This research report is based on a preliminary analysis of class action securities fraud litigation filed during calendar year 1996. We expect that some of our findings will change as we expand our database and continue to refine our analyses. We also believe that certain trends observed during the first year of litigation will not continue in the future. We therefore caution all readers to recognize the preliminary nature of these results and to check at http://securities.stanford.edu for more recent analyses, if available.

We also caution readers of this report that the information on which it is based may be materially incomplete in several respects. There is no centralized depository of information regarding federal or state class action securities fraud litigation, and we recognize that we have likely missed some litigation, particularly at the state level outside of California. We are most grateful for any information and copies of filings, briefs, orders, or other litigation material that appear not to be reflected in this report, and for any suggestions as to how this report can be improved. Please direct all such communications by post to Associate Director, Securities Litigation Project, Stanford Law Library, Stanford, CA 94305-8612.

The Private Securities Litigation Reform Act of 1995 became effective on December 22, 1995, after Congress for the first time overrode a veto by President Clinton. Opponents of the legislation characterized it as providing a "license to lie." Its supporters claimed that it provided necessary protection against meritless litigation.

This report compares patterns of class action securities fraud litigation in federal and state courts since the Act's effective date with litigation patterns observed prior to the Act's adoption. Although our analysis is based on partial data and is subject to further amendment and modification, we can report ten significant preliminary findings. As discussed in greater detail in the body of our Report, many of the policy implications of these findings are subject to dispute and interpretation. It is prudent to await further information regarding the operation of the Act before reaching any conclusions regarding the Act's "success" or "failure."

Our significant preliminary findings are as follows:

Overall litigation rates are little changed.

Our best estimate is that class action securities fraud litigation in federal and state court is being filed at an annual rate of 148 to 163 defendant issuers per year. Prior to the Reform Act, litigation was being filed at a rate of approximately 176 defendant issuers per year. The total volume of litigation activity in 1996 is thus down by about 7% to 16%, but is not very different from the level of activity observed in 1991, 1993, and 1995. In addition, increasing stock market prices in 1996 may have depressed litigation activity. It is therefore too soon to draw any firm conclusions as to whether litigation reform has had any material effect on the aggregate securities class action litigation rate.

About 26% of litigation activity has moved from federal to state court.

The relative stability of the aggregate litigation rate masks a significant shift of activity from federal to state court. Approximately 26% of class action claims are state court proceedings without parallel federal claims filed in 1996. This increase in state court litigation is likely the result of a "substitution effect" whereby plaintiffs' counsel file state court complaints when the underlying facts appear not to be sufficient to satisfy new, more stringent federal pleading requirements, or otherwise seek to avoid the substantive or procedural provisions of the Act. Plaintiffs may also be resorting to increased parallel state and federal litigation in an effort to avoid federal discovery stays or to establish alternative state court venues for the settlement of federal claims.

Allegations of accounting irregularities or trading by insiders now explain the lion's share of federal class action litigation.

Approximately 67% of post-Reform Act Section 10(b) complaints involving publicly-traded companies allege accounting fraud as a basis for liability. In sharp contrast, similar allegations are found in only 34% of pre-Reform Act cases. Allegations of trading by insiders now appear in about 57% of post-Reform Act cases, whereas these allegations are found in only 21% of pre-Reform Act cases. Alleged trading by insiders is particularly important in cases against high technology companies, appearing in 73% of those cases, but that statistic must be interpreted with caution because of the prevalence of option-based compensation in the high technology sector.

Pure "false forecast" cases explain a relatively small percentage of pending Reform Act claims.

Complaints alleging false forward-looking statements as the sole basis for liability account for only about 14% of all post-Reform Act complaints analyzed, and only about 6.5% of post-Reform Act federal complaints involving publicly-traded companies.

Litigation typically follows larger price declines than observed prior to the Reform Act.

Prior to the Reform Act, the average stock price decline preceding the filing of a claim was about 19%. During 1996, the average decline jumped to 31%. This increase is consistent with the observation that heightened pleading requirements induce plaintiffs' counsel to pursue cases that are correlated with larger price declines, and therefore seem to be more apparent instances of fraud.

Federal claims are now rarely filed against the largest issuers.

The average company sued in a federal securities fraud class action in 1996 had a market capitalization of $529.3 million. Prior to the Reform Act, the average market capitalization was $2,080 million. This decline appears to be attributable almost exclusively to a reduction in litigation naming issuers with market capitalization in excess of $5.0 billion. Prior to the Reform Act, these large corporations represented about 8.4% of federal court activity, but very few of these companies appear to have been sued in 1996.

High technology issuers continue to be the most frequent targets of class action litigation.

High technology companies represent 34% of all issuers sued in federal court since the effective date of the Reform Act. That statistic is not materially different from the pre-Reform Act experience.

The dominant plaintiffs' class action law firm, Milberg Weiss Bershad Hynes & Lerach, appears to have increased its significance nationally and in California in particular.

Milberg Weiss' appearance ratio nationwide stood at approximately 31% prior to the Reform Act. Aggregating parallel federal and state activity, Milberg Weiss' appearance ratio today stands at about 59% nationwide and 83% in California. Milberg Weiss' increased significance can be explained by the fact that it is likely the best capitalized plaintiffs' firm and therefore best able to finance the delays associated with slower procedures under the Reform Act. It also has the most diversified portfolio of plaintiffs' claims and is therefore better able to absorb the risk associated with litigation under the new regime. In addition, it is best situated to internalize the externalities associated with the need to invest to create new precedent interpreting the Reform Act's novel provisions.

In the courthouse, judges appear to be resolving legal questions regarding the interpretation of the "strong inference" requirement in favor of plaintiffs.

The most frequently litigated issue to date--the interpretation of the "strong inference" pleading requirement--has with but a single exception been uniformly interpreted to apply the Second Circuit standard, not some higher pleading requirement. This is the position espoused by plaintiffs. Moreover, no complaint subject to the "strong inference" pleading standard has been dismissed without permitting plaintiffs the opportunity to replead a material portion of the claims asserted in the original complaint.

The growth of parallel state and federal litigation, with concomitant disputes over discovery stays and other matters, suggests that attention to federal preemption issues is warranted.

In addition to the growth in "pure" state class action fraud claims, at least 28% of federal class action securities fraud cases also have pending parallel state securities fraud class action claims. Parallel state court securities fraud class actions were quite rare prior to the Reform Act. This parallel litigation appears to be brought to avoid the Reform Act stay on discovery, and perhaps for other strategic and settlement-related reasons as well. This boom in state class action securities fraud litigation raises issues regarding the optimal coordination of federal and state litigation regimes and suggests that a systematic review of the issue by Congress is in order.

I. INTRODUCTION AND OVERVIEW

II. LITIGATION RATES

Federal Litigation Rates

State Litigation Rates

III. CHARACTERISTICS OF THE COMPANIES SUED

Market Capitalization

Stock Price Declines Associated With Litigation

Beta of Defendant Issuers

The Incidence of Litigation by Industry

IV. THE NATURE OF FRAUDS ALLEGED

V. THE INCIDENCE OF ALLEGED TRADING BY INSIDERS

VI. THE INCIDENCE OF CLAIMS ALLEGING FRAUD IN INITIAL OR FOLLOW-ON PUBLIC OFFERINGS

VII. APPEARANCE RATIOS OF PLAINTIFF LAW FIRMS

VIII. THE GEOGRAPHIC INCIDENCE OF LITIGATION

IX. SETTLEMENTS

X. DECISIONS INTERPRETING THE REFORM ACT AND RELATED MATTERS

Interpretation of the "Strong Inference of Fraud" Pleading Standard

Other Motions to Dismiss

The Scope of the Discovery Stay on a Motion to Dismiss

Conflicts Between State and Federal Court

Adequacy of Notice and Appointment of Lead Plaintiff

Survival of the Recklessness Standard for Scienter

The Safe Harbor for Forward-Looking Information

Retroactive Application of the Reform Act

Cognate Decisions

XI. CONCLUSIONS AND POLICY ISSUES

APPENDIX A. COMPANIES NAMED IN FEDERAL SECURITIES FRAUD CLASS ACTIONS

APPENDIX B. COMPANIES NAMED IN STATE SECURITIES FRAUD CLASS ACTIONS

APPENDIX C. COMPLAINTS CONTAINED IN DATABASE

APPENDIX D. DESCRIPTION OF REFORM ACT DATABASE SUBSAMPLES

The Private Securities Litigation Reform Act of 1995 (the "Reform Act" or the "Act") took effect on December 22, 1995,¹ after Congress for the first time overrode a veto by President Clinton. The Act put in place a variety of procedural and substantive hurdles aimed at curtailing what its proponents considered abusive practices in class action securities fraud litigation. For example, the Act creates a heightened pleading standard that generally makes it more difficult for plaintiffs to file allegations of securities fraud without having solid information beforehand on which to base such a claim. Hand-in-hand with this provision, the Act also provides for a stay of discovery while a motion to dismiss is pending. With the stay, Congress sought to prevent plaintiffs from building their case solely on information provided by defendants during discovery and to address the concern that in the past, innocent defendants may have been induced to settle meritless claims simply to avoid the high cost of discovery.

