Securities Act of 1933

This law states registration and distribution requirements for public issuers of securities. Companies that issue securities to the public are required to disclose accurate information in their filings and other communications to shareholders. The dissemination of information known to be inaccurate can give rise to a cause of actions by investors who were adversely affected through their reliance on this information. In the wake of scandals such as Enron, the accuracy and adequacy of disclosed financial data have become a major issue.² Fiduciaries and professionals, such as accountants, may be held liable for the inaccurate information. However, defendants in such actions may try to assert defenses such as stating that investors were aware of the inaccurate information or that they (the defendants) did not know the information was false. Such suits require that experts analyze the damages arising from cases in which investors lost some or all of their investment as a result of inaccurate or false information.

Securities Exchange Act of 1934

This law regulates the trading of securities in the secondary market following their issuance in the primary market. This law requires companies to regularly file statements containing financial and other relevant information with the Securities and Exchange Commission (SEC) – an entity created by this law. The Securities Exchange Act has been amended several times since 1934. Its current form makes illegal such practices as insider trading and market manipulation. Examples of market manipulations include what is known as pump‐and‐dump transactions, whereby individuals or firms engage in deceptive trades and disseminate inaccurate information to try to push up the price of a security, only to sell it to other investors who are not aware that the market interest and rising prices were contrived. Such manipulations have plagued the penny stock market – unscrupulous investment bankers can more easily manipulate the market for lesser‐known publicly held companies.

Rule 10b‐5 of this law prohibits a variety of fraudulent practices, such as disseminating untrue or misleading information or simply engaging in what this section of the law considers fraudulent securities activities. Violations of this part of the law are often referred to as 10b‐5 violations.

Private Securities Litigation Reform Act of 1995

The Private Securities Litigation Reform Act of 1995 (PSLRA) was the first major revision of the above‐mentioned securities laws. It was designed in part to curb abusive litigation practices by private plaintiffs – especially those related to class actions under Rule 10b‐5. Securities lawsuits can be divided into two categories: public and private lawsuits. Public lawsuits are those pursued by a governmental entity, such as the SEC. However, individual plaintiffs can pursue their own securities lawsuits. PSLRA was designed to curb abuses in private securities lawsuits. PSLRA imposed new requirements on such lawsuits and had an important impact on class action suits. Although this law imposed new regulations on class actions, some of the largest class actions occurred after the law’s passage.

The typical class action follows a drop in an issuing company’s stock price. Plaintiffs file suit alleging that the company made overly optimistic disclosures in light of the information that led to the fall in the stock price. An example of the extreme abuses that the act sought to correct is the lawsuit against Philip Morris that followed the “Marlboro Friday” announcement on the morning of April 2, 1993: Philip Morris announced a $0.41 drop in the price of a pack of Marlboro cigarettes.³ Just five hours later, a class action suit was filed, followed by four more suits later in the day and five more the following Monday.⁴ The courts found the poor quality of the complaints laughable. Plaintiff’s counsel apparently overwrote files from a prior case featuring a toy company, discussing Philip Morris’s future success in the toy industry. Clearly, the speed with which the cases were filed prevented any meaningful research into the grounds for the fraud claims.

Among the regulations this law added is the requirement that settlements be disclosed to the class. Plaintiffs in class actions are required to demonstrate that they did not purchase the securities at the direction of plaintiff’s counsel, that they reviewed and authorized the complaint, and that they disclosed all of their trades of the security in question as well as their activity in other securities lawsuits. In part, the law was designed to try to curb law firms that were pursuing such suits to receive “fee‐mail” in settlements and employing professional plaintiffs who purchase the shares with the intention of suing. Under PSLRA, a defendant can object to the selection of a lead plaintiff if it believes that he or she is a professional plaintiff. The law also governed the use of forward‐looking statements. Such statements need to be clearly labeled as forward‐looking, and they must include appropriate cautionary language.

From a damages perspective, this law requires the plaintiff to demonstrate that the actions of the defendant directly caused the plaintiff’s damages. For example, in cases where a plaintiff alleges that the defendants disseminated financial statements featuring inflated profits, he or she must demonstrate that the stock moved in a manner that adversely affected him or her owing to this false information (rather than market movements). PSLRA also imposed a strong scienter requirement where the plaintiff must be prepared to demonstrate that the defendant acted with a particular state of mind that was bent on fraudulent intent.

