In a very interesting May 15, 2008 paper entitled “Do Options Backdating Cases Settle for Less?” (here), NERA Economic Consulting takes a look at the options backdating-related securities class action lawsuits settlements to date, and concludes that “in the cases that have settled to date, the amounts paid to plaintiffs have been substantially lower than in comparable non-backdating class actions.” NERA’s analysis is that the options backdating class action lawsuits are settling for half the amounts forecast by NERA’s own prediction model.
Having made this rather provocative observation, NERA then concedes that only a fraction of the options backdating-related securities class action lawsuits filed have yet settled. Clearly one factor that may be involved is that the weakest cases may have settled first, a consideration that the NERA study expressly acknowledges.
Nevertheless, in attempting to understand the variation between the settlements to date compared to the expected range of settlements based on NERA’s model, the NERA report does consider the possibility that “shareholder suits with backdating allegations are perceived as weaker on the merits than other class actions.” The report also considers the possibility that future options backdating settlements, which might include more serious cases, could be more in line with other securities class action settlements.
I have several observations about the NERA analysis, the first of which is that is important for all of us to keep a running tally of outcomes, to make sure we all know and keep track of what is happening. The fact that this study comes from NERA suggests that it will (appropriately) carry weight and credibility.
That said, it should also be noted that the NERA study is based on a small sample, only six settlements out of 37 options backdating related securities class action lawsuits. (The total number of lawsuits according to my tally, here, is only 36, but I am willing to go with their number for these purposes, which is close enough anyway.)
Not only is the sample small, but it seems to have been amputated at a couple of critical points. That is, for reasons that are not explained in the report, the NERA dataset does not include either the Mercury Interactive settlement ($117.5 mm) or the Vitesse Seminconduct settlement ($10.2 mm). If I know NERA, there are probably some very good reasons why they excluded these settlements, but the report does not explain or even refer to the omission of these settlements. Given the size of the Mercury Interactive settlement in particular, the omission of these settlements could have had a significant impact on the analysis, so their omission could be significant.
UPDATE: Dr. Branko Jovanovic, one of the author's of the NERA report, was kind enough to call me and politely point out that the report actually refers, in footnote 3, to the fact that the report's authors chose to exclude the Mercury Interactive and Vitesse Semiconductor settlements from the analysis because some but not all defendants had settled. (That's what I get for trying to write blog posts in a hotel room in Toronto without the ability to print out and read documents in hard copy form. Reading the report on a laptop screen, I just missed the footnote). Dr. Jovanovic points out that if the two settlements had been included, it would have increased the difference between expected and actual settlements.
The other thing about NERA’s analysis is that as a result of the small dataset, extreme individual results could be skewing the average. In particular, according to the report, the Rambus options backdating related securities class action lawsuit of $18 million, was only 8.3% of predicted. For me, an outlier result like that suggests that it is not representative, and in fact some case specific factor may explain the outcome. In any event, an extreme result like that clearly pulls down the average. While the exclusion of the Rambus result would still not eliminate the variation from the predicted range, it would reduce the difference.
I also think it is significant in considering whether the options backdating cases are or are not deviating from expectations that dismissals should be taken into account as well as settlements. According to my running tally of option backdating related settlements, dismissals and denials (which may be accessed here), six of the 37 (or is it 36?) options backdating related securities class action lawsuits have been dismissed. (Some of these dismissals are without prejudice). I am not 100% sure which way this cuts, but I think the number of dismissals is a relevant consideration to any analysis of whether or not outcomes are within predicted ranges. The dismissals may also provide some explanation, or at least context, for the variation between settlements to date and predicted ranges.
All of that said, I reiterate my appreciation to NERA for their effort to keep track of what has happened so far. The value of NERA’s analysis is in its provision of a status update, which is a sevice that we can all hope that NERA will continue as the cases develop.
For the sake of completeness, I urge all readers interested in these topics to review the analysis of options backdating securities class action settlements on the Securities Litigation Watch blog (here), which among other things notes that these cases are settling more quickly on average than other cases, which clearly might be a factor in explaining settlement outcomes. The SLW's analysis not only takes into account the settlements that NERA's report omits, but it also considers the dismissals as well.
One final observation is that NERA's analysis relates solely to options backdating securities class action settlements, and does not refer to or include options backdating derivative lawsuit settlements. For further information regarding options backdating derivative lawsuit settlements, please refer to the table I am maintainting, here.
CFO.com has an article discussing the NERA report here.