To address concerns about the influence of "professional plaintiffs" and class action attorneys, the Act contains a "lead plaintiff" provision and class notification process aimed at giving the plaintiffs with the largest financial interest at stake (presumably, institutional investors) the right to control the course of the litigation and to select lead counsel for the class, subject to court approval. The Act also creates a limited "safe harbor" for the release of forward-looking information about a firm's prospects. Congress provided the safe harbor to prevent companies from being subjected to class action securities litigation simply because their forecasts proved inaccurate, regardless of any proof of intent to mislead or the presence of appropriate cautionary language.

These and many other provisions of the Reform Act were hotly contested by plaintiffs' counsel, issuers of publicly traded securities, accounting firms, the Securities and Exchange Commission, state securities regulators, and many other constituencies.² Opponents of the legislation bemoaned its passage and questioned the extent of the perceived problems. They predicted that the United States' securities markets would become a magnet for fraud and that meritorious litigation would no longer be heard in federal courts.

This article is a first attempt at an empirical analysis of the litigation patterns observed since the adoption of the Act, and a comparison with experience that preceded the Act's adoption. In particular, this article focuses on the changes in the number and nature of actions commenced in the year following passage of the Act (December 22, 1995 through December 31, 1996), as well as judicial interpretation of the Act through January 31, 1997.³

This twelve month period is short given the historic life cycle of securities fraud litigation.⁴ Indeed, procedural delays associated with some of the Reform Act's more novel provisions such as the class notification process, lead plaintiff selection procedures, and stays on discovery are likely to slow the litigation process further. In addition, disputes over the proper interpretation and application of many of the Reform Act's innovative provisions--such as new pleading standards, discovery stays, and the scope of the safe harbor--are already generating uncertainty and will further slow the litigation process, at least until generally accepted precedent evolves.

The data that underlie this report are therefore, by their very nature, preliminary and suggestive, and can describe only the earliest phases of the litigation process. The data do not, to any significant degree, describe the outcome of motions for summary judgment, settlements, or verdicts because relevant data are either unavailable or too sparse to support a reliable inference.⁵ This research will, however, be updated on a regular basis to reflect ongoing experience with the Act, including the progress of litigation, settlements, and verdicts. We will also introduce more detailed statistical analyses of the characteristics of the companies being sued. Updates on this research can be found at the following Internet address: http://securities.stanford.edu.

This report seeks primarily to measure and to explain observed patterns in litigation behavior. It does not offer policy judgments regarding the desirability of observed trends or of the merits of litigation filed under the new Act. The debate over class action securities fraud litigation often has been heated, and Senator Dodd has observed that conflicting constituencies often are unable to agree even on "the basic facts."⁶ We hope that this report and its successors can serve as an objective point of reference that provides common factual ground for all parties in this continuing debate.

The remainder of this report is organized as follows: Sections II through IX compare the pre- and post-Reform Act experience in terms of litigation rates, characteristics of companies that have been sued, nature of frauds alleged, incidence of claims alleging trading by insiders, incidence of claims alleging fraud in initial or follow-on public offerings, the role of plaintiff law firms, the geographic incidence of litigation, and settlements. Section X discusses judicial interpretation of the Act. Section XI provides concluding comments and a brief discussion of the policy issues raised by this review.

Perhaps the most basic question regarding the impact of the Reform Act is whether it has caused a decline in class action securities fraud litigation. To measure changes in litigation activity, we focus on the number of issuers sued rather than the number of class action complaints filed.⁷ In a typical class action securities litigation, a single issuer will be sued in multiple complaints filed by different named plaintiffs represented by different plaintiffs' law firms. These complaints are generally consolidated and litigated as a single proceeding. The number of issuers sued is therefore a superior predictor of the volume of post-consolidation litigation, of the costs imposed by litigation, of recoveries by plaintiffs, and of the number of dispositions either by dismissal, settlement, or verdict.

Because the Reform Act is a federal legislative initiative that may or may not preempt state law, plaintiffs may attempt to avoid the heightened pleading requirement and other provisions of the Reform Act by seeking alternative remedies through state court proceedings. If these alternative remedies are equally effective, we should observe a substitution effect that simply shifts litigation from the federal forum to state court without reducing the aggregate incidence of filings. To the extent that state court causes of action are less attractive than federal claims, the substitution effect may be correlated with a decline in the aggregate volume of litigation. The Reform Act's effect on overall litigation rates can therefore be measured only by aggregating post-Reform Act litigation activity at the federal and state level and comparing that litigation rate with aggregate federal and state activity prior to the Act.

The aggregate litigation rate must also distinguish between issuers sued in state court only--cases which provide strong indicia of substitute litigation--and issuers sued in parallel federal and state court proceedings, a pattern which may be characteristic of litigation strategies unrelated to a pure substitution effect. In addition, issuers can be sued in state court on derivative claims which are not subject to the provisions of the Reform Act. The state court data must therefore be carefully filtered to identify only those claims and causes of action which are either substitutes for or parallel to claims that could otherwise be asserted as federal class action securities fraud claims.

The data described below suggest that the Reform Act has at most only modestly reduced the aggregate rate of securities class action fraud litigation, but has induced a material substitution effect that may have shifted weaker claims to state court. The data are also consistent with the hypothesis that the Reform Act and the Supreme Court's recent decision in Matsushita v. Epstein⁸ may have created strategic incentives to file parallel state and federal actions against a company in order to gain advantage in discovery and settlement.

Federal Litigation Rates

Between December 22, 1995, and December 31, 1996, at least 109 companies were named as defendants in securities fraud class actions filed in federal court. Appendix A lists the companies sued in federal complaints and notes the dates of the first identified filings and the courts in which the litigation was brought. Cases were included in the database if we were able to obtain information concerning the filing date, the parties, the court in which the action was filed, and the nature of the allegations.⁹ Table 1 presents the total number of companies sued per calendar quarter in 1996 and shows that the filing rate of sixteen in the first quarter of 1996 was approximately half the average quarterly rate of thirty-one recorded in the remaining three quarters of the year.

The low filing rate in the first quarter of 1996 appears to be the consequence of both an inventory effect and a "learning curve" effect. Anecdotal evidence suggests that there was a significant increase in the filing rate for securities class action litigation in the period just prior to passage of the Reform Act, as plaintiffs' attorneys sought to avoid application of the Act to cases they were preparing to file.¹⁰ This phenomenon would have decreased the number of securities class action cases filed during the first quarter of 1996 if cases that would have otherwise been filed during that period were instead filed in December 1995 to beat the Reform Act's effective date. The relatively low filing rate in the first quarter may also be the result of a learning curve effect, whereby plaintiffs' attorneys delayed filings until they had analyzed how to best plead cases under the Reform Act's new provisions and resumed activity once they had developed pleading strategies that they believed were more suitable to the new statutory environment.¹¹

The filing rate that prevailed during the last three quarters of 1996 is therefore likely a better predictor of the overall post-Reform Act litigation rate. Extrapolating from the last nine months of 1996 suggests an annual rate of 124 companies named in federal securities fraud class actions.

Obtaining a measure of the pre-Reform Act filing rate, however, is not so straightforward.¹² The traditional source of statistical information concerning overall civil and criminal filing rates in federal court are the yearly statistical abstracts of litigation activity published by the Administrative Office of the United States Courts. The legislative history to the Reform Act contains Administrative Office data suggesting that for fiscal years 1990 and 1991, 315 and 299 securities class actions were filed, respectively.¹³ These statistics are not reliable, however. The Administrative Office data are generated from tabulations of Civil Action Cover Sheets that are completed by the counsel filing the complaint. It is common practice for more than one complaint to be filed against a company alleged to have committed a fraud and for those complaints to be later consolidated into a single action. In some situations as many as fifteen complaints may be filed against a single company, and all may later be consolidated.¹⁴ To the extent that each counsel filing a complaint that is later consolidated into a single action also files a Civil Action Cover Sheet identifying a securities fraud class action, the Administrative Office data may overstate the volume of litigation activity by including duplicative lawsuits in the count.

At the same time, however, Administrative Office data can also understate litigation activity because plaintiffs' counsel can fail to identify their claims accurately on the Civil Action Cover Sheet as securities fraud class actions. A recent Federal Judicial Center study of all terminated class action litigation in four district courts from July 1, 1992, to June 30, 1994, concluded that the Administrative Office data tended to undercount substantially the amount of class action litigation.¹⁵ There appear to be no reliable techniques that can be applied to the Administrative Office data to generate accurate estimates of the number of issuers that historically have been sued in federal courts alleging class action securities frauds. We therefore do not rely on these data.

Instead, for estimates of baseline data describing the volume of litigation activity prior to passage of the Reform Act, we turn to two surveys that tabulate different measures of federal class action securities fraud litigation activity but generate highly consistent measures of average litigation rates over the five year period spanning 1991 through 1995. First, as described in Table 2, a survey of federal court filings of security class action suits found a range of 153 to 220 filings per year, with an average annual litigation filing rate of 177 filings per year. Second (also in Table 2), a survey of total dispositions of federal class action securities fraud litigation shows a range of 138 to 220 dispositions per year, with an average of 176 dispositions per year over the 1991-1995 period. While annual disposition rates can differ dramatically from annual filing rates because of unstable lags in the litigation process, in a steady state the number of filings per year will, on average and over time, equal the number of dispositions of securities fraud class actions per year, whether by settlement, motion to dismiss, summary judgment, jury verdict, or other means. Taken together, the two surveys support an estimate of 176.5 filings per year in the five years prior to the effective date of the Reform Act.