In addition to the preceding provisions, this law also sets forth parameters within which damages may be computed in fraud‐on‐the‐market claims. These parameters included a cap on damages. This cap limits a plaintiff’s recovery to the difference between the sale price of a security and its average price over a 90‐day period after the disclosure that is at issue in the lawsuit. The use of such an average limits the short‐term impact of any possible overreaction to the news.

Congress, in passing PSLRA over President Clinton’s veto, attempted to accomplish four goals:

1. Reduce the number of lawsuits that occur after a sharp drop in a company’s stock price (called strike suits) but where the company did not engage in any wrongdoing.
2. Reduce the number of suits aimed at “deep‐pocketed” defendants who may be easily able to settle a suit for an amount that would be relatively small to them but significant compensation to the plaintiff and its counsel.
3. Reduce the use of aggressive discovery practices that amount to burdensome fishing expeditions.
4. Reduce the overly aggressive and unethical practices of some members of the plaintiff’s bar.⁵

Various studies have attempted to measure the impact of PSLRA on the markets. Some initial studies found that the Act had a wealth increasing effect on the value of equities.⁶ Choi, Nelson, and Pritchard found that while there was not enough evidence to conclude that PSLRA led to a reduced incidence of claims that end up being settled for nuisance value, they did find that non‐nuisance claims were less likely to be filed after the passage of the law. Thus they concluded that the law had a measurable screening effect.⁷

A study by Johnson et al found that PSLRA resulted in a shift away from lawsuits based upon forward looking earnings disclosures. They also found that after PSLRA there was a greater correlation between securities litigation and earnings restatements as well as abnormal selling by insiders.(^*)

Sarbanes‐Oxley Act of 2002

Although the Securities Litigation Reform Act of 1995 placed needed curbs on abusive plaintiffs and their law firms and thereby aided potential defendants, it imposed greater requirements on potential defendants. Financial statements that are filed with the SEC must now be certified by the chief executive and chief financial officers of the issuer, who must clearly state that the statements do not contain any information known to be false. The law applies to both quarterly and annual financial statements and the filings that contain them – 10Qs and 10Ks, respectively. The officers who sign such certifications are responsible for establishing controls that ensure their accuracy. Officers who violate this law are subject to fines that can be as high as $5 million; they also can face imprisonment for up to 20 years. If the financial statements have to be restated due to inaccuracies, the officers may also have their compensation forfeited.

The law also imposes restrictions on accountants who certify such statements. It attempts to eliminate a conflict of interest on the part of accountants who are auditing companies and at the same time profiting from other consulting work for the issuer. The law sets forth a variety of services, including actuarial, valuation, broker/dealer, and investment banking, which auditors are expressly prohibited from providing to their audit clients. Accounting firms may perform nonaudit services, but they must be approved by the audit committee and be disclosed in the periodic reports that are filed.

Trends on Class Action Lawsuits

The number of class actions filings declined in 2009 compared to 2008 as the nation entered the subprime crisis. Thereafter the levels remained somewhat below the levels during the years 2000–2008 (with the exception of 2006 when the numbers were unusually low). However, starting in 2005 the volume of class actions picked up and surged to high levels over the years 2018–2019 when the numbers were more than double what they were in 2008–2014 (see Exhibit 10.1)(^*).

Trends in Class Action Settlements

The dollar value of class action settlements has been volatile with no clear trend. Relatively high dollar amounts have been followed by lower values in the next year and vice versa. Naturally, there is a lag between the volume of filings, numbers which were significantly higher over the years 2017–2019, and subsequent settlement values. It is noteworthy that Cornerstone found that the disclosed market capitalization losses in the years 2018 and 2019 were also up sharply.(^*) Thus far, these have not translated into higher settlement values but this could be due to the fact that there has not been sufficient research on the causality and timing of filings and settlement values.

Basic, Inc. v. Levinson

One of the most famous cases in this area was Basic, Inc. v. Levinson. In this case, representatives of Basic, Inc. denied in public statements that merger negotiations were ongoing. However, starting in the fall of 1976, representatives of Basic, Inc. had active discussions with Combustion Engineering, Inc. regarding the possibility of a merger. In December 1978, some investors were surprised to learn that Basic, Inc.’s board of directors had approved a merger with Combustion Engineering.

