The following comes to us from Scotland M. Duncan, an attorney at Jones Day in Pittsburgh, PA.*

It is axiomatic that calculating the amount of loss within securities fraud cases is essential to criminal sentencing.  Until about a year ago the U.S. Sentencing Guidelines (the “Guidelines”) did not expressly provide for any method of loss calculation.  Rather, the limited guidance was that a “court need only make a reasonable estimate of the loss,” “based on available information.”[1]  With no real guidance or formula, courts introduced a multiplicity of calculations, which resulted in capricious sentences.  Recognizing the problem, and no doubt in light of the fact that § 2B1.1 was one of the most applied Guidelines,[2] the United States Sentencing Commission (the “Commission”) sought comment in early 2012 on whether it should amend the Guidelines to set forth a formula for calculating loss in securities fraud.   On November 1, 2012, the Guidelines were amended and a “modified rescissory method” was adopted whereby loss is measured by average stock price during discrete periods of time.  Courts should now multiply the difference between (i) the average share price during the fraud period and (ii) the average share price during the 90-day period after the fraud was disclosed by the number of shares outstanding.[3]  The math seems straightforward and is easy for jurists to apply.  The new law’s arithmetic, however, suffers from a number of severe flaws.

In my article Recalculating “Loss” in Securities Fraud, recently published by the Harvard Business Law Review and available here, I argue that the amended Guideline imprudently opts for ease of judicial application over precision when the two need not be mutually exclusive. The Commission’s math can and will result in upward bias with respect to the number of damaged shares and skewed sentencing disparity (both upward and downward) due to the inclusion of extrinsic factors wholly unrelated to a defendant’s conduct.  For example, the Guideline’s arithmetic analyzes average prices over extended stretches of time, which is inconsistent with the efficient markets hypothesis.[4]  It is well recognized that the longer the span of time, the more likely it is that extraneous factors might affect the loss calculation.[5]  A defendant whose offense conduct spans a period that includes extraneous negative market forces that depress stock price during the inflationary period will invariably receive a much shorter sentence, ceteris paribus, than a defendant who engages in the same conduct during a span void of such extraneous forces.  The reverse also holds true: an offender might receive a harsher sentence due to an artificial increase in the loss calculation if the market is booming during the inflationary period.

In addition, by multiplying the difference in average price by total shares outstanding the Commission’s calculation presupposes that every outstanding share incurred harm.  In fact, and as discussed in more detail in my article, the actual number of shares that incurred losses is likely to be significantly lower, resulting in potentially draconian sentences.

In lieu of the Commission’s proposed arithmetic, my article instead proposes conforming criminal sentencing for securities fraud to its civil counterpart, as promulgated by the Supreme Court in Dura Pharmaceuticals, Inc. v. Broudo.  In Dura, the Supreme Court unanimously held that when evaluating civil securities fraud allegations, courts must consider “the tangle of factors affecting [stock] price.”[6]  Though Dura was a civil securities fraud case, much of the Court’s analysis therein is apropos to quantifying loss in the criminal context.  Dura reduced the amount of frivolous civil securities litigation,[7] in part, by setting forth a method for arriving at the appropriate economic loss.  Of the methods considered by the Commission, the “market-adjusted method” was and is the only one that calculates the actual loss “that resulted from the offense.”[8] This method focuses on normalized change in a damaged security’s value and is a more precise way to calculate actual loss as it removes the possibility of sentencing disparity (both upward and downward) due to the inclusion of non-fraud related factors.

Under the new loss calculation regime, we can expect diverse sentences for defendants engaged in interchangeable conduct due to both macroeconomic and microeconomic factors wholly unrelated to the defendant’s conduct.  Fortunately, Judge Patti B. Saris, Chair of the Commission, explained that these amendments are “the first step in a multi-year review of the fraud guideline.”[9] Hopefully that means future refinement of the Guidelines consistent with my article.

_______________