psychology.
As for the other explanations, again, I find the paper a bit light on the discussion of evidence, which is odd, because I do believe that the central (and most important) contribution of Walker’s paper is his empirical work. Walker seems to be impressed with the naivete / cognitive bias explanation throughout this section, but as I have noted elsewhere, I do not find this explanation persuasive in light of the time series evidence (have compensation committee’s become more naive? Or for that matter, employees or executives?). In any event, to the extent that many of these explanations touch upon the economic explanation for the increase in executive compensation more generally, simple explanations like this one (that apparently explain much of the data) should be addressed, as should evidence of the stock price effects.
The strength of this paper, by my lights, is Walker’s empirical analysis. His contribution to our understanding of what is going on within a particular industry with a lot of backdating is an important one. In fact, I would be inclined to organize the entire paper around this analysis — which is really his unique contribution. I know, nobody asked me. Just a thought.