It was argued that had this fact been disclosed, the stock would have traded at a higher price. This is due to the fact that in acquisitions, target shareholders receive a premium above the stock’s price.⁹ Shareholders who sold their shares, believing that a takeover premium was not forthcoming, may have incurred damages by selling their holdings at a lower sales price. The plaintiffs in this case were one‐time shareholders in Basic, Inc., a manufacturing company whose shares traded on the New York Stock Exchange. They successfully argued that they sold their shares at artificially depressed prices due to the fact that the market was not aware of the impending tender offer from Combustion Engineering.

In Basic, the Supreme Court considered a Sixth Circuit Court of Appeals decision that affirmed the granting of class status to shareholders of Basic, Inc. The district court granted a plaintiff’s motion seeking class certification. The matter ultimately was brought to the U.S. Supreme Court.

Market Conditions and Market Efficiency

Market conditions can affect how efficient markets really are. Market makers and arbitragers play important roles in keeping markets efficient. Market makers bring buyers and sellers together. Arbitragers may bring about the law of one price by trying to identify overpriced and underpriced opportunities. When we have very unusual market conditions, such as what occurred during the subprime crisis, market participants may “run away” from the market, leaving it somewhat devoid of these key participants.¹⁸ When they are not active in the market, it can use lose its efficiency. Thus, for this reason, and others, market efficiency should not be blindly assumed.

Cammer Factors

One year after the Basic decision, a district court in Cammer v. Bloom provided further clarity to applying the court’s reasoning set forth in Basic with respect to how market efficiency can be determined. In Cammer, a defendant, Coated Sales, Inc., filed for bankruptcy following an announcement of its overstated financials over the prior two years.¹⁹ The efficiency of markets became an issue in determining if the markets reflected the false financial information. The defendant’s securities traded in NASDAQ. In reaching its decision the court identified five factors that have come to be known as “Cammer Factors.” They are:

• High weekly trading volume
• A number of market makers and arbitragers
• Analyst coverage
• Ability of the company to file an S‐3 registration statement
• Evidence of a historical empirical relationship between new information and past movements in the company’s stock price.

A high trading volume creates an opportunity for stock prices to possibly better reflect new information. This level of efficiency may be enhanced by market makers who buy and sell the securities of the company. Arbitragers may also trade the securities when they spot inefficiencies. For example, if an arbitrager sees that the stock price is below the value the arbitrager places on the shares, then this investor will see this as a buying opportunity. The opposite would be the case for an overvaluation, which could create a profitable selling opportunity.

The court in Cammer recognized that it could be useful to be able to demonstrate that in the past the stock did respond in an efficient manner to prior announcements of relevant information. This can be done by a statistical analysis that identifies past releases of information and relates that the movements in the company’s stock price – movements that go in the expected direction.

An S‐3 form is the most simplified registration form that public companies file with the SEC in the U.S. To be eligible to file an S‐3 form a company must have been required to report to the SEC for at least 12 months and have filed all their required reports, such as 10Qs, 10Ks, and 8Ks, on a timely basis. There are several other relevant requirements such as having a public float of at least $75 million. Others include the requirement that the company must not have defaulted on any of its debt obligations and must not have missed any dividend payments on its preferred stock. In the Cammer case, Coated Sales had not filed with the SEC for a year and a half, and, therefore, could not file a Schedule S‐3.

A study conducted at NERA tried to ascertain how courts apply Cammer factors. In an analysis of decisions over the period 2002–2011, and the respective factors at play in each case, they found that “in over 98 percent of the cases, the ultimate ruling on efficiency can be predicted by the number of factors that the court found efficient less the number of factors that the court argued against efficiency. When the figure was positive, the court found the security at issue to have traded in an efficient market in all but one instance, while when the figure was zero, the court always found the security to have traded in an inefficient market. Moreover, just limiting the analysis to three Cammer factors (turnover, analysts and market makers) yielded similar results.”²⁰

Other Factors: Krogman Factors

At times courts have considered other factors that might reflect market efficiency. In Krogman v. Sterrit, the U.S. District Court for the Northern District of Texas looked to factors such as market capitalization, bid/ask spreads, and float. Under the Krogman court’s reasoning, companies that have larger market capitalizations may be considered more efficient. The court also viewed companies with more narrow spreads as trading in more efficient markets. Lastly, this court considered that the higher percent of shares that are traded, or floated, in markets, the more efficient the market.

