chevron-down Created with Sketch Beta.

The Business Lawyer

Fall 2022 | Volume 77, Issue 4

Market Prices vs. Fundamental Value: The Case for Using Discounted Cash Flow Analysis in Securities Class Actions

Frank Partnoy


  • This article proposes discounted cash fl ow (DCF) analysis as a substitute for, or complement to, event studies in securities class actions. In many contexts, DCF analysis would be more consistent with the principles articulated in the Supreme Court’s decision in Goldman as well as the academic literature. DCF analysis also could avoid many problems associated with event studies and provide a more intuitive and precise framework for assessing both price impact and merits issues in securities class actions. The central idea is to shift from the analysis of market prices to the analysis of fundamental value.
Market Prices vs. Fundamental Value: The Case for Using Discounted Cash Flow Analysis in Securities Class Actions

Jump to:


In this article, I propose that courts accept expert testimony based on the well-established methodology of discounted cash flow (DCF) analysis as a substitute for, or complement to, event studies in securities class actions. I show how DCF analysis fits within the judicial framework for expert testimony related to class certification and merits issues in securities litigation. My basic normative point is that judges should shift their focus to examining the fundamental value of firms instead of the market prices of their shares.

This point is well understood in other contexts, where courts sometimes rely on market prices and sometimes rely on DCF analysis. For example, in appraisal litigation under Delaware law, courts routinely use DCF analysis in situations where market prices cannot be relied on as evidence of fair value. The idea I advance here is that, although courts in securities class actions might look to market prices (and event studies) to examine market price changes, as in the analysis of price impact at class certification, when examining the “real” or fundamental value of companies to determine issues on the merits, market prices (and event studies) are less likely to be helpful. Instead, in determining the “out-of-pocket” measure of damages in securities class actions, the focus is, and should be, on the difference between the price paid for securities and the “real” or fundamental value at the time of purchase.

In a securities class action, courts and litigants frequently assume that market prices reflect available information, yet, until there has been a corrective disclosure, prices by definition do not reflect the fundamental value of a company. After a corrective disclosure, fundamental valuation and the analysis of stock market prices might generate similar results, but before the corrective disclosure they definitively do not. Indeed, that is the central point in securities litigation: that market prices allegedly did not reflect misstatements or omissions related to the fundamental valuation of a company. Put another way, although experts could examine both DCF analysis and event studies based on market prices on dates after a corrective disclosure, on dates before a corrective disclosure, any event studies necessarily are based on prices that are assumed to have been fundamentally wrong.

More generally, the financial asset pricing literature suggests that market prices encode a range of information, including confounding information, and one central effort has been to decompose the factors that matter to market returns. Asset pricing models are highly variable and recent research suggests a degree of skepticism about how market prices and returns encode information associated with various factors. In any event, it is clear from the finance literature that DCF analysis is deeply connected in a theoretical sense to asset pricing: consistent with the capital asset pricing model, stock prices can be expressed in DCF terms.

Here is how an analysis based on DCF analysis could work (perhaps along with event studies). In a simple case with one corrective disclosure, the expert methodology could proceed in three phases. First, experts could perform one or more DCF analyses of the defendant company immediately following the corrective disclosure, based on publicly available data, including investor and analyst assessments and projections. The market price could serve as a “check” on the DCF analysis; in other words, experts on both sides could assume that in an efficient market the stock price reflected information associated with the corrective disclosure.

Second, the experts could perform similar DCF analyses as of the date immediately before the corrective disclosure, based on the market price at that time. Again, the market price could serve as a check, assuming that the price incorporated information other than the corrective disclosure. The DCF analysis before the corrective disclosure would require experts to estimate different inputs to the DCF model, based on available evidence.

Finally, the experts could perform one or more counterfactual DCF analyses as of the beginning of the class period, assuming there were none of the alleged misstatements or omissions. The difference between this counterfactual valuation and the market price at the beginning of the class period would establish the initial point estimate for the beginning price inflation “ribbon,” which then could be adjusted throughout the class period, depending on changes in counterfactual DCF valuations as of each date. In some cases, the width of the “ribbon” might be the same at the beginning of the class period as compared to the width based on a valuation immediately before the corrective disclosure; in other cases, the width of the “ribbon” might vary, depending on the facts and analyses.

Conceptually, the DCF methodology would overlap with the use of event studies, but would differ in two important respects. First, the DCF methodology would emphasize point estimates based on data and stock prices at particular times, rather than comparative changes in stock prices over time relative to the market overall, industry indices, or other measures. The DCF methodology would account for market and industry impacts in the valuation analyses, so that the magnitude of the inflation “ribbon” would vary based only on factors related to the alleged misstatements and omissions (instead of accounting for market and industry impacts in event study analyses by attempting to parse how much of a price decline was associated with changes in the market or relevant indices). Second, the DCF methodology would emphasize the details of changes in the metrics relevant to corporate valuation such as cash flow estimates and discount rates, rather than focusing on present and historical stock prices and their relationship to disclosures. In terms of a metaphor, if the event study methodology examines changes in the performance of a car over time, the DCF methodology emphasizes looking under the hood at specific times to see why the car performed differently.

