Pomerantz LLP

Leakage Theory is No Longer Just a Theory

Attorney: Michael J. Wernke
Pomerantz Monitor September/October 2015

It all started ten years ago with a question posed by  Justice Stevens during oral arguments in Dura Pharm., Inc. v. Broudo: “What if the information leaks out and there’s no specific one disclosure that does it all and the stock gradually declines over a period of six months?” Until last month, this question remained in the “what if” category of securities fraud jurisprudence. We now have an answer. In Dura, the Supreme Court held that a plaintiff in a securities fraud action must plead the element of loss causation, i.e. that the company’s stock price declined once the truth was revealed through a corrective disclosure. At trial, the plaintiff must ultimately prove that the decline in stock price was a result of the fraud – not market, industry or company-specific nonfraud factors. Since Dura, courts have universally held that loss causation can be established even if the truth is revealed through multiple “partial” corrective disclosures that drove the stock price down.

Courts have also acknowledged that, in theory, a company’s stock price could decline as a result of the truth “leaking” into the market without any actual disclosures of the fraud. For example, the stock price may move because insiders traded on the inside fraud related information prior to a disclosure, or because investors gradually lost confidence in the company’s previous misrepresentations even though the truth was not yet officially disclosed. However, in practice, courts have until now required plaintiffs to connect any decline in stock price to an identifiable “corrective” disclosure. 

The Seventh Circuit’s June 21, 2015 decision in Glickenhaus & Co. v. Household InternationalInc. has lifted that restriction, creating the possibility that investors may recover losses resulting from the gradual decline in a company’s stock price that is not directly connected to any corrective disclosure, but which can be attributed indirectly to the unraveling of the underlying fraud. 

In Household, the defendants appealed to the Seventh Circuit a jury verdict finding them liable for securities fraud on the basis that the causation/damages model adopted by the jury failed to establish loss causation. The plaintiffs had presented two models to the jury. The first, a “Specific Disclosure Model,” identified fourteen partial corrective disclosures that revealed the truth to the market and calculated the price declines that followed within the next day, removing price movements attributable to market and industry factors. This model determined that disclosure of the fraud led directly to investor losses of $7.97 per share. The second analysis, the “Leakage Model,” attributed to the fraud all the price declines during the year-long period of partial disclosures, except for declines caused by market or industry factors. Using this model, plaintiffs calculated that losses per share were $23.94. The jury adopted the Leakage Model and damages were ultimately determined to be $2.46 billion.

In their appeal, the defendants argued that the Leakage Model was flawed because it included price declines that did not immediately follow any of the partial disclosures of the fraud. While the Leakage Model eliminated market and industry factors, it did not identify and eliminate the effect of company-specific, nonfraud news on the stock price, which may have contributed to the decline in stock price during the periods between the fourteen partial corrective disclosures. Instead, plaintiff’s expert testified in general terms that he considered the issue but was unable to conclude that non-fraud news would have altered the analysis. The question before the court was whether that was enough or whether the model itself must fully account
for the possibility that company-specific, nonfraud factors affected the stock price.

The court refused to answer simply “yes” or “no,” as doing so would create an unfair advantage for plaintiffs or defendants. Accepting the defendants’ position would likely doom the leakage theory because it may be “very difficult, if not impossible,” for any statistical model to separate damage caused by “leakage” from damage caused by release of company-specific news unrelated to the fraud. On the other hand, if it’s enough for an expert to offer a conclusory opinion that no company-specific, nonfraud related information affected the stock price, then plaintiffs may be able to easily evade their burden of proving that the loss for which they seek recovery was a result only of the alleged fraud.

The court chose a middle ground, creating burden-shifting process to be used at trial. It held that if the plaintiffs’ expert testifies in a nonconclusory fashion that no company-specific, nonfraud related information contributed to the decline in stock price, then the burden shifts to the defendants to identify some significant, company-specific, nonfraud related information
that could have affected the stock price. If the defendants can, then the burden shifts back to the plaintiffs to account for that specific information or provide a model that doesn’t suffer from the same problem. Significantly, the court stated that one solution for the plaintiffs would be to simply exclude from the model’s calculation any stock price movements directly related to the company-specific nonfraud information identified by the defendants. 

While the defendants won the battle – the case was remanded to the trial court – investors may have won the war. Plaintiffs’ recoveries in a securities fraud action are no longer limited to stock price declines immediately following specific disclosures of the fraud. Moreover, the
Seventh Circuit provided a clear roadmap for the creation and use of a leakage model that can withstand judicial scrutiny (at least in the Seventh Circuit). 

Notably, this decision came only a year after the Supreme Court’s decision in Halliburton II, which dialed back the more rigid views of market efficiency which had previously been employed by many of the lower courts, and installed a similar burden-shifting process for that analysis. The Seventh Circuit’s decision could be viewed as a road marker in a forming trend of courts taking a more practical view of how securities markets function and investors’ burdens in proving their losses from frauds.