Tests of Market Efficiency

Above we have discussed examples of market inefficiency. Clearly, all markets are not at all times efficient. Markets can vary in the extent to which they are efficient as a function of the specific market, such as New York Stock Exchange as opposed to NASDAQ, as well as the time period at issue. It should be kept in mind that markets can be relatively efficient at times and less efficient at other times. The varying degrees of efficiency are underscored by the award of the 2013 Nobel Prize in economics, which was, ironically, shared by Eugene Fama, of the University of Chicago, who is widely known as the father of the efficient markets hypothesis, and Yale’s Robert Shiller, who is one of the leading researchers on market inefficiencies.

Given that there are three different versions of the efficient markets hypothesis, the research literature in the field of financial economics features abundant tests of each form. For many years this area of financial research served as a research forum for studies by young PhDs seeking tenure and/or promotions in academia.

Class Actions and Reliance

Pursuant to Rule 23(b)(3) of the Federal Rules of Civil Procedure, class action certification requires the plaintiffs to show that the facts or questions of law at issue predominate over the issues that are specific and unique to each plaintiff. For fraud‐on‐the‐market class action lawsuits, a key issue is establishing the reliance. Reliance means that the plaintiff must establish that misstatements at issue motivated an investor’s decision to engage in given transactions. It is not enough that the plaintiff establish damages; they must also demonstrate that the damages were the result of the misstatements.

In Basic the Supreme Court created the rebuttable presumption of reliance by all class members in actions involving misstatements of omissions of relevant information related to their purchase or sales of an issuer’s securities. The significance of this is that a plaintiff does not necessarily need to show that he or she actually read or heard the misstatement. Thus, Basic shifts the burden to the defendant to disprove reliance. Without reliance plaintiffs cannot proceed with 10b‐class actions seeking damages against an issuer for misstatements that may have inflated the price of its shares. If the plaintiffs cannot get the class certified, the suit will be much less likely to brought.

Theoretically, reliance could be much more challenging for a class action as opposed to an individual plaintiff lawsuit.²⁵ However, this was clarified by the decision in Stoneridge Invest. Partners v. Scientific‐Atlanta, wherein the court indicated that reliance is presumed when the information at issue is made public.²⁶ That is, the court in Stoneridge assumed that the information would be reflected in the market price of the security. The court concluded that buyers and sellers would naturally take relevant information that is public into account when they buy and sell the security. This is an important issue, as it could allow the trier of the facts to determined damages through one common process for the whole class.²⁷ In other words, if the matter is a class action, the damages the class is seeking must be calculable in a common process.²⁸

Further clarification was provided by the Supreme Court’s 2014 decision in the Halliburton case.²⁹ Halliburton II provided clarity on the issue of determining class certification but also how event studies should be used in determining damages in fraud‐on‐the‐market loss analyses and in class actions in particular.

That case arises out of a lawsuit related to a merger between Halliburton and competitor Dresser Industries. The incestuous nature of the parties was troubling. Halliburton’s CEO was none other than Dick Cheney – perhaps better known as George W. Bush’s vice president – and for his role in the disastrous Iraq war and the many U.S. soldiers and also civilians who died as a result of their stupid decisions.³⁰ Dresser had huge potential asbestos liabilities that the plaintiffs, including the Archbishop of Milwaukee, alleged were not disclosed. The plaintiff sued for damages based upon what they considered insufficient disclosure.

In Halliburton II the Supreme Court accepted the case so as to consider whether to overrule or modify its holding in Basic that was the law for roughly a quarter of a century. In a majority decision, and over the objections of three dissenting judges, the Court in Halliburton did not overrule Basic and its presumption of reliance. However, the Court did open the door to future defendants challenging class actions by showing that there was not a price impact.

In light of Halliburton, the parties may employ financial economists and econometricians to conduct even studies at the class certification stage (and later) which can be used to measure a price impact or to show that lack of one. For defendants this could be done through an econometric showing that a corrective disclosure (offsetting the misleading prior positive information) did not have a negative impact on the stock price. The opposite would be the goal for plaintiffs to try to show the there was a negative impact. A showing of a negative impact implies that the misstatements did indeed cause the price to be inflated. Obviously, this process is not as simple as a showing of the movement of the defendant’s stock price on the relevant day. Econometrician may be employed to statistically filter out other influences such as the movement of the market as well as other relevant factors. In addition, our discussion here is simple and a given case may involve multiple misstatements and possibly also multiple different corrective disclosures.

