Why Price Impact Matters: The Legal Framework

Most class action securities claims alleging a violation under Section 10(b) and Rule 10b-5 of the Securities and Exchange Commission Act require a plaintiff to show reliance on the defendant’s misrepresentation. In a market with thousands of investors, requiring each to prove individual reliance would make class treatment nearly impossible. The law resolves this through the fraud-on-the-market theory, which presumes that in an efficient market, the price of a security reflects all publicly available information, including any material misrepresentation by a company. An investor who buys a stock at the market price is presumed to have relied on the integrity of that price, and therefore on the misrepresentation embedded in it.

This presumption is what makes securities class actions viable, and it rests on two economic premises. The first is that the market for the security is efficient, essentially meaning its price incorporates public information. The second is that the alleged misrepresentation affected the price. Both premises are economic questions, and both are typically addressed through event-study evidence.

The presumption is not irrebuttable. Defendants may attempt to rebut it by showing that the alleged misrepresentation had no price impact, that is, that the misstatement did not distort the market price at all. The Supreme Court confirmed that defendants are entitled to introduce evidence of a lack of price impact at the class certification stage to defeat the fraud-on-the-market presumption.[1] The practical consequence is that price impact, measured through event studies, has moved to the center of the certification battle. A defendant who can show through a  properly conducted event study that the corrective disclosure produced no statistically significant price movement may defeat certification entirely.

Market Efficiency: The Threshold Question

Before the fraud-on-the-market presumption can apply, the plaintiff must establish that the security traded in an efficient market. Courts have developed a set of factors to guide this inquiry, commonly drawn from two well-known district court decisions. The first set known as the Cammer factors considers indicators such as average weekly trading volume, the number of securities analysts following the stock, the presence of market makers and arbitrageurs, the company’s eligibility to file a simplified registration form, and, critically, evidence of a cause-and-effect relationship between company-specific news and price movements. The second set known as the Krogman factors adds more quantitative measures, including market capitalization, the bid-ask spread, and the proportion of a company’s outstanding shares of stock held by the public, which is also known as the float.

The most probative of these factors, and the one most directly addressed by an event study, is the demonstration of a cause-and-effect relationship between the release of new, unexpected company-specific information and a prompt change in the stock price. This is precisely what an event study is designed to test. If the security’s price reacts quickly and measurably to new information, such as earnings surprises and other material news, that reaction is strong evidence that the market processes public information efficiently. An event study showing statistically significant price reactions to a sample of news events is therefore often a key piece of the plaintiff’s market-efficiency showing.[2]

The Mechanics of an Event Study

At its core, an event study measures whether the price movement of a security on a particular day is larger than what normal market and industry factors would predict. If the actual daily movement or return significantly exceeds the predicted movement, the difference, known as the abnormal return, is attributed to the company-specific information released that day. The technique proceeds through several steps.

Defining the Estimation and Event Windows

The expert first selects an estimation window, a period of trading days used to establish the normal relationship between the security’s returns and broader market and industry movements typically measured by an index like the S&P 500. This window should reflect typical trading behavior and should generally exclude the days on which the alleged fraud-related information reached the market, so that the baseline relationship is not contaminated by the very events under study.

The expert then defines the event window, the day or narrow set of days on which the relevant information, such as an alleged misrepresentation or a corrective disclosure, became public. Selecting the correct event date is consequential. If the information leaked earlier, or if the market did not fully absorb it until the following day, an event window that misses the actual price reaction will produce a misleading result.

Building the Market Model

Using the estimation window, the expert fits a market model, typically a regression of the security’s returns on the returns of a broad market index and sometimes an industry or peer company index. The model captures how much of the security’s normal day-to-day movement is explained by market-wide and industry-wide factors that are unrelated to company-specific news. The regression yields coefficients describing the security’s sensitivity to those factors (often referred to as the company’s “beta”).

Calculating Abnormal Returns and Testing Significance

On the event date, the expert uses the estimated regression model coefficients to compute the return the security would have been expected to achieve given how the market and the industry moved that day. The difference between the actual return and this predicted return is the abnormal return, the portion of the movement attributable to firm-specific information. The expert then tests whether that abnormal return is statistically significant, meaning larger than could plausibly be explained by ordinary random fluctuation, typically by comparing it to the normal volatility of the security’s returns during the estimation window.

A statistically significant abnormal return on the day corrective information reached the market supports the conclusion that the information had a price impact. The absence of a significant abnormal return, conversely, demonstrates the alleged misrepresentation did not distort the price (within the event study framework) and that the presumption of reliance should not apply.

Price Impact, Loss Causation, and Damages

Event studies do work at more than one stage of a securities case, and it is important to distinguish the roles.