Expert use of DCF analysis would be consistent with the overall jurisprudential approach in securities class actions. For decades, judges have cited expert testimony based on event studies in applying the “out-of-pocket” measure of damages in such actions, limiting recovery to the difference between the price paid for securities and the “real” value at the time of purchase. The DCF methodology likewise would assist judges (and in the rare case of trial, juries) in assessing the “real” value of securities at the time of purchase in order to determine the “out-of-pocket” measure of damages. Relatedly, the DCF approach could be used to assist judges with respect to determinations of other issues in securities class actions where event studies are currently used, including price impact, materiality, and loss causation. My primary argument here is that expert testimony based on the DCF methodology has the potential to improve on testimony based on event studies, by more directly assessing several key areas of dispute among experts in securities class actions, including fundamental questions about the width of the stock price inflation “ribbon” during the class period (and changes in the width of this “ribbon” over time), as well as thorny inquiries into materiality and loss causation.

The DCF approach also would be consistent with, and indeed arguably flows from, the Supreme Court’s 2021 opinion in Goldman Sachs Group Inc. vs. Arkansas Teacher Retirement System. In particular, the four important paragraphs in Justice Barrett’s opinion that set forth principles for courts to use in assessing “price impact” at the class certification stage of securities class actions are broadly consistent with this DCF proposal. Although those paragraphs might seem at first blush to be limited in scope, they mark a new trail for lawyers and experts to follow in future cases in assessing the potential “mismatch” between misstatements and omissions compared to corrective disclosures, determinations that are made at both class certification and later merit stages.

First, in terms of “price impact,” the concept at issue in Goldman, some aspects of DCF analysis resemble prior efforts by experts at the class certification stage to assess, either explicitly or implicitly, the probability that a misrepresentation had no price impact on the stock at issue. As I show below, DCF analysis could assist courts in determining whether defendants have borne the burden of persuasion to make a showing that in fact it is more likely than not that an alleged misrepresentation had no price impact.

Second, experts and courts could transition to using DCF analysis to assess key merits questions, particularly loss causation and damages. The DCF methodology is a promising alternative, especially in difficult cases where there is confounding information at the time of corrective disclosures, or there has been informational leakage or other factors that limit the utility of event studies in such cases. Indeed, one potential criticism of the use of the DCF methodology—that it lacks precision with respect to what the “real” cash flow estimates would have been absent any misrepresentation—illustrates the strength of the methodology, namely that it provides judges with a more “common sense” explanation about the differences between actual and hypothetical expert estimates than the more opaque event study approach. To the extent the DCF methodology might appear imprecise, that is primarily because it is more transparent; the statistical underpinnings of the event study methodology are less intuitive and accordingly the inherent imprecision can be largely hidden from view.

Both the shift to “common sense” suggested in the Goldman litigation, and this proposal for adding DCF analysis to the mix of expert testimony in securities class actions, follow, as with so many aspects of securities regulation, from the prescient writings of Professor Donald Langevoort. He saw all of this coming, more than a decade ago, just as he has predicted so many other difficulties arising from various securities practices and judicial decisions. Specifically, he described the likely problems associated with slavish reliance on the outputs from event studies and the dearth of “common sense” approaches in securities cases; indeed, the source of the term “common sense,” emphasized in Goldman, is him.

As Professor Langevoort wrote in 2009, “[t]he downside to making the science so determinative is that judges and other non-expert policymakers may not notice when very important questions of law are assumed away in all the number crunching.” Adopting his language, experts using the DCF methodology would be forced by its very transparency to do more than merely crunch numbers. Instead, they would be forced to crunch numbers in ways that emphasize the fundamental issues in securities class actions and therefore would be potentially more helpful to judges.

In Part I, I provide some background and context. I describe how both the principles in Justice Barrett’s opinion and the initial application of those principles at the district court level, particularly in Judge Crotty’s decision in the Southern District of New York re-certifying the class in Goldman, set the stage for the new DCF approach to expert witness testimony in securities class actions. In Part II, I discuss a few theoretical considerations at the heart of the principles set forth in Goldman, including a methodological framework for understanding the innovations implicit in Judge Crotty’s opinion. His “sliding scale” approach is consistent with peer-reviewed academic literature on causal inference, particularly in the area of securities markets, and also can be framed in terms of a more general framework based on standard approaches in probability and statistics. It also, accordingly, is consistent with the DCF approach. Finally, I discuss in Part III how the specifics of DCF analysis are consistent with both basic principles in securities litigation and these theoretical considerations. I also describe some practical potential benefits of implementing the sliding scale approach, focusing on Judge Crotty’s analysis of the expert evidence in Goldman Sachs. I describe how the DCF methodology could be used and applied in future cases.