Disgorgement

In SEC enforcement actions, the defendant, if found guilty, may be required to disgorge the ill‐gotten gains. This precedent was established in the Texas Gulf Sulphur cases in which various employees of Texas Gulf Sulphur purchased stock and call options in the company prior to an announcement of a major mining discovery.³¹ In these cases, the courts required the defendants to disgorge their profits. The amount to be disgorged can be either the actual profits that the defendant enjoyed or what is referred to as paper profits, defined as the difference between what the defendant paid for the shares and the value for them that the court assesses. This value is determined by relevant stock price data around the time of the event. In the case of positive information on which a defendant traded prior to the release of that information to the public, the paper profits are the difference between the price paid for the shares prior to the public announcement and the full information price. The full information price is determined by judgment based on the time that the market has finished reacting to the new information. Paper profits can thus be different from actual profits. If the illegal trader sells at the full information price, the two might be the same. However, if the trader holds on to the shares and other factors cause the stock price to move to another level, the actual profits will vary and will be determined at time of sale. Assume that the trader holds on to the shares and other factors cause the stock price to fall below the purchase price. If that person sells the shares, he or she may be forced to disgorge positive paper profits when his or her overall trading in the stock actually resulted in losses.

Measuring Damages in Fraud‐on‐the‐Market: Out‐of‐Pocket Damages

The most common method of measuring damages for fraud‐on‐the‐market is an out‐of‐pocket measure of damages.³² This measure draws on the reasoning of Judge Sneed in Green v. Occidental Petroleum Corp.³³ Judge Sneed set forth an acceptable way of measuring such damages as the difference between what he called the price line and the value line. The price line reflects the “corporate defendant’s wrongful conduct.” In a case involving inflation of corporate profitability, this would be the stock price that was a function of the exaggerated income. The value line is the stock’s price in the absence of exaggerated prices. This difference is depicted in Exhibit 10.3, where the shaded area shows the magnitude of the damages that investors may have incurred. The two curves, called lines by the court, start off at the same point t_i and it is assumed that this is the date when the misrepresentation occurred. At this time, it is assumed that an overly optimistic picture of the company’s performance is portrayed, thus causing the market price (as reflected by the price line) to be above what its true value would be absent the misrepresentation.

The measurement of damages can be divided into two parts:

1. Establishment of the loss period. The loss period usually begins with the date when the inaccurate information was released to the market. It usually ends with the date of disclosure.³⁴ If one assumes that markets are very efficient, then the time period narrowly focuses on these two dates. If, however, markets are not assumed to be very efficient, then the loss period becomes less clearly centered on these dates.

   The establishment of the loss period may not be very clear‐cut. There is usually not just one improper announcement or material misrepresentation. Rather, there is often a series of such false statements. As a result, the expert must make a judgment as to the correct start of the loss period. This judgment is made more difficult when there is a series of improper statements or misrepresentation; the expert must make a determination of when the inflating effects on the market began. The end of the loss period may also be unclear. It may be the case that the issuer makes a series of statements that address the initial misrepresentation. This could cause the market to correct in a series of steps. However, it could cause the market to overreact by thinking that the misrepresentation was greater than what it actually was.

2. Measurement of damages. In this model, the measure of damages is the difference between the price line and the value line multiplied by the number of securities in question. The number of shares is estimated using an analysis that attempts to determine how many shares were actually affected and to eliminate shares in which gains offset losses.³⁵ The major challenge here is to calculate the value line. There are different ways to compute this. One is to arithmetically reconstruct what the stock price should have been during the loss period. This method examines the percent change of various other comparable companies and, in doing so, computes the relative change in the stock prices of these companies. This can be done using basic percentages or by employing regression analysis. More formally, the expert may employ what is known as comparable index approach. It attempts to predict the security’s return using explanatory variables such as the market return and the industry return.

Security and Market Returns

The percentage return of security is defined as the combined effects of the income that is received, such as the dividends from a share of stock, and the price changes of the securities (in the form of capital gains or losses). This percentage return has two components, one of which is the dividend yield. It is the component of the return that is attributable to the dividend income.³⁶ This can be expressed as shown in Equation 10.1.

(10.1)

The other component of the total return is attributable to the security’s price movements; it is called the capital gains return. This is expressed as shown in Equation 10.2.

(10.2)

Both components of a security’s return can be combined to form the percentage return of the security. This is shown in Equation 10.3.