At class certification, the event study addresses market efficiency and price impact, the questions that determine whether the fraud-on-the-market presumption applies. Here the focus is often on whether the alleged fraud, and especially the corrective disclosure that revealed the truth, moved the price at all.  If so, the class of investors commonly relied on the misrepresentation because it was embodied in the price paid by all investors.

At the merits stage, the event study speaks to loss causation, the requirement that the plaintiff’s economic loss following a share price decline was caused by the revelation of the truth rather than by unrelated negative market or company developments. A plaintiff must connect the price decline to the disclosure that corrected the earlier misrepresentation. An event study isolating a significant price drop on the corrective-disclosure date, while controlling for market and industry movements, is the standard tool for making that connection.

Finally, the event study underlies the calculation of damages.

Confounding Information: The Central Criticism

The most common and most serious challenge to an event study is the problem of confounding information. An event study attributes a day’s abnormal return to the specific disclosure at issue, but stock prices generally respond to all information that becomes public on a given day. If a company announces a disappointing earnings restatement, that is potentially the corrective disclosure, but if it simultaneously announces the departure of its chief executive, a dividend cut, or lowered guidance – information not necessarily related to the misrepresentation – the price reaction embodies the combined effect of this news.

When multiple pieces of material information reach the market on the same day, the expert must disentangle the portion of the price movement attributable to the fraud-related disclosure from the portion attributable to other, non-actionable news. This is often the decisive battleground between opposing experts. A plaintiff’s expert may attribute the entire price drop to the corrective disclosure (assuming the broader market did not fall that day), while a defense expert could argue that most or all of the decline reflects confounding information unrelated to the alleged fraud. A study that fails to carefully account for potentially confounding news invites the argument that its abnormal return measures something other than the price impact of the fraud, and such a study is exposed to challenge under the standards governing expert testimony.

Methodological Points That Determine Credibility

Several technical choices distinguish an event study that withstands scrutiny from one that does not.

The choice of market index used for the regression model and to compute abnormal returns is important to distinguish price movements attributed to market conditions from those attributed to company-specific information. The market and industry indices used in the regression should genuinely reflect the broad market factors driving the security’s normal returns. An ill-fitting industry index can confound industry-wide movements with company-specific news, distorting the abnormal return.

The estimation window – the time period over which the regression model coefficients are estimated – is important for identifying a pre-event relationship between the company’s stock and the overall market. A window that is too short may produce unstable coefficients, while one that spans a period of structurally different trading behavior may misstate the security’s normal sensitivity to the market. The window should also avoid contamination by the event itself.

The statistical significance threshold and testing is important to demonstrate whether or not abnormal returns during the event window were the result of random chance. The choice of confidence level (the upper and lower boundaries within which abnormal returns generally lie), the treatment of the security’s volatility, and whether the expert accounts for the possibility that volatility itself changed around the event all affect whether an abnormal return is judged statistically significant. Opposing experts frequently contest these choices, and a test result that is significant under one reasonable specification but not another can be open to criticism.

Event-date selection is important because it instructs the model where to look for the relationship between the price and the event. Because information can reach the market gradually or ahead of an official announcement, such things as the definition of the event, the timing of its occurrence, the way the information reached the market, as well as the expert’s treatment of possible leakage, can determine whether the study captures the true price reaction.

Robustness matters throughout, starting from the definition of the event, the collection of share price data, the identification of the market index, the implementation of the econometric regression model and other considerations discussed in this blog. As with any econometric analysis performed in litigation, an event study is stronger when the expert tests the sensitivity of the results to reasonable alternative specifications and when the methodology is transparent and reproducible, so that an opposing expert working from the same data, assumptions and methodology could follow, replicate and evaluate every step in the event study process.

Conclusion

The event study occupies a prominent place in class action securities litigation. It is the economist’s analytical tool through which overarching questions of market efficiency, reliance, causation, and damages are used to produce empirical economic evidence. Whether a court applies the fraud-on-the-market presumption, whether a class is certified, whether loss causation is established, and how much investors may recover all turn, in large part, on what an event study shows about price impact.

That centrality is also one reason the technique is scrutinized so closely. Event study techniques are well established, but the expert’s implementation decisions, the estimation window, the choice of indices, the event date, the significance testing, and above all the treatment of confounding information, are where cases are contested and where credibility is won or lost. A rigorous, transparent event study that isolates the price impact of the alleged fraud and transparently confronts competing explanations is among the most persuasive forms of evidence in this field. One that ignores or does not adequately address confounding news or rests on fragile model specification decisions is among the most vulnerable. The difference, as always in expert economic analysis, lies in discipline, transparency, and a methodology built to meet the court’s high standards for expert testimony and withstand the adversary rebuttal criticisms of the opposing side.