At this stage, the natural inclination is to compare the volume of 1996 litigation reported in Table 1 with the baseline data reported in Table 2 and conclude that the volume of litigation has decreased from approximately 176 companies per year to either: (a) 109 companies per year (a 38% decline) based on observed annual filings; or (b) 124 companies per year (a 30% decline) based on an annualized litigation rate that ignores the low-level first quarter activity due to potential inventory and learning curve effects. Either interpretation of the data is fundamentally flawed because much of this decline appears to be the result of a substitution effect from federal to state court proceedings.

State Litigation Rates

It has historically been more profitable for plaintiffs to initiate class action securities fraud litigation in federal court rather than in state court. Counsel with substantial experience in litigating securities fraud matters suggest that the volume of class action securities fraud litigation in state court has, until passage of the Reform Act, been de minimis.¹⁸ Four recent developments, however, suggest an increased incentive to initiate class action securities fraud litigation in state court either instead of or in addition to federal litigation.

First, the new pleading requirements, rules governing joint and several liability, discovery stays, and other provisions of the Reform Act impose costs on plaintiffs that can potentially be avoided in state court. To the extent that the Act shifts the relative profitability of class action litigation in favor of state court, there should be a substitution into state court from federal court. For the defendant, however, the burdens and costs of defending against a state law claim are likely just as large as those in a federal action, and the size of a settlement or judgment can also be just as great.

Second, the automatic stay of discovery contained in the Reform Act provides an incentive to file a parallel state law action as a means to avoid the federal stay. Recent litigation suggests that this incentive is in fact at work. State actions filed for this reason are not, however, a measure of increased litigation activity, rather they are evidence of a new litigation strategy. Any attempt to measure the effect of the Reform Act on class action securities fraud litigation activity must therefore be careful to distinguish between state cases that represent new claims against companies not otherwise sued in federal court and parallel claims that are brought primarily for strategic advantage. Nonetheless, parallel state actions represent an increase in the costs of litigation to the extent that the Act provides an incentive to file duplicative actions.

Third, the Supreme Court's recent decision in Matsushita v. Epstein¹⁹ may strengthen the incentives to file substitute or parallel claims by establishing that state court settlements may discharge federal securities law claims that could otherwise only be brought in federal court. A plaintiff who seeks to avoid the new settlement procedures adopted in the Reform Act, who prefers not to deal with the lead plaintiff designated in the federal action, or who otherwise concludes that a more favorable settlement may be available in the state forum, may attempt to resolve all federal and state claims through state procedures that may not respect the Reform Act's innovations.

Fourth, the prospect of passage of California's Proposition 211, a ballot initiative that sought to establish California state securities laws that were substantially more plaintiff-favorable than current federal law, may have also increased the incentives to commence litigation in California state court.²⁰ With the defeat of Proposition 211 in November of 1996, this incentive to file in California state court is not currently present. The pendency of Proposition 211 may therefore be the cause of a transitory and non-recurring increase in the volume of California state litigation activity.

Because plaintiffs are not required to disseminate notice of the filing of state court securities class actions, it is difficult to obtain accurate information on the number and nature of these state court cases.²¹ As a result, the data upon which this report is based may significantly undercount the actual number of state court filings, particularly for state court cases outside of California. For this reason, we do not attempt to draw any conclusions from these data concerning seasonal filing rates or trends in state court filings. Nonetheless, the available evidence suggests a significant increase in state litigation activity over the de minimis levels of prior years.²² Table 3 shows that from December 22, 1995, to December 31, 1996, sixty-nine companies were sued in securities fraud class action lawsuits filed in state court.²³ Of these, thirty-nine were sued solely in state court with no parallel federal complaint in 1996. Although some of this increase in activity may be due to factors like Proposition 211, it appears that a significant portion of this increase is in some way attributable to the Act.

Table 4 describes actual and estimated class action securities fraud litigation rates in federal and state courts. We have observed a total of 148 companies actually sued in state or federal court in the study period, after eliminating double-counting due to the presence of parallel state and federal proceedings. Annualizing the April through December federal litigation rates yields an expected total of 163 companies sued in federal and state court. These figures represent a decline of 7% to 16% when compared to the average number of filings and dispositions in the period from 1991 through 1995. However, the 1991 though 1995 data demonstrate a significant range, and the 1996 data are not materially different from the data for 1991, 1993, and 1995, when filings were 153, 158, and 162, respectively.

It is too soon to draw any firm conclusions from these data with respect to the effect the Reform Act has had on the aggregate number of securities class action lawsuits filed per year. The difficulty in drawing such conclusions is compounded by the difficulties associated with obtaining a complete census of state court class action activity. Nonetheless, these data suggest that the Reform Act appears to have had a modest effect at best on aggregate securities litigation activity. The more significant effects associated with the Reform Act appear to be the substitution effect that has shifted the venue for much of this litigation from federal to state court and the newly created incentives to file parallel litigation in state court.

In addition to questions concerning the Act's effect on the level of securities litigation activity, another important question is whether the Act has caused any material decline in the number of "meritless" lawsuits. Direct evidence on this issue may be difficult to obtain, particularly in light of disputes between plaintiff and defendant constituencies as to whether any individual claim has merit. Nonetheless, an investigation of the market characteristics²⁴ of the firms sued before and after the Act reveals interesting patterns that may, over time, shed light on the extent to which plaintiff class action securities fraud litigation is, or is not, "merit-driven."

Market Capitalization

Data on the capitalization of companies named as defendants in federal class action securities fraud litigation suggest that issuers sued since the Reform Act became effective are, on average, smaller than issuers that were sued prior to the Reform Act. The reduction in the size of the average defendant issuer appears to be attributable entirely to a dramatic decline in litigation against the largest issuers, i.e., those with a market capitalization in excess of $5 billion.

Table 5 presents data comparing the size of issuers sued before and after the effective date of the Reform Act. The mean market capitalization of issuers sued in cases raising Section 10(b) claims but no Section 11 claims prior to the Reform Act was approximately $2,080 million, with a median of $180.0 million. Since the effective date of the Reform Act, the mean capitalization of issuers sued in comparable Section 10(b) cases has dropped to $529.3 million while the median has increased to $193.0 million.

As is apparent from Table 6, this sharp decline in mean capitalization accompanied by relative stability in median capitalization is almost entirely attributable to the absence in the sample studied of any litigation against issuers with a capitalization in excess of $5 billion.²⁷ Preliminary analysis thus suggests that the Reform Act has cut off the tail of the size distribution of defendant firms by sharply reducing claims against the largest firms.

Because the average IPO has a smaller capitalization than the average firm already traded on the market, we expect that the pre- and post-Reform Act differentials in size of defendant firms would diminish if capitalizations are calculated for pooled Section 10(b) and Section 11 claims. Table 5 indicates just such an effect, but demonstrates that post-Reform Act defendant firms are, on average, still significantly smaller than defendants sued prior to the Reform Act's effective date.

As explained in detail below, this new pattern in defendant selection is consistent with our observation that the preponderance of post-Reform Act litigation involves allegations of accounting irregularities and trading by insiders.²⁹ Larger, more established firms are less likely sources for material accounting irregularities or statistically significant trading by insiders. Larger firms are therefore less likely to be named as defendants. In addition, the stock market has experienced a substantial increase in value since the effective date of the Reform Act, with much of the market's strength centered on the most well-capitalized issuers. Between December 29, 1995, and December 31, 1996, the total returns for the S&P 500 were 22.92%. In contrast, the returns for the Russell 2000 Index were 16.52% over the same period.³⁰ That price pattern is also consistent with a shift toward litigation targeting smaller issuers.

Recent speculation that plaintiffs' class action law firms are avoiding litigation against the largest issuers because of concerns regarding the vigor with which they will defend such litigation is therefore not necessarily correct. Other factors may be sufficient to explain this pattern in the data.

Stock Price Declines Associated With Litigation

Among the more controversial claims made in the debate leading up to adoption of the Reform Act was that a significant stock price decline was frequently sufficient to trigger a class action securities fraud claim.³¹ Recent research suggests that this assertion is overstated.³² A more accurate characterization is that a significant stock price decline over a short period of time can be a necessary but not sufficient condition leading to class action securities fraud litigation.

Analysis of data describing litigation instituted since the Reform Act became effective suggests that, whatever the causal mechanism linking stock price declines and the institution of class action securities fraud litigation, the decline triggering post-Reform Act litigation is greater than the decline triggering pre-Reform Act litigation. As illustrated in Table 7, the average one-day stock price decline around the end of the class period in a sample of 161 pre-Act cases alleging violations of Section 10(b) but not of Section 11, was approximately 19%. A sample of forty-six equivalent post-Reform Act lawsuits indicates an average one-day decline around the end of the class period of about 31%.

A similar increase in stock price declines is found if Section 11 and Section 10(b) cases are pooled in a single sample. A study of 200 class action lawsuits instituted after 10% price declines finds that the average decline surrounding the end of the class period was 24.8%.³⁵ The equivalent statistic for a sample of fifty-four post-Reform Act cases shows a decline of 33.5%. A sample of fifty-nine post-Reform Act cases, which includes five cases brought after one-day declines of less than 10%, shows a one-day average decline of 30.7%.