(10.3)

The return on the market can be computed in a similar manner using an accepted market proxy such as the Standard & Poor’s 500 market index. This index uses the market value of the securities to measure the performance of the market.

Criticism of the Comparable Index Approach

As with any approach that relies on econometric analysis to estimate the best relationship among variables, there are often disagreements as to whether the expert estimated relationship is the most accurate one. Some contend that the estimate is made more accurate if more explanatory variables, such as more than one index, are added to the model.³⁷

Impact of the Private Securities Litigation Reform Act of 1995 on Damage Computations

The Private Securities Litigation Reform Act of 1995 addressed the way that damages are calculated in Rule 10b‐5 cases. Section 21D (e) of the law states that a damage award shall not exceed the difference between the purchase price of the security and the mean trading price of that security over a 90‐day period. This period begins with the day that the full disclosure, which corrected the misstatement, was made to the market. An exception to this occurs when the security holder sells the shares or repurchases the security prior to the end of this 90‐day period. In this case, the damages cannot exceed the difference between the purchase price and the average of the security’s price at the beginning of the full disclosure period and the price in the subsequent sales transaction.

In Exhibit 10.4, an investor purchases a stock at $30 and the stock price rises to $40. The plaintiffs allege that this rise was due to the false information disseminated by management. When the curative disclosure occurs at time t₀, the price falls to $20. Under the parameters of the new law, damages cannot be measured as the difference between the $40 price and the post disclosure price – $20. Exhibit 10.4 shows that over the 90‐day period, t₀ + 90, the stock price rebounded to $25. Given this rebounding price rise, the losses are the difference between the purchase price ($30) and this price ($25) for total per‐share damages of $5, not $20.

An exception occurs when the shares are sold during the 90‐day period. In Exhibit 10.5, we assume that the investor bought the stock at the same price – $30. It rose to $40 and later fell to $20 on day t₀ after full disclosure. In this example, however, we assume that the stock price falls to $15 per share on t₀ + 30 when the investor sold the stock. The average price in the 30‐day period before the investor sold the shares is $17.50. Damages are the difference between the purchase price ($30) and this average price, or $12.50. This example differs from that of Exhibit 10.4 in that not only was the stock sold prior to the 90 days, but the stock price continued to fall in the 30‐day period – although not as sharply as it did on the announcement date. The further decline in the stock price expanded damages, although not by as much as the difference between the purchase price and the eventual sale price of $15. If the price rebounded steadily after t₀ as it did in Exhibit 10.4, damages would be less, although still higher than if the stock had been held for the full 90‐day period.

Abnormal Returns of the Event Study Methodology

The market model allows for the separation of the total return of a security into systematic and unsystematic components. The systematic component is the part that can be explained by the market’s return – α_i + βi Rmt. According to the market model, the part that cannot be explained by the market’s influences is attributed to firm‐specific effects and is statistically subsumed with the “remainder factor” in the market model’s equation: εit. That is, εit captures the variety of firm‐specific factors including those that are the subject of the litigation – the alleged fraudulently inflated profits of the previous example. The impact of εit can be discerned more readily by arranging Equation 10.5 as follows:

(10.6)

The historical data on Rit and Rmt for the period prior to the event in question, which is taken to be t = 0, are used to econometrically estimate α and β. The values of these two parameters are then used to compute a predicted return, Rit, and its resulting deviation, εʹit, which can then be compared to εit. This represents the deviation of the actual return, inclusive of the effects of the fraudulent behavior, from the predicted return as measured using the market model’s estimated parameters α and β. This discussion is sometimes also expressed through the estimation of abnormal returns where abnormal returns are equal to:

(10.7)

where:

• AR_it = abnormal returns
• αʹi = the estimated alpha
• βʹ = the estimate of beta

EXAMPLE OF A ONE‐PERIOD ABNORMAL RETURN

As an example of a one‐period abnormal return, assume that:

The one‐period abnormal return is simply:

The abnormal returns just computed apply to just one period, i, which may be one day. When the event window is longer, then it is necessary to compute the cumulative abnormal returns (CAR). This is a running total of the one‐period abnormal returns for the length of the study period. This is mathematically expressed as:

(10.8)

where:

• T = the length of the study window

When we compute CART, we have a measure of the total impact of the event.

