Economic Analysis Of Loss In The United States Sentencing Commission's Proposed Methodologies

Introduction

On 19 January 2012, the United States Sentencing Commission asked for public comment on various topics. It stated that it had received comments noting that "determinations of loss in cases under §2B1.1 [of the Federal Sentencing Guidelines Manual] involving securities fraud and similar offenses are complex and a variety of different methods are in use, resulting in application issues and possible sentencing disparities."

This article argues that if they are to be reasonably accurate, the determinations of loss in cases involving securities fraud must necessarily be somewhat complex because different types of fraudulent schemes result in different types of losses. It examines the four methods discussed in the Commission's request for comments, showing how they can either underestimate or overestimate losses depending on the fraudulent scheme involved. It then outlines a method that will tend to reduce the degree of inaccuracy, though noting that no generic method will be appropriate in all scenarios. Ultimately, the question is whether courts are willing to hear evidence that will result in more appropriate estimates of loss in specific cases or rely on more generic calculations that reduce perceived, but not actual, disparities.1

Per-Share Inflation

Financial fraud typically raises the price of a security such as a stock above its true level, with the difference between the two called the inflation in the price of the security. For example, suppose that a stock was trading at $10 but a false statement raises the price to $13; then, there is $3 in inflation in the share price. The inflation often leaves the share price with the release of one or more corrective disclosures. For example, an initial partial disclosure can lower the stock price to $12, meaning that the inflation in the stock would be $2 if the true value were still $10. Then a final corrective disclosure could bring the stock price back to its true value of $10, which we again assume has not changed. Investors who bought at $13 and held through the final corrective disclosure suffered a loss of $3 because they held their shares when they suffered a first decline of $1 and a second decline of $2, a total of $3.2

To see how inflation affects losses and loss estimates, consider the following two examples, which we examine using the four methods in the United States Sentencing Commission's letter.

Example 1: The stock of company X is trading at $10. On Day 1, company X makes a material misrepresentation that raises the stock price to $12. Thus, the initial inflation is $2, or $12 less $10. On Day 2, company X announces materially fraudulent earnings that raise the stock price to $16. Assume also that the true earnings were better than expected, and the stock would have risen to $15 had they been announced. (i.e., the inflation would have fallen from $2 per share to $1 per share, $16 less $15. This could occur for example, if true earnings unexpectedly began to catch up with previously inflated earnings and guidance, thereby reducing the gap between the truth and the information stated to the market). On Day 3, company X comes clean and the stock falls to its true value of $15. (See Figure 1.)

Now we can examine the losses as calculated under the four methods:

1. "a simple rescissory method (under which loss is based upon the price that the victim paid for the security and the price of the security as it existed after the fraud was disclosed)." Here, an investor who purchased on Day 1 paid $12 and the price after the fraud was disclosed was $15. There is no calculated loss associated with this investor's purchase, even though the investor overpaid by $2 and suffered a $1 price decline on Day 3. This method therefore underestimates per-share loss for this fraudulent scheme.

2. "a modified rescissory...