I am not suggesting that courts should apply DCF analysis in every securities class action. The DCF methodology will be more useful in some contexts than others, particularly in cases where there is confounding information, leakage, ambiguity about appropriate industry indices, or interpretive challenges related to the stock price decline on an alleged corrective disclosure date. My suggestion is simply that courts begin to accept expert testimony based on DCF as a supplement to event studies, and then over time begin to rely more on DCF analysis on a case-by-case basis going forward. In future research, I hope to provide specific guidance about the circumstances under which DCF analysis is likely to be most helpful, and to provide examples of precisely how the analysis could and should be done.

I. Goldman and the Invitation for “New Ideas”

The Supreme Court’s opinion in Goldman focuses on “price impact” at class certification, so I begin there. Expert witness evidence regarding “price impact” plays an important role in securities class action litigation, stemming from the Supreme Court’s decision in Basic Inc. v. Levinson, and the rebuttable presumption “that an investor presumptively relies on a misrepresentation so long as it was reflected in the market price at the time of his transaction.” The Basic presumption has “particular significance” at the class certification stage, at issue in Goldman, where questions centered on allegations about Goldman’s conflicts of interest and the alleged corrective disclosures on dates when Goldman’s stock price declined.

Investors making a claim under section 10(b) of the Securities Exchange Act of 1934 obtain the Basic presumption if they can show that alleged misstatements were (1) publicly known and (2) material, (3) that the stock was traded in an efficient market, and (4) that the plaintiffs in fact traded the stock between the time of the alleged misrepresentations and subsequent corrective disclosures. Plaintiffs and defendants frequently submit expert reports and testimony regarding these factors. If the plaintiffs satisfy Basic, as they did in Goldman, defendants may rebut the presumption at the class certification stage by showing that the alleged misstatements in fact had no impact on the stock price.

The Supreme Court in Goldman set forth the parameters for defendants submitting price impact evidence: defendants must show a complete lack of price impact, and they face a “preponderance of the evidence” burden. There have been many sources of confusion and disagreement regarding the Goldman litigation over the years, and a sharp divide among lawyers and experts who work on securities class actions about various aspects of the expert witness approaches in the case. Some of the disagreement stems from arguments about the “mismatch” in the case between the plaintiffs’ allegations about alleged conflicts of interest and the information revealed on the dates alleged to be corrective disclosures, when Goldman’s stock price declined. Allegations of generic misstatements present special difficulties at the class certification stage.

But Justice Barrett’s opinion articulated several principles in general terms, in ways that are not limited to cases involving generic allegations at class certification, and these principles arguably apply, in at least some ways, to all securities class actions, at all stages. Moreover, as Justice Barrett noted, the parties and all of the justices agreed in Goldman that “courts may consider expert testimony and use their common sense in assessing whether a generic misrepresentation had a price impact.” Justice Barrett indicated that these general principles hold even though expert testimony about price impact also can be relevant to the merits. Moreover, courts may consider such expert testimony as to price impact even if it also may be relevant to materiality, which, as noted in Amgen, is an issue reserved for the merits.

Justice Barrett’s reference to “common sense,” the money quote of the opinion, comes from Professor Donald Langevoort’s important 2015 law review article, which has been cited in numerous contexts, including the Allstate litigation, which Justice Barrett also references. The quote is sufficiently important that it is worth setting forth the original, in full:

Event studies may help, but there is no reason in the class certification inquiry to limit evidence to those, especially in “confirmatory lie” cases. Courts should be open to all probative evidence on that question—qualitative as well as quantitative—aided by a good dose of common sense.

For class certification, rebutting price impact is understandably difficult for defendants. The presumption of reliance arises because, as Professor Langevoort notes, one can fairly presume that a material misrepresentation or omission would necessarily distort the market price, unless the market somehow already reflected the truth. Notably, all of the justices joined this part of Justice Barrett’s opinion, which includes both the phrase “common sense” and the language “qualitative as well as quantitative” from Professor Langevoort.

The oral argument in Goldman presaged much of the agreement among the justices and the parties as to “common sense” and the notion that judges should consider expert testimony based on a sliding scale related to the extent of the “mismatch” between alleged misstatements and corrective disclosures. Justice Kagan used the phrase “common sense.” Justice Kavanaugh referenced “sliding scale” and the importance of “adjectives” in the litigation. Justice Barrett also referenced “common sense” in asking about expert testimony, which elicited a response that “price impact requires comparing the actual price to what the price would have been had there been no deceit. And so the nature of the deceitful statement is relevant, though not by itself dispositive, to that inquiry.”