Decision Rules in Assessing the Significance of a Particular Return Value

The Z test statistics just discussed can be used to assess whether a particular return value is as high or low as it is due to chance or to some nonrandom event. In statistics, the null hypothesis is that the return value is different from some other value by sheer chance. Differences from this selected value – here the average daily return of the security – that are so large that they are above a certain threshold value are then determined not to be a function of normal random movements of the market. They must therefore be caused by some nonrandom process.

A common standard is to use a 5% decision rule, which is based on the normal distribution.⁴⁴ This rule states that if a certain value is 1.96 standard deviations from the mean (above or below – that is, in absolute value), there is only a 5% probability that such a difference is caused by chance. We can make the rule even more stringent by going to a 1% standard; this equates to 2.58 standard deviations, in absolute value, from the mean. However, we can relax the decision rule to a 10% level of significance; this corresponds to 1.65 standard deviations from the mean.

It is important to remember that a decision rule, such as the use of the aforementioned 1.96 value, is really a rule of thumb. More specifically, it is dependent on the degrees of freedom. As the degrees of freedom get larger, we approach the 1.96 value. However, for fewer degrees of freedom, other higher values apply.

When a researcher has more abundant information and knows, for example, the population mean and standard deviation, the aforementioned Z score can be used to determine if a sample value is significantly different from a population mean. Normally in research one does not have such abundant information, and you have to estimate the population standard deviation. Here the t statistic and a t test are used. Also, in cases where the sample size is small, a t test is used. As noted above, the expert needs to be aware of the degrees of freedom and the sample size. With small samples, one cannot blindly apply rules of thumb, such as assuming statistical significance when the t statistic has a value above 1.96. The smaller the sample, the higher the required threshold to assume significance.

Using an Enhanced Market Model

In the comparable index model, both market and industry effects are included as explanatory factors in calculating a predicted return. This has an advantage over the basic market model in that there may be important industry factors that are unique to the industry and that cannot be captured by the market’s variation. These factors can be filtered econometrically by the direct inclusion of an industry return as a separate explanatory variable, as shown in Equation 10.11.

(10.11)

The more important industry factors are in explaining the variation in the firm’s return, the more important it is to use the industry‐enhanced market model. Failing to do so can lead to the erroneous conclusion that all of the variation between the predicted and actual return is explained by firm‐specific factors, including the alleged fraudulent behavior. Including the industry return allows us to filter out industry influences and better isolate the firm‐specific factors.

Factors to Be Considered in Applying the Event Study Methodology

In using the event study methodology, a number of factors have to be considered. These factors can affect the usefulness of this approach as a tool in measuring damage. One of the most fundamental of these factors is the availability of the necessary information. Another is the presence of confounding events that may skew the results. Still another is the time period or event window.

Measuring the Number of Damages Shares

The preceding discussion presents a model for measuring per‐share damages. However, the number of damaged shares still must be explained. Depending on the legal framework of the claims, the number of damaged shares may include only those that traded during the class period. If the number of these shares is known, then the computation involves simply applying the per‐share loss to the number of shares.⁴⁶ Unfortunately, it may not be easy to quantify the exact number of damaged shares. The expert may have to resort to a simplifying process in order to estimate the number of damaged shares. To arrive at such an estimate, certain assumptions about the trading behavior of investors need to be made.

Equal Trading Probability Model

Certain models based on different assumptions about trading behavior exist.⁴⁷ These models differ in how they treat the trading behavior of investors. In one version, the equal trading probability model, sometimes called the one‐trader model, it is assumed that shares are traded only one time during the class period. This assumption can lead to the total number of shares traded exceeding the number of shares outstanding. The assumption is made more realistic by taking into account the fact that shares can be retraded. Under the one‐trader model, the expert assumes that all shares entering the class have an equal probability of being traded as shares that have not yet entered the class. This results in the number of shares entering the class being a function of trading volume and the number of remaining shares that are not yet in the class. One of the criticisms of the one‐trader model is that it does not differentiate between the different types of investors that hold shares in a company. It treats active traders and those who utilize a buy‐and‐hold strategy the same. This assumption has a significant impact on the damages that result from the use of this model.

Dual Probability Trading Model

An alternative to the equal probability model is the dual probability model. Under the dual probability trading model, sometimes referred to as the two‐trader model, traders are categorized into two different classes: active traders and traders who utilize a buy‐and‐hold strategy. The model requires the expert to assume a certain probability of trading for active traders and inactive traders. The expert may also assume the percent of total shares outstanding held by these two groups of traders. Under this model, shares are retraded among active traders, resulting in a lower number of shares entering the class. With a lower number of shares in the class, the dual probability trading model results in a lower damage estimate.