This increase in one-day stock price declines observed around the end of the class period is consistent with the theory that plaintiffs must, on average, demonstrate more dramatic wrongdoing in the post-Reform Act environment in order to satisfy the new federal pleading standard. Further statistical analysis is necessary, however, to support this conjecture. In particular, it would be valuable to know whether the average stock price decline associated with state court filings that have no parallel federal claims are systematically smaller than those associated with federal claims.

Beta of Defendant Issuers

Beta is a measure of the riskiness of a given firm's stock based on correlation between the movement of its stock price and the movement of a broad-based stock market index. A beta in excess of one indicates greater than average risk in that the stock price tends to move more dramatically than the index as a whole, while a beta of less than one indicates below-average risk in that the stock price tends to move less dramatically than the index as a whole.

Preliminary analysis of a sample of forty-six companies suggests that companies sued after the Reform Act are riskier than the market, with an average beta of 1.39, but perhaps no riskier than the targets of pre-Reform Act litigation, with average betas variously reported as 1.190,³⁶ 1.25,³⁷ 1.408.³⁸ Thus, while targets of class action securities fraud litigation continue to be riskier, on average, than the market index as a whole, the data do not yet indicate a systematic change in the average beta of target companies.

The Incidence of Litigation by Industry

High-technology companies were among the most vocal proponents of securities litigation reform,³⁹ in large part because experience prior to the Reform Act indicated that high-technology companies were involved in a disproportionately large number of securities fraud class action cases. One study of securities fraud class action litigation from 1989 through 1992 concluded that high-technology companies were sued twice as often as firms in other industries.⁴⁰ A study of 348 settlements of open-market fraud securities class actions found that 30.5% were high-technology companies, 22.4% were in financial services, and 47.1% were in some other industry.⁴¹ Thus, an important question is whether the overall rate of litigation against high-technology or other industries has changed in the post-Reform Act period.

To determine the post-Reform Act incidence of litigation activity by industry, we use the Standard Industrial Classification ("SIC") codes that were employed in the Jones and Weingram (1996) study. We rely on the primary SIC codes listed on the issuer/defendant's Form 10-K as it appears in the LEXIS database or on company profiles contained in the LEXIS Company library. Where no SIC information was available on-line for a particular defendant, we rely on SIC information compiled by the Center for Research in Security Prices ("CRSP") at the University of Chicago.

Table 8 indicates that of the 109 issuers sued in federal court from December 22, 1995, through December 31, 1996, approximately 34% are high-technology companies, an estimate within the range found in pre-Reform Act samples. The Reform Act thus seems to have had little effect on the percentage of "high-tech" firms named in securities fraud class action lawsuits.

Actions involving finance companies constitute 10.1% of the complaints. In contrast to the rate of litigation against high-technology companies, the rate of litigation against finance companies appears to have dropped significantly. It does not appear, however, that this decrease is attributable to passage of the Reform Act. This result is, instead, consistent with the results of earlier studies which observed a decreasing rate of litigation against commercial banks⁴³ attributable primarily to the end of the savings and loan crisis and the associated reduction in loan loss reserve litigation.⁴⁴

Another measure of the impact of the Act on securities litigation is whether there have been material changes in the frequency of particular allegations of fraudulent conduct. Our analysis of the nature of the frauds alleged is based on our review of sixty-five, post-Reform Act, federal court complaints that we were able to obtain.⁴⁵ A comparison of these post-Reform Act complaints and available baseline data from pre-Reform Act cases suggests that there has been a significant change in the frequency of particular allegations, apparently in response to the Reform Act's new heightened pleading standard.⁴⁶

As illustrated in Table 9, one of the most common forms of fraud alleged in the sixty-five complaints analyzed are misrepresentations or omissions in financial statements which appear in 58.5% of the complaints. Thirty complaints (or 79% of those alleging false and misleading financial statements) allege a violation of Generally Accepted Accounting Principals ("GAAP"). Of the thirty-eight complaints alleging misrepresentations or omissions in financial statements, thirty-six (or 95%) allege improperly recorded sales, revenues, or earnings. Allegations of misstated financials account for 67.4% of the forty-six complaints that are based solely on alleged Section 10(b) violations.

The frequency of these accounting-driven allegations appears to have increased markedly in the post-Reform Act period. A study of 174 pre-Reform Act Section 10(b) cases found that only 33.9% of the sample contained allegations of misstated financials.⁴⁷ This change in allegations does not mean that there is more accounting fraud in the post-Reform Act period. The significant increase in the number of cases involving misrepresentations and omissions in financial statements, particularly those that allege improperly recorded sales, earnings, or revenues or a GAAP violation, is consistent with the Reform Act's higher pleading standard which requires plaintiffs to plead facts giving rise to a strong inference of scienter. Plaintiffs may believe that courts will be more likely to find that they have satisfied their pleading obligations in cases involving such misrepresentations or omissions.

One unexpected result found in the complaints analyzed, however, is the significant number that also contain allegations involving false or misleading forward-looking statements. The Reform Act created a safe harbor for the release of forward-looking information in certain circumstances. Indeed, one of the chief concerns expressed in the Act's legislative history is that companies that released forward-looking information in the pre-Reform Act period were susceptible to securities fraud class actions if those forecasts proved to be inaccurate, regardless of any proof of the company making an intentionally misleading forecast.⁴⁸ Nonetheless, as detailed in Table 9, forty-two of the sixty-five post-Reform Act complaints analyzed (or 64.6%) contain allegations of false forward-looking statements. Among the forty-two complaints alleging false forward-looking statements, thirty-two (or 76%) allege that the misleading forward-looking statement concerned earnings, sales, or revenue forecasts.

One possible reason that allegations of false forecasts remain a significant part of the litigation landscape is that many of the forecasts alleged to be false occurred prior to the passage of the Reform Act, making it unlikely that any of the companies complied with the safe harbor requirements. It is also important to recognize that allegations of false forecasts were rarely the sole allegation contained in the complaint. Indeed, only nine complaints (or 13.8%) had false forward-looking information as the sole basis for their allegations of securities fraud. In the sample of forty-six Section 10(b) cases that figure drops to three (or 6.5%). In the remaining cases, allegations of false forward-looking statements, were combined with some other allegation of fraud. Thus, where other allegations may be sufficient to satisfy the new pleading standards, plaintiffs may consider that there is little downside in alleging false forecasts as well.

It may also be significant that the nine complaints alleging false forward-looking statements as the sole basis for liability all arose after June 1, 1996. These complaints were thus all filed after the Central District of California's May 21 decision in Marksman Partners, L.P. v. Chantal Pharmaceutical Co.⁴⁹ As discussed more fully below,⁵⁰ that decision is significant because the court's interpretation of the new pleading standard was consistent with pre-Reform Act precedent from the Second Circuit. The court rejected language in the Act's legislative history suggesting that the pleading standard was even more stringent than the Second Circuit's interpretation. Some plaintiffs' attorneys may have taken this ruling to suggest that some forward-looking statement cases might be able to survive even under the new pleading standard.

In addition to an observed increase in the frequency of allegations concerning misrepresentations and omissions in financial statements, passage of the Act also appears to be correlated with an increase in allegations of insider trading. As illustrated in Table 9, 52.3% of the complaints studied in this report contain allegations of trading by insiders during the period the alleged fraud was alive in the market. For the forty-six complaints brought solely under Section 10(b), the percentage rises to 56.5%. The only pre-Reform Act study of which we are aware that analyzes insider trading patterns suggests that insider trading allegations appeared in only 20.7% of a sample of 174 Section 10(b) pre-Reform Act cases.⁵¹ This significant increase in alleged insider trading is consistent with the theory that plaintiffs are increasingly relying on trading by insiders to support the "strong inference" pleading requirement of the Reform Act.

There also appears to be a significant correlation between the frequency of alleged trading during the class period and the industry of the issuer.⁵² Table 10 shows that 73.1% of the complaints involving high-technology companies contain allegations of insider sales, compared to approximately 38.5% for all other issuers. Given that stock option compensation in the high-technology sector is more common than in other sectors of the economy, this finding is not surprising. At the same time, however, this compensation practice also means that the baseline level of "normal" insider sales in the high-technology sector is greater than in other sectors. It would therefore be incorrect to draw any inference from these data that opportunistic behavior by insiders is more common in the high-technology sector than in other sectors.

The observation that trading by insiders occurs even when there is no allegation of fraud raises an important and as yet unresolved question with regard to the interpretation of the "strong inference" pleading standard. If trading by insiders during a period when a fraud is allegedly alive in the market is consistent with patterns observed in situations not involving fraud, how is a court to decide whether such conduct will support a strong inference of a state of mind connoting scienter? The resolution of this complex but exceedingly important issue--especially for the high-technology sector--will likely require further statistical research and the evolution of additional legal doctrine refining the notion of strong inference of fraud in the context of "plain vanilla" sales by corporate insiders.

One possible consequence of the Reform Act would be to shift litigation toward claims that do not come under the new heightened pleading standard.⁵³ In particular, claims brought under Section 10(b) must adequately allege scienter, while Section 11 or Section 12(2) claims alleging fraud in initial or follow-on offerings are not subject to this requirement. The heightened pleading requirement thus imposes a differentially greater burden for pleading Section 10(b) claims than for the pleading of Section 11 or Section 12(2) claims. One predictable consequence of the differential pleading standard would be a shift toward Section 11 and Section 12(2) cases.