These suggestions from oral argument were reflected in Justice Barrett’s opinion. But when the U.S. Supreme Court’s Goldman decision was published, on June 21, 2021, it still remained unclear among practitioners how the lower federal courts might follow several aspects of Justice Barrett’s opinion: her guidance regarding price impact evidence, her emphasis on the importance of assessing the “mismatch” between the contents of a misrepresentation and a corrective disclosure, and her advice that courts should be “aided by a good dose of common sense.” The Second Circuit characterized the Supreme Court’s decision as supplementing the courts’ previous understanding with “new ideas.”

These new ideas were first implemented on December 8, 2021, in an opinion by Judge Paul A. Crotty. This opinion signaled a potentially new judicial approach to securities class actions, particularly with respect to the role of expert witnesses. In Goldman, the plaintiffs’ expert identified three statistically significant price drops immediately following public disclosures. The experts disagreed about the extent to which information revealed at the time of these stock price declines was related to the alleged misrepresentations.

Judge Crotty applied these new ideas to re-certify the class, implementing the Supreme Court’s guidance. As noted above, according to Judge Crotty, the new approach involves assessing the price impact of misrepresentations in securities class actions based on (using his emphasis) a “sliding scale,” where the court analyzes expert testimony to assess any “mismatch between the contents of the misrepresentation and the corrective disclosure.” The new approach presents both theoretical and practical innovations. In terms of theory, establishing such a “mismatch” explicitly calls for the analysis of counterfactuals and causality, as well as the timing and content of disclosures and related price reactions. In terms of practice, the new sliding-scale approach envisions expert witnesses engaging more specifically with the factual aspects of particular disclosures in securities class actions to show that “the evidence and common sense speak as one.”

Judge Crotty’s approach is a potential road map for courts addressing price impact in future cases. He assessed the expert opinions in Goldman and the extent to which their analysis was consistent with “common sense”; he applied the preponderance standard articulated by the Supreme Court (“whether it is more likely than not that the alleged misrepresentations had a price impact”), and then re-certified the class.

As of this writing, it remains unclear how Judge Crotty’s approach will be received. Defendants immediately filed a petition with the Second Circuit, challenging Judge Crotty’s certification of the class, and the Second Circuit granted this petition. In a 2022 decision in the Apple Inc. securities litigation, Judge Gonzales Rogers applied the Supreme Court’s decision in Goldman to certify a class of shareholders (though not option holders), based in part on assertions by plaintiffs’ expert that the price inflation “ribbon” could be constructed using not only event studies, but a “full array of generally accepted and widely used valuation tools can be applied, if necessary, to calculate the but-for stock prices under the assumption of prior full disclosure.” Neither the experts nor Judge Rogers explicitly referenced the DCF approach I suggest here, but the language in the case is arguably capacious enough to include this approach.

To sum up the relevant background, the Goldman litigation has been a dramatic, decade-plus journey. Before the Supreme Court’s decision, it was unclear whether the Basic presumption would continue to apply, or how courts would assess evidence of price impact, or “mismatch,” questions at various stages of securities litigation. There was the possibility that Supreme Court would reject the price maintenance theory entirely; it did not. Confusion about the applicable standards obviously has led to considerable delay and cost for parties. But now the Supreme Court, the Second Circuit, and Judge Crotty have essentially invited the courts to find a more cogent theoretical and practical approach.

II. A New “Sliding Scale” Theory

I next briefly discuss a few theoretical questions that are implicit in the new “sliding scale” approach. The most important theoretical point is that the “sliding scale” test is essentially probabilistic. The central question about the “mismatch” between the contents of a misrepresentation and a corrective disclosure can be framed as a question about probability: how likely is it that a misrepresentation in fact was associated with an impact on the price of the stock, given the evidence associated with the corrective disclosure? Essentially, this question asks how likely it is that a stock price decline was due to the misrepresentation versus some other cause.

Different answers to this question correspond to different states of the world in a probabilistic setting. For example, one answer might be that there is a 0 percent chance that a misrepresentation caused any price impact; another might be that there is a 100 percent chance that a misrepresentation caused the full stock price decline on the date of the corrective disclosure. Other answers might be in between these two poles. One might conclude that there is a 51 percent chance that a misrepresentation caused no price impact, and a 49 percent chance that it caused some non-zero price impact. Such a determination would overcome defendants’ burden at class certification. Alternatively, one might conclude that both the chance that a misrepresentation caused no price impact and the chance that it caused some price impact are equally likely, which would result in certification of the class. (This is the dividing line that generated considerable debate between Justice Barrett and Justice Gorsuch in Goldman.) Of course, other, more specific alternatives are possible, further from the 50 percent level, given the degree of uncertainty about the relationship between the misrepresentation and the stock price decline, and there potentially could be a distribution of possible outcomes.

Framed in such probabilistic terms, one can conceptualize the new sliding scale approach as resembling a natural experiment in which there is a “true” state of the world regarding the price impact of a misrepresentation, but it is not possible to observe this state. Instead, the best scientific approach is to indirectly establish what the likely “true” state of the world is by considering various alternative states of the world, and assessing how likely they are. Fact finders, both judges and juries, implicitly engage in this kind of approach in numerous settings, though the approach is typically not framed in probabilistic terms.

