Ninth Circuit rejects use of Dura Pharmaceuticals' civil loss-calculation method to fix "loss" for criminal sentencing
Nearly five years, the Supreme Court held in Dura Pharmaceuticals. Inc. v. Broudo, 544 U.S. 336 (2005) that a civil securities fraud plaintiff was required to show both that an alleged fraud was disclosed to the market and that the disclosure caused a loss to shareholders, that is, that the share price fell after the truth about a defendant's fraud became known. No longer could civil plaintiffs allege the requisite loss on the theory that they had purchased stock at a price artificially inflated by undisclosed losses, because someone who purchased overvalued stock could quickly sell that stock and sustain no out of pocket loss. Two courts of appeals have applied Dura's loss-causation theory to limit the "loss" and resulting Guidelines range for criminal securities fraud defendants. United States v. Olis, 429 F.3d 540 (5th Cir. 2005); United States v. Rutkoske, 506 F.3d 170 (2d Cir. 2007).
But in a recent case, the Ninth Circuit rejected the approaches of the Fifth and Second Circuits and held that Dura's loss formulation methodology does not apply to sentencing in criminal cases. United States v. Berger, 2008 WL 4141478 (9th Cir., Nov. 30, 2009). The defendant was the CEO, President, and Chairman of the Board of Craig Electronics, where he had actively concealed the company's true, negative financial condition from lenders. When an outside auditor forced a restatement of earnings, Craig's share price fell but, still, the fraud committed by Berger was not disclosed to the market until the company's stock was finally delisted. Under Dura. there was no loss that the government could show was suffered by the market at the moment following disclosure of the fraud; instead, the government pursued a stock overvaluation theory to establish millions in losses for sentencing purposes.
But the Ninth Circuit reasoned that Dura was limited to the civil context, where a given plaintiff was required to show that he or she sustained an actual, economic loss. In contrast, in the criminal setting, the Guidelines expressly endorse the calculation of loss by taking a snapshot of the then-overstated and misrepresented value of an item, such as a worthless check represented to have value, and assessing the loss as the difference between actual and overstated values (see U.S.S.G. § 2F1.1, cmt. 7(a)). So, a criminal court may assess "loss" by measuring the amount of loss caused to the market as a whole, which could appropriately be done by measuring the overvaluation of Craig's stock at a given point in time before disclosure of the misdeeds.
Although agreeing that the court below correctly eschewed reliance on Dura, the Ninth Circuit did reject the unusual approach taken by the district court in measuring loss to Craig's shareholders by looking to the typical change in stock value of other companies where frauds had been disclosed. This was rejected as an inapt methodology because Craig's irregularities were never disclosed before trading was ended. On remand, the trial court was obliged to try again to gauge the difference between Craig's share-price as inflated and what it would have been absent the misrepresentation, with no guidance from the Ninth Circuit as to how precisely to accomplish the task.
The appeals court did note in footnote 7 that Dura might well control Berger's restitution analysis, since the MVRA focuses on actual harm sustained by any victims, as opposed to broader losses caused by the defendant. In short, the Berger decision is bad news for defendants seeking through application of Dura to cabin their loss numbers for Guidelines purposes, although Dura may still be helpful in reining in the restitution figure.