Prior to the Reform Act, approximately 21% to 24% of class action securities fraud cases involved claims arising from the sale of securities in initial public offerings or follow-on offerings.⁵⁴ Of these underwritten offerings, IPOs accounted for approximately 14%⁵⁵ and follow-on offerings 8% to 10%.

To date, we have not observed any increase in the percentage of litigation raising claims of fraud in the sale of securities governed by Section 11 or Section 12(2). In a sample of sixty-five post-Reform Act complaints, fifteen (or 23%) alleged violations of Section 11 or Section 12(2). Allegations involving IPOs are significantly more frequent, accounting for eleven of these claims. Subsequent versions of this report will analyze all 109 companies sued in federal court in 1996 and state court complaints that may assert Securities Act claims.

It was generally understood that prior to passage of the Reform Act, a single law firm, Milberg Weiss Bershad Hynes & Lerach ("Milberg Weiss"), played a dominant role as plaintiffs' class action counsel. Available data suggest that during the period from April 1988 through September 1996, Milberg Weiss represented clients in approximately 31.4% of 842 class action securities fraud cases pending nationwide.⁵⁶

Since passage of the Reform Act, Milberg Weiss appears to have become even more dominant in the class action securities fraud litigation process. As illustrated in Table 11, of the total of 109 companies sued in federal court, Milberg Weiss has entered an appearance in at least 51 of these federal cases, for a national federal appearance ratio of 46.8%. If the national sample is expanded to include state claims with parallel federal proceedings, we are able to identify Milberg Weiss appearances in 64 of 109 cases, for a 58.7% appearance ratio.

These national appearance ratios understate Milberg Weiss' particularly active role in litigation pending in California federal and state courts. In the 24 cases identified as filed in a California federal court, at least 17, or 70.1%, involve appearances by Milberg Weiss. If the California sample is expanded to include California state court claims with parallel federal proceedings, we are able to identify Milberg Weiss appearances in 20 of 24 cases, for a 83.3% percent appearance ratio.

These data must be interpreted with caution, however. In addition to the previous comments regarding the preliminary nature of this research, it is important to emphasize that an appearance ratio should not be confused with a market share for several reasons. First, it is common practice for many law firms to enter appearances on behalf of different plaintiffs in the same case. If the average consolidated class action brings together claims filed by different law firms, then the aggregate appearance ratio for the market as a whole will be greater than 100%. Any individual law firm's appearance ratio, calculated as a percentage of the market's aggregate appearance ratio, will therefore be lower than its appearance ratio calculated on a stand-alone basis. At present, we lack a sufficiently large sample of complete dockets to develop a reliable estimate of any firm's appearance ratio as a fraction of average aggregate appearance ratios.

Second, not all appearances are equal. Prior to the Reform Act, a lead counsel was often designated in class action securities fraud litigation, and that lead counsel typically asserted disproportionate control over the litigation and collected a disproportionate share of the fees awarded, if any, to plaintiffs' counsel in the litigation. It is generally believed that Milberg Weiss, which is the largest of the nation's law firms specializing in class action securities fraud litigation, is often designated lead counsel and often plays a dominant role in the prosecution of these claims. Accordingly, an appearance by Milberg Weiss in any given lawsuit may, on average, indicate a greater influence by that firm than any other making an appearance in the litigation. Again, however, we lack data necessary to adjust for this observation.

Third, the Reform Act established a procedure whereby a lead plaintiff has the obligation to select class counsel subject to approval by the court.⁵⁷ To the extent that the addition of this statutorily mandated approval process either increases or decreases the probability of Milberg Weiss being named lead counsel, or changes the dynamics of inter-firm management of plaintiff litigation, appearance ratios observed prior to the Reform Act may have a significance that is different from the appearance ratios observed since the Reform Act's effective date. In particular, to the extent that economies of scale for larger firms, such as Milberg Weiss, make it easier for them to attract coalitions of plaintiffs willing to act as lead plaintiff and to designate Milberg Weiss as counsel, this provision of the Act may increase the firm's significance in the litigation process. On the other hand, to the extent that large institutional investors step forward to influence the counsel selection process away from Milberg Weiss, this provision may diminish Milberg Weiss' future influence. Again, we must await further experience before drawing firm conclusions regarding these effects of the Reform Act.

No doubt, data describing the percentage of total plaintiff class action attorney fee awards captured by each class action counsel would play a useful role in estimating a true "market share" as opposed to an appearance ratio statistic. The data necessary for such calculations are not, however, publicly available. While courts are required to approve the aggregate fee award paid to counsel in class action litigation, the court is typically not called upon to approve the division of that fee among the many law firms potentially representing plaintiffs, and the public record typically contains no information regarding these allocations.

Notwithstanding these important cautionary statements, Milberg Weiss' increased dominance as measured by the appearance ratios observed to date is consistent with economic theory regarding firm behavior in a market subject to externalities. Specifically, if Milberg Weiss is the largest and best capitalized of the plaintiffs' class action firms, which it appears to be, then it will be best situated to bear the additional cost of delay and uncertainty associated with litigation in the post-Reform Act era. Further, to the extent that in the earliest years following adoption of the Reform Act any one decision reached by any court interpreting a provision of the Reform Act may have a disproportionately large effect on the resolution of other cases, the Milberg Weiss firm has an incentive to internalize this externality by assuring that early cases do not establish precedents adverse to the interests of the firm. No other firm will have an equally strong incentive to invest for this reason. Put another way, in the early years following adoption of the Reform Act, there will be relatively less well established doctrine on which smaller firms, less capable of financing or undertaking riskier litigation, will be able to "free-ride." Thus, just as the Reform Act may create economic incentives for Milberg Weiss to expand its appearance ratio, it may also establish an incentive for smaller firms to shrink their appearance ratios.

If the preceding analysis is correct, and if historical patterns are not disturbed by lead plaintiffs' decision to exercise a new degree of control over the counsel designation process, then we would not be surprised to find a material period of increased dominance by Milberg Weiss in the market for class action securities fraud litigation.

Table 12 presents data describing the incidence of litigation by judicial circuit from December 22, 1995 through December 31, 1996. As that table demonstrates, the geographic distribution of post-Reform Act litigation by circuit is similar to the distribution of litigation prior to the Reform Act.⁵⁸ This result is not surprising given the relative stability of high-technology firms among the companies sued. As in the pre-Reform Act period, the Ninth Circuit (which includes California) is the most active forum for securities class actions; however the percentage of cases in that circuit has dropped from 36% to 27.6%. The Second Circuit (which includes New York) remains the next most active circuit, although there has been a modest increase in cases from 15% to 18.1%. Among circuits with significant numbers of filings, modest increases have also been observed in the Fifth (which includes Texas) and Eleventh (which includes Florida) Circuits.

Table 13 lists the top five districts for securities class action filings and describes the number of securities class action filings per thousand civil case filings in those districts. The districts in which the most companies have been sued are the Northern District of California (Silicon Valley's home district) and the Southern District of New York (which includes Manhattan). Each district had fifteen filings or about 13% of the total. In total, California district courts account for about 21% of securities class action litigation in the post-Reform Act period. New York district courts have the second highest incidence of securities class action litigation during the study period with 17.9% of the post-Reform Act litigation activity.

Securities class actions are, however, a significantly larger portion of the docket in Northern California, which has nearly twice as many securities class action filings per thousand civil cases (2.872) as the Southern District of New York (1.460).⁶¹ While these figures suggest that securities class actions account for only a small percentage of total civil filings, they do not capture the magnitude of the burdens these cases place on the judiciary. The average class action demands considerably more judicial time than either the average civil case or the average non-class action securities fraud case.⁶²

It has frequently been observed that the large majority of federal class action securities litigation filed prior to the Reform Act was resolved by settlement. One study has reported that 87.6% of the securities class actions filed from April 1988 through September 1996 ended in a settlement,⁶³ with the large majority of the remainder being resolved by dispositive motions or voluntary dismissal. Very few class action securities fraud cases go to trial. Another study has found that the median time between filing and settlement was 21.7 months.⁶⁴ Any trends in the number of cases settling, the length of time between filing and settlement, and the average settlement amounts will provide important data as to the effects of the Reform Act.

The first year of practice under the Reform Act, however, is too short a period within which to expect substantial settlement activity, particularly given the novelty of many of the Reform Act's provisions. We are aware of only five settlements in 1996⁶⁵ of cases brought under the Reform Act, and we expect that it may take several years of experience before we are able to draw broad conclusions as to the effect of the Reform Act on settlement behavior. Nonetheless, early experience supports the tentative inference that plaintiffs are quickly dismissing certain claims with little or no recovery. Such conduct is consistent with a plaintiff filing a lawsuit only to discover quickly that the claim lacks merit or is otherwise not profitable to pursue. It therefore remains an open question as to whether the Reform Act has successfully deterred the filing of claims that can quickly be determined to be weak.

The settling cases are as follows:

1. Caramonta v. Dingus.⁶⁶ On May 30, 1996, Touchstone Software, a developer and publisher of utility software, announced a settlement of three shareholder class action and derivative suits brought against it and several of its officers and directors. The agreement calls for the establishment of a settlement fund consisting of $500,000 to be paid by Touchstone and issuance of 200,000 new shares of the company's common stock. The agreement also calls for plaintiff review of certain policies.⁶⁷

2. Levy v. United HealthCare Corporation.⁶⁸ On December 18, 1996, United HealthCare announced that all claims against it in the Levy action had been voluntarily dismissed with prejudice. Pursuant to a Stipulation and Order, the parties agreed to dismiss all claims prior to class certification. No payment was made to the named plaintiff or his counsel.