Suppose that a court is considering expert testimony about a corrective disclosure date on which the stock price decline was X%. Expert witnesses might present several different probabilistic interpretations of the extent to which the back-end price drop corresponds with front-end inflation maintained through alleged misstatements. Each interpretation might differ based on the extent to which the alleged misstatements in fact impacted the stock price. As described above, one interpretation might be that the alleged misstatements had no impact on the stock price. Another might be that the alleged misstatements had the full price impact on the stock prices. Others might be that the alleged misstatements had price impact of some other percentage of the stock price drop.

One way to understand the sliding scale approach is to use a classic probability thought experiment: suppose that the possible price impact interpretations are represented by balls in an urn. A ball with a 0 written on it represents a “complete” lack of price impact. A ball with the actual stock price decline, X, written on it represents a price impact that is equal to X%. A ball with some other number represents a price impact of that number. Effectively, what the court is being asked to decide is whether the probability of drawing a ball with a 0 is greater than 50 percent, the preponderance standard.

This discussion and the probabilistic example illuminate an important aspect of expert testimony under the sliding scale approach. Effectively, experts are using statistical techniques, and common sense, to help the court assess the probability distribution of potential price impacts. For example, an expert for the plaintiffs might opine that the distribution of potential price impacts is likely a normal distribution, with a modal price impact of, say, negative 5 percent.

In contrast, an expert for the defendants might opine that the distribution of potential price impacts is likely a skewed distribution with a modal price impact of zero. Experts might consider the extent to which the observation of the corrective disclosure was influenced by various factors. A court applying the sliding scale approach effectively is assessing the facts and expert evidence to determine whether the distribution is more accurately described by the first distribution or the second, or perhaps some distribution in between. This theoretical approach is consistent both with the language of the Supreme Court’s opinion in Goldman and with scientific research. In other words, it is a cogent approach theoretically.

The difficult question is how to apply this approach in practice. To date, the main way for experts to overcome Daubert challenges, and follow this theoretical approach empirically, has been to use event studies. The idea is that event studies can assist the trier of fact in assessing the likelihood that a particular view of price impact is the “true” state of the world. As noted above, scholars increasingly have been criticizing various aspects of the use of event studies in securities class actions. Although those critiques typically are not formed in terms of theoretical causal inference, they effectively are precisely that, undermining the grounds for a judicial conclusion based on event studies that, for example, there is no evidence of price impact.

I next show how the DCF approach could be used as an alternative based on this theoretical model. The DCF methodology could be implemented in ways that are superior to the event study approach in many cases and would supplement the event study approach in others.

III. The DCF Approach

As the Supreme Court explained in Goldman, expert witnesses can opine regarding two categories of price impact: “inflation introduction” means that price impact is measured by the additional stock price inflation introduced by an allegedly false statement, and “inflation maintenance” means that price impact is measured by “the amount that the stock’s price would have fallen without the false statement.” Inflation maintenance, not inflation introduction, was at issue in Goldman.

The analysis of inflation maintenance involves the use of statistical and other expert approaches to the kind of counterfactual causation analysis that has become important in academic research. Specifically, as noted in Part II, the counterfactual “price impact” analysis involves a consideration of what likely would have happened if the defendant had spoken truthfully.

As a practical matter, this analysis typically focuses on one or more “corrective disclosures,” including event studies of the price decline on the corrective disclosure dates, expert analysis of documents and testimony related to the corrective disclosures, and potentially other statistical techniques. For example, Judge Crotty analyzed expert testimony about three corrective disclosures, as well as the lack of abnormal price movement on thirty-six other dates on which there were disclosures of information related to plaintiffs’ claims about news of Goldman’s conflicts of interest. Judge Crotty cited his previous, vacated certification of the class, which found that expert evidence had “at the very least, establishe[d] a link between the news of Goldman’s conflicts and the subsequent stock price declines, and that Defendants had failed to present evidence sufficient to sever that link.” He then noted that the Supreme Court’s clarification that “courts may not end their inquiry there.”

The generic nature of a statement was one practical hurdle to “matching” the statement to a later stock price decline. Judge Crotty described how the “fresh guidance” for courts applies when a generic statement is at issue:

They must also evaluate whether the subsequent inference required for class certification under the inflation-maintenance theory—that back-end price drop corresponds with front-end inflation maintained through alleged misstatements—is fatally undermined by the generic nature of those misstatements, a “mismatch” in genericness between misstatement and corrective disclosure, or other common-sense factors.

Judge Crotty analyzed the experts’ testimony regarding the three statistically significant stock price declines at issue. The crux of this analysis was the court’s inquiry into whether there was a “link” between the public revelations concerning Goldman’s conflicts of interest and the subsequent stock price declines. The key parts of the expert testimony involved questions about whether the three corrective disclosures included news about conduct “for the first time” and whether market commentary evidence was consistent with “match” or “mismatch”: this language echoes the Supreme Court’s suggestion that the question is whether “there is a mismatch between the contents of the misrepresentation and the corrective disclosure.”