3. Alexander v. Health Management.⁶⁹ Health Management, Inc., announced a settlement of all pending shareholder class actions filed against it. The settlement, subject to final court approval, calls for a $2 million cash payment and the issuance of 2.2 million shares of common stock and 2.2 million warrants to purchase common stock.⁷⁰

4. Trieff v. Cirrus Logic Inc.⁷¹ Cirrus Logic agreed to pay $31.3 million to settle shareholder lawsuits filed against the company in 1993, 1995, and 1996. The post-Reform Act complaint was filed in California state court and alleged claims arising out of Cirrus Logic's announcement that it would restate quarterly results. On August 22, 1996, the Alameda Superior Court dismissed four of plaintiffs' claims for relief. Plaintiffs then filed a second class action complaint against the company in September 1996 alleging the company misstated demand for certain of its products. The company will pay $2.3 million of the settlement.

5. Fradin v. HighwayMaster Communications, Inc.⁷² Plaintiffs agreed to dismiss voluntarily their class action lawsuit without costs to either side. The complaint alleged that HighwayMaster omitted to disclose in its IPO prospectus certain product defects in mobile communications equipment the company manufactured.

As of the date of this paper, we are aware of twenty-six decisions interpreting provisions of the Reform Act. Because these decisions arise in cases that have all been filed within the last year, they involve exclusively the earliest judicial determinations in class action securities litigation under the Reform Act. These twenty-six decisions involve the: (a) interpretation of the "strong inference of fraud" pleading standard; (b) scope of a discovery stay on a pending motion to dismiss; (c) conflicts between federal and state court; (d) adequacy of notice and the appointment of a lead plaintiff; (e) survival of the recklessness standard for scienter; (f) safe harbor for forward-looking information; and (g) retroactive application of the Reform Act. Two additional decisions, although not decided under the Reform Act, may provide important precedents for its interpretation. This section discusses the decisions chronologically within each of these categories.

To the extent that it is possible to generalize from this early experience, three developments appear to be worth mention. First, of the seven courts interpreting the strong inference pleading requirement, six have concluded that the statute incorporates the Second Circuit's pleading requirement and not some higher standard implied in the legislative history. This is the conclusion preferred by plaintiffs. Second, only one motion to dismiss has been granted without leave to replead the major portion of the case. The sole exception involved a Section 11 claim that was not subject to the new strong inference standard. The new heightened pleading standard therefore does not yet appear to be functioning as a useful device for quickly dismissing claims drafted with an eye toward satisfying the statute's new requirements.

Third, complex issues regarding the interplay of federal and state jurisdiction are rapidly emerging. The application of the discovery stay to state litigation while a federal claim is pending and the interpretation of the federal safe harbor in a state proceeding when there is no state law safe harbor are but two examples of the many issues now arising in class action securities fraud litigation. This tension between the federal and state courts will likely grow over time and may well draw attention from Capitol Hill.

Interpretation of the "Strong Inference of Fraud" Pleading Standard

One of the primary reforms Congress put in place to curtail the filing of "meritless" lawsuits was a new heightened pleading standard. Section 21D(b)(2) of the Reform Act requires, among other things, that plaintiffs in securities fraud actions "state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind." The "strong inference" language was taken from Second Circuit case law interpreting the requirements of Federal Rules of Civil Procedure 8 and 9(b) in the context of securities fraud cases.⁷³ President Clinton cited this provision in his veto of the Reform Act because, in his view, language in the legislative history indicated that Congress was adopting a standard that was more stringent than that of the Second Circuit. Since the veto and subsequent override, commentators and courts have disagreed over what facts will satisfy Section 21D(b)(2) and whether the Second Circuit tests for finding a strong inference of fraud survived passage of the Act. Through January 1997, at least seven courts addressed the new pleading standard, with six finding that the Second Circuit tests survived passage of the Reform Act. In addition, it is significant to note that no motion to dismiss in a case subject to the new pleading standard has been granted without permitting plaintiffs the opportunity to replead at least some of their claims.

1. Marksman Partners, L.P. v. Chantal Pharmaceutical Corp.⁷⁴ Chantal is the first decision to interpret the Reform Act's more stringent pleading requirements. The Central District of California denied a motion to dismiss and found that plaintiffs had satisfied their pleading obligations. In so holding, the district court determined that Congress did not adopt a more stringent pleading standard than had existed in the Second Circuit. The court also found that two tests the Second Circuit employed to determine whether the plaintiff satisfied its pleading obligations--the "motive and opportunity" test and the "strong circumstantial evidence" test--survived passage of the Reform Act.⁷⁵ The court held that allegations that the Chairman and CEO sold 20% of her stock during the class period for proceeds of $6.3 million were sufficient to allege a strong inference of fraud, particularly where she had not sold any stock during the previous three years.

2. Zeid v. Kimberley.⁷⁶ The court dismissed with leave to replead⁷⁷ a securities class action brought against Firefox Communications, Inc. and certain of its officers and directors. The complaint alleged that defendants had misrepresented the demand for the company's products and the success of its sales and marketing program in the United States. The complaint also alleged that Firefox improperly recognized certain revenues and failed to keep adequate reserves in violation of GAAP and SEC rules.

The court found that plaintiffs failed to plead with sufficient specificity: (i) when or how allegedly misleading statements were communicated to the market; (ii) that the company adopted certain analysts' statements; and (iii) the reasons why certain statements were misleading. Allegations that Firefox violated GAAP by "parking" inventory with distributors were insufficient because the complaint did not "name a single customer or sale where Firefox 'parked' its inventory or prematurely recognized revenue."⁷⁸ The court did find that the complaint alleged with particularity facts surrounding the company's failure to maintain adequate reserves. However, these claims were dismissed as well for failure to plead a strong inference that defendants' actions were intentional or reckless. As in Chantal, the court in Zeid found that the Second Circuit's tests for pleading scienter survived passage of the Reform Act.

Finally, the court dismissed with prejudice plaintiffs' claims that certain warnings and disclaimers contained in the company's Form 10-Qs were false and misleading. Plaintiffs argued that these statements were merely boilerplate and that the company's reports should have contained specific disclosures of the adverse factors which were then negatively affecting the company's business. The court rejected this argument as "absurd," and held that the warnings were not actionable as a matter of law because plaintiffs did not allege that the warnings were wrong. Instead, they alleged that the warnings "should have been more specific." This was not a proper claim under Section 10(b) and Rule 10b-5, which do not protect against "statements that are too abstract."⁷⁹

An amended complaint has been filed and is now pending before the court.

3. Sloane Overseas Fund, Ltd. v. Sapiens International Corp.⁸⁰ The Reform Act's provisions did not apply to this case because the action was filed prior to the effective date of the Act. In deciding a motion to dismiss, however, the court noted in dicta that the Reform Act "codif[ied the] Second Circuit standard for pleading scienter."⁸¹

4. In re Silicon Graphics, Inc. Securities Litigation.⁸² The court in Silicon Graphics reached a substantially different conclusion with respect to plaintiff's pleading obligations than the courts in Chantal and Zeid. Unlike those cases, the Silicon Graphics court held that the Reform Act did not simply codify the prior Second Circuit standard.⁸³ Instead, the court found that Congress meant to erect a higher pleading barrier that requires "that plaintiff must allege specific facts that constitute circumstantial evidence of conscious behavior by defendants."⁸⁴

The court found that plaintiff's attempt "to couple allegations of defendants' awareness of negative internal reports with their false and misleading statements and stock sales ... [was] not specific enough to raise a strong inference of fraud."⁸⁵

The court granted plaintiff leave to replead, and an amended complaint was filed on October 17, 1996. A second motion to dismiss the amended complaint is now pending.

5. STI Classic Funds v. Bollinger Industries, Inc.⁸⁸ In Bollinger, a magistrate judge ruling on a motion to dismiss held that the motive and opportunity test survived passage of the Reform Act. The magistrate judge then found that the facts pleaded in the amended complaint at issue satisfied this standard with respect to claims against the company and certain of its officers, but not with respect to a subsidiary company.

Defendants Glenn and Bobby Bollinger founded and were the two senior officers of Bollinger. The complaint alleged that the company invoiced fraudulent sales transactions with one or more customers in order to inflate reported sales and earnings. These misstatements of financial performance allegedly stemmed from inadequate financial and accounting controls and violations of GAAP of which the company and its senior management were aware. Bollinger was also allegedly aware prior to an announcement in March 1995 that supervisors at its NBF subsidiary were falsifying production reports. In June 1995, two of the company's outside directors and its auditor resigned, allegedly because of these problems.

With respect to motive, the court noted that allegations of materially inflated financial health "benefited the value of Bollinger's shares and likewise increased the value of the Brothers Bollinger's interest in the company."⁸⁹ The court rejected defendants' argument that a similar finding of motive would apply to any small, family-dominated business. A strong inference could be drawn from the individual defendants' positions within the company "that they were knowledgeable about the methods and billing practices utilized by Bollinger which led to the over-stated sales and revenues reported in the SEC filings signed by them."⁹⁰ The court found the allegations against the individual defendants with respect to NBF inadequate to satisfy the Reform Act's pleading standard because they failed to plead sufficiently that the defendants knew that the statements were false when made.