In this context, Judge Crotty described the “sliding-scale test” in various ways, based on the Supreme Court’s use of phrases such as “starts to break down,” “less likely,” and “less reason.” Judge Crotty emphasized in italics the terms “starts” and “less.” He also stated that the Supreme Court’s “mismatch” guidance is consistent with certifying a class based on “narrower” corrective disclosures that only correct a portion of the prior inflation in the stock price. In other words, although damages ultimately might be impacted by such narrowness, corrective disclosures that “poke targeted, meaningful holes in overarching impressions reinforced through broader prior statements” can support class certification, especially where there is compelling evidence of price impact. Questions about generic versus specific disclosures are relevant because of the potential mismatch: “it is less likely that the specific disclosure actually corrected the generic misrepresentation, which means that there is less reason to infer front-end price inflation—that is price impact—from the back-end price drop.”

Note how the event study methodology itself is not particularly useful in informing Judge Crotty’s above inquiries. The statistics on their own do not help “match” (or not “match”) the contents of any corrective disclosure associated with the observed back-end stock price drop to the inflation. Instead, what Judge Crotty emphasized was the analysis of news coverage and commentary surrounding the alleged corrective disclosures, which he found “convincingly links Goldman’s post-disclosure plight back to the alleged misstatements.” That is certainly using “common sense.” The question is whether this kind of analysis might be done more precisely. The answer is yes, using DCF analysis.

What if DCF analysis is added to the above discussion, meaning that expert testimony would include DCF analysis at various points in time? Then, when Judge Crotty was examining particular dates, the questions would not simply have been about the combination of statistical and common sense interpretations of the information. Instead, the analysis would have included details about the extent to which the corrective disclosures related to information that would have impacted expected cash flows and discount rates at earlier dates. Expert analysis of expected cash flows and discount rates would generate helpful information, transparency, and insight not provided by event studies. If the experts in Goldman had included DCF analyses, they would have reached separate opinions about that impact, and Judge Crotty would have had competing assessments (to repeat the earlier metaphor, looking under the hood of the car) regarding the extent to which particular revelations of information impacted expected cash flows and discount rates. This kind of analysis would have been far more transparent, and also squarely within areas where experts might have helpful expertise, than simply asking whether news stories might be relevant to a particular stock price decline, as was done with the event studies.

In other words, because DCF analysis involves valuation techniques based on expected cash flows and discount rates, it provides an alternative for the court to assess the extent to which a decline in value after a corrective disclosure “matched” allegations in the case. For example, Goldman’s stock price declined from $184.27 on April 15, 2010 to $160.70 on April 16, 2010. Various approaches using the event study methodology might conclude that this stock price decline was statistically significant, or not. However, the event study methodologies used in Goldman did not prove particularly useful in helping to determine whether there was no “price impact” for class certification purposes. Instead, the experts made various attempts to show lack of price impact based on other information, none of which were persuasive to Judge Crotty.

Instead, suppose the experts constructed DCF valuation analyses to support their analysis of the difference between (1) a $184.27 valuation on April 15, 2010 and (2) a $160.70 valuation on April 16, 2010. The experts could assess the extent to which different changes in expectations related to future cash flows or discount rates were matched, or not, with the allegations and the stock price change. Specifically, the experts would adjust cash flow and discount rate assumptions in various ways that would have been consistent with the actual valuation changes, and then test and assess those adjustments based on a range of available data, just as experts test and assess DCF valuations in other contexts. (For example, experts in appraisal proceedings employ different cash flow and discount rate assumptions to help judges make an assessment of “fair value” and then compare it to the actual deal price.)

Similarly, such a DCF approach could be used, not just at class certification, but at the merits stage of securities class action litigation. Experts could opine regarding materiality based on DCF valuations by showing whether the factors that changed in their DCF valuation (such as expected cash flows) rose to the level of factors that a reasonable investor would find important in the total mix of information. Experts could opine regarding loss causation based on DCF valuations by analyzing the impact of particular disclosures on expected cash flows or discount rates in their DCF models. Experts could opine regarding damages based on DCF valuations by showing what a reasonable price inflation “ribbon” would be based on changes in the valuation of the company as information about the alleged misstatements emerged. As noted above, in order to construct a price inflation “ribbon,” the experts could create both an actual and counterfactual DCF as of the beginning of the class period, and then trace changes in valuation based on the DCF model over time, based on changes in disclosures, analyst and investor expectations, and other economic factors related to expected cash flows and discount rates.