The district court adopted the magistrate judge's findings on November 12, 1996.

6. Fischler v AmSouth Bancorporation.⁹¹ Plaintiff brought a class action on behalf of all purchasers of non-deposit investment products from AmSouth, claiming that AmSouth failed to disclose certain surrender charges applicable to these instruments. In a brief opinion, the court determined that the complaint satisfied the requirements of Federal Rule of Civil Procedure 9(b) and the Reform Act. The court noted that Rule 9(b) had three purposes: (i) ensuring that the allegations of fraud are specific enough so that defendants will be able to respond effectively; (ii) eliminating those complaints filed as a pretext for discovery of unknown wrongs; and (iii) protecting defendants from unfounded charges of wrongdoing. The court then applied the Second Circuit tests for pleading a strong inference of fraud without discussing whether those tests survived passage of the Act. The also court noted that the "motive and opportunity" test was a "common method for establishing a strong inference of scienter".⁹²

7. Rehm v. Eagle Finance Corp.⁹³ The court in Eagle Finance found that the Reform Act "does not impose a more rigorous pleading requirement than that enunciated by the Second Circuit."⁹⁴ The Reform Act, however, also "declines to bind courts to the Second Circuit's interpretation of its standard."⁹⁵ Nonetheless, the court employed the motive and opportunity and strong circumstantial evidence of fraud tests in refusing to dismiss the action, which alleged that Eagle materially understated its credit losses and overstated its earnings.

Following Second Circuit precedent, the court found that certain generic motives that were likely held by executives generally were not enough to create a strong inference of fraud. Similarly, alleged insider sales did not satisfy the motive and opportunity test. The court first noted that two of three individual defendants did not sell any stock during the class period. The third defendant sold only 6% of his personal holdings. While this was the defendant's first sale, the court did not find this unusual because the stock was originally issued only sixteen months earlier.

The court found that plaintiff's combination of allegations constituted strong circumstantial evidence of conscious misbehavior or recklessness. First, Eagle allegedly overstated its 1995 earnings by 91%. This allegedly serious GAAP violation, when combined with allegations that defendants were responsible for calculating and releasing the financial information, supported the conclusion that defendants acted with scienter. Second, the company reported "massive" year-end increases to credit loss reserves and decreases to earnings. The magnitude of these reporting errors lent weight to the allegations of recklessness, especially where defendants were in a position to detect the errors. Third, the court pointed to the nature of Eagle's loan servicing business and defendants' statements downplaying the significance of the accounting errors to support a finding of scienter. "[T]he crucial significance of accurate credit loss accounting in determining the financial viability of Eagle, combined with defendants' careful statements mitigating the seriousness of the credit loss problem, raises a strong inference that defendants acted with knowledge of their public misstatements or were willfully blind to the truth."⁹⁶

Other Motions to Dismiss

1. Steckman v. Hart Brewing, Inc.⁹⁷ Hart Brewing is the only post-Reform Act case to date where a complaint was dismissed without leave to replead. Ironically, the Act's new heightened pleading standard did not apply to the case because it alleged only violations of Sections 11 and 12(2) of the 1933 Act in connection with Hart's initial public offering. Nonetheless, the court noted that the Reform Act "encouraged the use of motions to dismiss in certain securities cases."⁹⁸ The court also echoed one of the Reform Act's rationales when it noted that courts should dismiss 1933 Act claims to "minimize the chance that a plaintiff with a largely groundless claim will bring a suit and conduct extensive discovery in the hopes of obtaining an increased settlement."⁹⁹

The Hart complaint alleged that defendants failed to disclose material facts indicating an adverse trend of declining sales. Plaintiff alleged that disclosure of partial fourth quarter 1995 results would have shown that Hart could not sustain its past high growth and that this disclosure would have reduced the IPO stock price.

The court's decision hinged on the interpretation of Item 303 of Regulation S-K, which requires companies going public to disclose trends that are presently known to management and reasonably likely to have a material effect on financial condition or operating results. The court found that Item 303 only required a company to disclose intra-quarter results where those results represent an extreme departure from earnings of prior quarters. The complaint was found to allege no facts indicating either that defendants actually knew that the quarterly results would be an extreme departure or that the results were in fact an extreme departure. The court also noted the company had no duty to disclose trends about future performance or to supply other forward-looking information.

The Scope of the Discovery Stay on a Motion to Dismiss

In passing the Reform Act, Congress was concerned that "[t]he cost of discovery often forces innocent parties to settle frivolous securities class actions."¹⁰⁰ To combat this problem, the Reform Act provides in new Section 27(b) of the 1933 Act and new Section 21D(b)(3) of the 1934 Act that "all discovery and other proceedings shall be stayed during the pendency of any motion to dismiss, unless the court finds upon the motion of any party that particularized discovery is necessary to preserve evidence or to prevent undue prejudice to that party."

At least three federal courts addressed the scope of this discovery stay in 1996. These decisions suggest that courts will interpret broadly the Reform Act's mandatory stay provision.

1. Medical Imaging Centers of America, Inc. v. Lichtenstein.¹⁰¹ Plaintiff, the subject of a corporate control contest, filed an action alleging violations of Section 13(d) of the Exchange Act. Defendants called a special shareholders meeting and filed a proxy indicating their intent to unseat plaintiff's current board of directors. Defendants also moved to dismiss the complaint and requested expedited discovery to be completed before the shareholders' meeting. Plaintiff moved to stay discovery based on Section 21D(b)(3)(B).

The court held that discovery was properly stayed while the motion to dismiss was resolved. The "undue prejudice" standard was held to require the party seeking discovery to show an improper or unfair detriment. The court characterized this showing as something less than irreparable harm. The court noted that if the party seeking discovery "had shown that the discovery stay would prejudice it because [the opposing party] would be shielded from eventual liability for any material violations of the securities laws, the Court would find that an 'undue prejudice' exception to the statutory stay had been shown."¹⁰² The court also rejected an argument that Congress did not mean for the statutory stay to apply to proxy contests. Although no absolute "carve out" exists for such actions, the court noted that it would be proper for the court to consider the nature of the action in determining whether an exception to the stay should be recognized.¹⁰³

2. Novak v. Kasaks.¹⁰⁴ The court stayed discovery in a class action involving AnnTaylor Stores because it found that plaintiffs had not satisfied their burden of showing "exceptional circumstances." In response to plaintiffs' concern that non-parties might not retain relevant documents during a stay, the court ordered the 30 non-parties that had previously been served with subpoenas to preserve all responsive documents.

3. Medhekar v. United States District Court.¹⁰⁵ In Medhekar, the district court held that the statutorily required stay did not apply to mandatory disclosure obligations under local civil court rules and the Federal Rules of Civil Procedure.¹⁰⁶ The court found that disclosure is distinct from discovery and that if Congress had meant to stay both it would have listed both in the statute.¹⁰⁷ The court also found that the statutory phrase "other proceedings" did not clearly encompass disclosure. Instead, it interpreted "other proceedings" to refer to formal procedures involving a hearing or other court activity.¹⁰⁸

Defendants sought and obtained a writ of mandamus directing the lower court to stay the initial disclosure requirements pending disposition of defendants' motion to dismiss.¹⁰⁹ The Ninth Circuit held that mandatory disclosures constitute "discovery" for purposes of the Act's stay provisions. The Ninth Circuit rejected the lower court's conclusion that disclosure and discovery were distinct; instead, it found that disclosure was merely a subset of discovery. The time and expense associated with disclosure were exactly the type of burden the Act sought to eliminate. The court thus found that: "Congress clearly intended that complaints in these securities actions should stand or fall based on the actual knowledge of the plaintiffs rather than information produced by the defendants after the action has been filed."¹¹⁰

The Ninth Circuit also rejected the lower court's interpretation of the phrase "other proceedings". The legislative history and context of the phrase suggested that the term was intended to include litigation activity relating to discovery. Thus, the term includes disclosures. The court was careful to emphasize, however, that the term does not include "all litigation activity in general."¹¹¹

Conflicts Between State and Federal Court

One strategic response to the Reform Act is to avoid it altogether. Some plaintiffs have sought to escape from the Act's discovery stay and other procedural restrictions by filing parallel or stand-alone state court actions alleging either claims under the 1933 Act or state common law or Blue Sky claims. This strategy raises interesting and difficult federalism problems. In 1996, at least six state courts appear to have addressed whether the Reform Act's discovery stay should apply in state court proceedings. So far, state courts have split on whether the Reform Act's discovery stay is a substantive policy decision that should be respected in state courts or merely a procedural mechanism with no applicability in state fora. The significant substitution effect into state court and the emerging strategy of filing parallel federal and state court actions suggest that resolution of this issue may significantly affect the practical effectiveness of the stay provision.