The above methodological description is not new. DCF analysis already is the “gold standard” for expert analysis in a range of cases, including in Delaware. DCF methodologies are regularly taught in courses, used in peer-reviewed research, applied by consultants and experts, and used by practicing investment bankers and lawyers. DCF valuations are a standard part of the curriculum in various educational programs. The point here is that DCF analysis also could be part of expert approaches in securities class actions.

Of course, DCF methodologies will not always be persuasive, in any context, including this one. Experts likely will disagree about forecasts and inputs, discount rates, and other variables. In some contexts, the use of DCF analysis could result in more battles of the experts, and greater costs. But even in these cases (or perhaps especially in these cases), the use of DCF analysis would better enable judges to make the kinds of “common sense” evaluations about class certification and the elements of securities class actions, in intuitive ways based on factors relevant to valuation, instead of based on less intuitive expert disagreements about statistical power and factual assertions related to event studies that are unmoored from intuition. As noted above, event studies and DCF analyses are closely related theoretically, and one way to view the use of DCF analyses in securities class actions is simply to make more transparent what experts are attempting to accomplish with event studies alone.

It is not clear which parties, plaintiffs or defendants, would benefit from the use of the DCF methodology in securities class actions. For example, defendants might seek to use the DCF approach to rebut the presumption of price impact at the class certification stage, by showing that changes in the assumptions of the DCF model that would have resulted from corrective disclosures were likely minimal at the time, so that any market price impact likely was due to information other than the revelation of the truth about alleged misstatements or omissions. As the defendants asserted in their reply brief in Goldman, “of the more than 2,000 securities class actions filed since Halliburton II, defendants have rebutted the presumption by showing no price impact in only five cases—and almost never in inflation-maintenance cases.” It is hard to imagine that the DCF approach would have a lower success rate. Plaintiffs also likely would introduce responsive expert testimony based on DCF analysis, but, in at least some cases, DCF analysis could be enough to tip the scales in favor of defendants at class certification.

DCF analysis is perhaps more likely to be influential at later stages, after the certification of a class, when both parties could use DCF methodologies to illuminate their disagreements about “mismatch.” To repeat, the basic task for the experts would be to start with a DCF analysis based on the stock price after the corrective disclosure and then work backwards to determine the amount of inflation throughout the class period. Experts would seek to demonstrate when there was a mismatch between the contents of misrepresentation and corrective disclosure, so that the inference “starts to break down.” Essentially, the idea would be to supplement strictly stock price–based analysis in event studies with more specific company-based valuation analysis in DCF analyses. In other words, the relevant question for the experts would be, not how a disclosure might have changed the stock price, but how a disclosure changed the DCF model, assumptions, and inputs.

Suppose that in Goldman an expert’s counterfactual analysis examined the effects on a DCF valuation if Goldman had disclosed more detail about its conflicts at the beginning of the class period, including details sufficient to infer at the time that the factual allegations that ultimately were disclosed from the SEC investigation were true. What would the counterfactual DCF analysis look like? Plaintiffs presumably would argue that the DCF cash flow projections would be significantly lower as a result, or that the relevant discount rate would be higher. An expert would need to support that argument with evidence, perhaps from investor and analyst reports, statistical analyses, or even other cases or analyses. In contrast, defendants presumably would argue that the DCF cash flow projections and discount rates would be essentially the same, even assuming specific disclosures about conflicts of interest. An expert could support that argument by analyzing evidence that these kinds of disclosures would not have been likely to influence assumptions about expected cash flows and discount rates. This kind of analysis could enable the judge to use “common sense” in reaching conclusions about how much of a change in the DCF variables likely would have occurred given the counterfactual. This conclusion could be challenging, depending on the facts, but it would be more transparent and direct than the current comparisons of experts’ event studies.

Competing expert reports based on DCF analyses would then put two very different interpretations of the valuation of the defendant company at various dates in front of the judge, in an intuitive way. The judge could then ask “common sense” questions to determine which DCF conclusion is more plausible. Which changes in expected cash flows seem most likely to be associated with disclosures regarding conflicts of interest? How much would such disclosures have changed the discount rate that reasonably would be applied to those cash flows? Both plaintiffs and defendants could use reliable methodologies based on the DCF approach, suggesting that both might survive Daubert challenges if they were implemented appropriately. The analyses likely would do more than simply break out various pieces of information in the corrective disclosure and then assert how they impacted price, as in an event study. Instead, they would transparently show what a particular approach to DCF valuation would assume that the cash flow and discount rate impact likely would be, and then test those conclusions, so that the judge could then assess the plausibility of the interpretations directly.

In cases where the allegations relate to the variables that are typically used in a DCF analysis, the DCF approach would be particularly apt. Suppose based on a company’s disclosures that investors expected it to generate $100 million of EBITDA annually, and that it had a $200 million market capitalization. In fact, if investors knew the truth, they would expect EBITDA of $50 million annually, and the company would have a $100 million market capitalization. Suppose further that, after some time passed, there was a corrective disclosure and investors revised their expectations, and the company’s market capitalization declined to $100 million. Instead of simply analyzing the corrective disclosure and price decline at the time based on an event study, and then attempting to associate that decline with allegations about misstatements and omissions, the experts might instead construct a set of assumptions based on the respective EBITDA amounts, and determine the likely associated projections associated with the respective DCF valuations. DCF analyses in such a case are arguably much more helpful than event studies.