1. Milano v. Auhll.¹¹² Plaintiff brought a class action against Circon Corporation and certain of its officers and directors alleging violations of the Securities Act of 1933 and the California Corporations Code in California state court. No parallel federal action was pending. Defendants moved to stay discovery pursuant to the Reform Act and the plaintiffs argued that the Act was inapplicable in state court proceedings. The court found that the Reform Act's stay provisions were applicable in state court actions and determined that discovery on both the federal and state claims should be stayed. This result was necessary in order not to undermine the congressional intent "to provide a broad and effective method of weeding out frivolous and unsupported lawsuits."¹¹³

2. Sperber v. Bixby.¹¹⁴ In this class action involving Brooktree Corporation, plaintiffs alleged solely state Blue Sky and common law claims. No parallel federal action was filed. Defendants filed both a motion to dismiss and a motion to stay discovery pending resolution of the dismissal motion. Defendants argued that plaintiffs filed in state court to evade the Reform Act's procedures, including the stay of discovery and the heightened pleading standard. In this situation, defendants argued that the Reform Act, although not controlling, provided persuasive authority that should be considered in ruling on stays in state securities class actions. The court stayed discovery pending resolution of the motion to dismiss.

3. Marinaro v. The Superior Court of Santa Clara County.¹¹⁵ In this case involving Network Computing Devices four federal complaints arising out of the same facts and circumstances were filed after the state court complaint. Defendants sought to stay the state case, arguing that the state claims were brought in a separate action in order to avoid the procedural provisions contained in the Reform Act. The lower court denied the stay without opinion and the intermediate appellate court denied a petition for a writ of mandamus. The California Supreme Court, however, directed the appellate court to reconsider defendants' request for a stay.¹¹⁶ As of the date of this paper, we are unaware of any decision from the appellate court.¹¹⁷

Adequacy of Notice and Appointment of Lead Plaintiff

Under the Reform Act, Congress sought to increase litigant oversight of plaintiffs' class action attorneys by reducing the number of "professional plaintiffs" and by encouraging institutional investors to become more active in securities fraud litigation. To achieve this result, the Reform Act requires that plaintiffs filing securities class actions file sworn certifications describing, among other things, their transactions in the security and their prior appearances as named plaintiffs in securities class actions.¹¹⁸ The Act also requires the named plaintiff to publish a notice "in a widely circulated national business-oriented publication or wire service" that informs potential class members of the right to move to be appointed "lead plaintiff."¹¹⁹ The law establishes a rebuttable presumption that the lead plaintiff will be the party who volunteers and who "has the largest financial interest in the relief sought by the class."¹²⁰ Lead plaintiffs can control the course of the class action, including selection of lead counsel, subject to court approval.¹²¹

Through January 1997, at least four opinions addressed these provisions.¹²² The activity to date suggests three significant trends are emerging. First, most of the activity has not involved the institutions Congress sought to get involved in securities litigation. Instead competing groups of individual plaintiffs have tended to vie for the lead plaintiff position. Indeed, one of the unintended consequences of the Reform Act has been to encourage traditional plaintiffs' law firms to cobble together large numbers of smaller claimants who, in aggregate, have the largest financial interest of any individual or group seeking the lead plaintiff position.

Second, firms' initial reaction to the notification procedures was to publish inconspicuous notices in Investors Business Daily. One court found these notices to be inadequate. But that decision may have little practical impact because before it was issued firms had already begun to employ the Act's notification procedures as a means of advertising their actions in order to attract additional plaintiffs or others with relevant information about the asserted claims. The notices therefore now tend to be widely-disseminated on wire services and contain lengthy descriptions of the factual bases for the complaint.

Finally, the decisions to date suggest that if institutions are interested in becoming lead plaintiffs, they should be able to do so in most cases. Significant questions remain concerning the kind of discovery institutions will be subject to if they assume that position, and it is unclear that many major institutional investors will conclude that the benefits of participation as lead counsel will outweigh the costs.¹²³

1. In re Cephalon Securities Litigation.¹²⁴ Under the Reform Act's provisions for the appointment of lead plaintiff, "discovery relating to whether a member or members of the purported plaintiff class is the most adequate plaintiff may be conducted by a plaintiff only if the plaintiff first demonstrates a reasonable basis for a finding that the presumptively most adequate plaintiff is incapable of adequately representing the class."¹²⁵ In Cephalon, the Court permitted discovery to resolve a dispute over whether a plaintiff that was seeking to be named lead plaintiff was in fact an institutional investor.¹²⁶

2. Greebel v. FTP Software.¹²⁷ Greebel is the first substantive decision addressing a number of important issues that arise from the Reform Act's lead plaintiff provisions. First, the court addressed the defendant's role in the lead plaintiff determination. The court held that defendants have standing to contest plaintiff's failure to file a certification with the complaint or the failure to provide adequate notice to the class. Defendants lack standing to challenge whether a particular plaintiff satisfies the requirements set forth in Section 21D(a)(3)(B)(iii), in particular, whether the proposed lead plaintiff "otherwise satisfies the requirements of Rule 23 of the Federal Rules of Civil Procedure." Because defendants have traditionally been permitted to challenge a plaintiff's satisfaction of the Rule 23 requirements at the class certification stage, the court held "that its determination to appoint a person or persons as lead plaintiff must be without prejudice to the possibility of revisiting that issue in considering a motion for class certification."¹²⁸

Second, the court held that only representative plaintiffs that file complaints are required to file the certifications required in Section 21D(a)(2). Finally, the court held that publication of a notice on Business Wire, a computer database service that distributes press releases to news media, on-line services, and subscribers in the investment community, satisfies the Reform Act's notice requirement.

3. Gluck v. Cellstar Corporation.¹²⁹ The State of Wisconsin Investment Board ("SWIB") sought in this case to be named sole lead plaintiff. Class members who had filed a complaint and were represented by Milberg Weiss opposed that motion. Milberg Weiss argued that SWIB was not an appropriate lead plaintiff because SWIB is a sophisticated institutional investor that had used derivatives and is not typical of the entire class of plaintiffs. Milberg Weiss argued that the court should therefore not appoint SWIB as sole lead plaintiff and should instead appoint a plaintiffs' committee that would serve as lead plaintiffs, with Milberg Weiss and SWIB's counsel appointed as co-lead counsel. The court appointed SWIB as sole lead plaintiff and required SWIB, subject to court approval, to appoint counsel for the class within 30 days of the order.

4. Chan v. Orthologic Corp.¹³⁰ In Orthologic, the City of Philadelphia pension fund and a group of non-institutional plaintiffs (the "Chan Plaintiffs") competed to be appointed lead plaintiff. It was undisputed that the City of Philadelphia had the largest financial stake in the outcome of the case and was therefore the presumptive lead plaintiff under the terms of the Reform Act.¹³¹ The Chan Plaintiffs sought to overcome the presumption by arguing that they represented a narrower class of purchasers that purportedly conflicted with the Philadelphia class. As in Cellstar, the Chan Plaintiffs argued that Philadelphia, as a sophisticated, institutional investor, was not typical of the class as a whole and that it was subject to unique defenses. In particular, Philadelphia's investment strategy was said to be atypical of the class as a whole because Philadelphia was a "speculator" that allegedly purchased Orthologic securities after certain bad news had already been disseminated to the market. As a result, the Chan Plaintiffs sought either to be appointed co-lead plaintiffs with Philadelphia or to be named to their own subclass.

The court held that these factors were insufficient to rebut the Act's presumption that Philadelphia was the most adequate plaintiff. First, the court held that the differing class periods did not create a conflict because the parties had similar injuries arising out of the same course of conduct. The court did, however, reserve the right to create appropriate subclasses if a conflict was subsequently revealed. Second, the court held that Philadelphia's sophistication was irrelevant in this case because all plaintiffs had consistent interests. Finally, the court rejected pre-Reform Act precedent that found institutional investors to be atypical of the class. Instead, it relied on Ninth Circuit precedent finding that differences in sophistication among purchasers have no bearing in fraud-on-the-market cases.¹³²

After appointing Philadelphia as lead plaintiff, the court also approved its choice of Milberg Weiss; Barrack Rodos & Bacine; and Bonnett Fairbourn Friedman & Balint as co-lead counsel. The court noted the firms' "extensive experience in the area of class action securities fraud" in finding that they would provide capable representation to the class as a whole.

5. Ravens v. Iftikar.¹³³ In this class action involving Syquest Technology, Inc., Judge Vaughn Walker examined the Reform Act's requirements for the content of the required notice. The notice at issue in Ravens simply stated the date the action had been filed, where it was filed, the alleged class period, the security at issue, and that the complaint asserted violations of Sections 10(b) and 20(a) of the Exchange Act. The notice did not contain a description of the facts underlying the claims. Judge Walker held that the notice did not satisfy the Reform Act's requirements because it was "inadequate to apprise investors of the claims asserted so that those who may wish to do so have a fair opportunity to intervene and assume control of the litigation."¹³⁴

Congress employed the lead plaintiff and notice provisions to curb "the disproportionate influence lawyers have exerted over securities class actions."¹³⁵ The Court found that these provisions were meant to "transfer primary control of private securities litigation from lawyers to investors" and were similar to other constitutionally mandated notices required under Federal Rule of Civil Procedure 23(b)(3). He held that the Reform Act notice must at a minimum contain three elements: (1) a notice of the pendency of the action; (2) a description of the claims asserted therein; and (3) a specification of the purported class period.

The bulk of the court's opinion focused on the second element. The court found that the notice provisions will only permit investors to make an informed decision about whether to seek control of the case if investors "are notified of the nature and character, not just the existence, of the claims asserted."¹³⁶ The notice must contain information describing the legal and factual basis f