In other contexts, the use of the DCF approach might be more challenging, and yet still more helpful than event studies alone. Suppose the same details as above, but on the date of the corrective disclosure, the company simultaneously disclosed a positive earnings surprise. An event study might show that the stock price decline on this date was statistically insignificant. Experts might opine that there was a “mismatch” between the misstatement that allegedly led to stock price inflation maintenance and the stock price decline. In contrast, the DCF approach would focus on the difference between the value of the company at the different times, essentially producing point estimates of value that then could be used to establish the price inflation “ribbon.” The DCF approach thus would enable parties to assess more directly the impact of alleged misstatements, apart from conflating information. Moreover, even when forecasting cash flows overall might be challenging, experts would only need to value the differentials in cash flows at different times based on the allegations; other difficult-to-value cash flows would effectively cancel out to the extent they were unaffected by the alleged misstatements.

The DCF analysis is likely to be more valuable in certain settings than others, as has been the case, for example, in Delaware appraisal proceedings. Courts might usefully distinguish between cases involving accounting misstatements versus cases involving estimates related to particular industries (e.g., mortgage banking or pharmaceutical manufacturing) versus cases involving generic statements (e.g., conflicts of interest). The wisdom and experience of the Delaware courts might be a useful starting point for federal judges applying the DCF methodology in securities class actions.

The DCF methodology is not without challenges. DCF analyses that use terminal values pose questions about how the terminal value is calculated, and to the extent event studies are used as part of such calculations the DCF methodology could pose similar questions to those posed by event studies. The use of the DCF methodology also raises questions about what an appropriate range of assumptions is in the relevant models. How should one calculate “beta” in the capital asset pricing model context, or otherwise derive an appropriate discount rate, or range of discount rates? Are the errors in assumptions about cash flows in discount rates likely to be normally distributed, or have errors that suggest estimates will be reasonably correct, on average? Or are there likely to be systematic skews or biases in the analysis?

Courts could approach such questions by isolating particular factors in expert testimony. For example, one approach might be to assume a particular discount rate or “beta” applies in the analysis, and then ask experts to debate valuation based on different cash flow estimates. Alternatively, one could assume that the cash flow differences are fixed and then consider different discount rate estimates. Courts also could recognize that statements about generic issues, perhaps such as statements related to conflicts of interest in Goldman, could be less likely to be directly related to expected cash flows or discount rates than statements related to the expected cash flows themselves. In other words, the use of DCF analyses might, on their own, make cases that involve generic misstatements less attractive for plaintiffs, because it would be more difficult to substantiate a conclusion that a corrective disclosure would have significantly changed any point estimates of cash flows or discount rates. This approach arguably would be more principled than simply carving out a jurisprudential exception for generic misstatements.

Notwithstanding the above challenges, there are no obvious barriers to parties implementing the DCF approach in particular cases, either as a substitute for event studies, or in addition to them. Lawyers litigating such cases could simply ask experts to do so, and then see how judges react. This article recommends that lawyers should try this new approach, following Goldman, and it suggests that judges should be open to such “new ideas.”


The Goldman class action securities litigation stemmed from alleged misstatements and omissions from more than a decade ago. During that time, several academic researchers have offered significant criticism of the use of event studies in securities class action litigation, and the district court in Goldman specifically criticized the use of event studies on remand and suggested a new approach. The challenges in Goldman plus the challenges from academia suggest that an alternative approach to event studies is needed.

My suggestion, consistent with Judge Crotty’s approach, is for courts to consider expert testimony based on DCF analysis as that alternative. Event studies are useful for many purposes, but they also can lead to an overemphasis on statistical detail over causal inference. The DCF methodology is reliable and also potentially more transparent and intuitive than events studies. In other words, the DCF methodology is more consistent with the notion that judges should look to “common sense” in adjudicating securities class actions after Goldman.

I am grateful to Ken Ayotte, Andrew Baker, Robert Bartlett, Bobby Bishop, Matthew Cain, Albert Choi, Madison Condon, Brad Cornell, Jay Dubow, Jennifer Fan, Jill Fisch, Jonah Gelbach, Zohar Goshen, John Griffin, Kevin Haeberle, Christine Hurt, Robert Jackson, David Katz, Bill Lafferty, Donald Langevoort, Ann Lipton, Dorothy Lund, Brett McDonnell, Joshua Mitts, Simone Sepe, Cary Shelby, Omari Simmons, Leo E. Strine, Jr., J.W. Verret, and Lori Will for helpful comments, including comments at the 2022 Tulane Corporate and Securities Roundtable.