The dual probability trading model is more realistic than the equal probability trading model. The latter results in higher losses and thus is more in favor of the plaintiff. However, the dual probability model is influenced by the different trading probabilities and share percentages assigned to the two groups of traders. The expert may want to draw on research to support these assumptions. One such source of data is depository records, which, when combined with trading volume data, may help the expert select the relevant parameters needed to use the dual probability trading model.

Proportional Trading Model

Still another stock trading model is the proportional trading model.⁴⁸ In this trading model, the number of shares traded is differentiated from those that were retained. The model assumes that there is an equal probability that a given share came from a pool of those shares that had already been traded during the loss period and those that were being traded for the first time. This assumption is applied on a daily basis throughout the loss period. The number of shares are retraded, as opposed to those that enter the class due to being traded for the first time, being a derived process using the equal probability assumption. This differs from the dual probability model in which probabilities that may not be equal are assigned to the two groups of traders.

Longer‐Term Effects

The event study model is designed to try to isolate the impact of a discrete event that occurs at a specific time. For example, event studies as applied to financial markets litigation may measure the response to an event such as a public announcement made to the market. The wider the time event window being considered, the greater the possibility that any abnormal returns that are measured may be a function of other, confounding events. This is a limitation of the traditional event study model. It is limited in its ability to capture a pattern of behavior that took place over an extended time period.

As an example, consider a case of a somewhat smaller, public drug developer/manufacturer that has developed a drug that needs to undergo drug testing for the FDA approval process. There are companies that specialize in such work. The drug manufacturer may hire a testing firm that is in the business of doing such large‐scale tests designed to meet the FDA stringent requirement. That firm will gather the subjects for the testing and conduct the tests. Such testing may take an extended time period. If, however, that company is negligent in conducting the testing, there may be a significant lag between the time the drug developer becomes aware of problems with the conduct of the tests and this, in turn, may result in a gradual dissemination of information to the market. A short‐term event study model may not capture this behavior.

In a situation such as described above, there is no single short‐term event window. Rather, there is a longer‐term deterioration in the market value of the company’s stock as, for example, the market and the industry participants grow in value while the plaintiff’s market value languishes, or declines, due to the ineptitude of the testing firm. In this situation, the expert needs to create a model that includes both a market measure, such as the S&P 500, and an industry index, such as a pharmaceutical index. Nowadays there are many available industry indices. The expert needs to be aware of the components of the chosen index to make sure it includes companies that are sufficiently comparable. In addition, the expert will test that appropriateness of the index and the market measure, using data from the estimation period, to ensure that they adequately explain the dependent variable.

In addition to doing tests of the statistical significance of the explanatory variables, such as testing the significance of the coefficients of the market index and the industry index, the expert can use test the accuracy of the model by examining forecasts using the model and its coefficients, and see how well it predicts during a period where the values of the plaintiff’s stock price are known. If the expert can demonstrate that the model predicts the company’s stock price well, there is greater assurance that it is a robust model. Conversely, if the model fails to predict well, an opposing expert can exploit this when criticizing the plaintiff’s expert.

Standard Elements of Lost Profits Damages Involved in Broker Raiding Cases

As we have noted, the main elements of lost profits damages claims apply to broker raiding. They include the requirement that damages must be measured within a reasonable degree of certainty. Such profits cannot be speculative or generally uncertain. The broad approaches of the before and after method and the yardstick approach can be relied on. Only lost net profits can be claimed. As a result, lost revenues net of the costs associated with the achievement of those revenues are the measure of damages. In addition, mitigation must be incorporated into the analysis. Claimants often make the standard arguments that the revenues and profits associated with business that the respondent asserts constitute mitigation could and would have been achieved while still enjoying the lost revenues and profits (“but for” the respondent’s conduct). While there are many common elements of all lost profits claims, there are some important factors in broker raiding claims that are worth emphasizing.

Merger Laws

Acquisitions of publicly held companies are subject to certain laws that regulate share purchases and takeovers. The most prominent of these is the Williams Act.

This law, enacted in 1968, is an amendment to the Securities Exchange Act of 1934. In an effort to regulate hostile takeovers, the law seeks to enhance the confidence of investors in securities markets by providing them with certain information and time to deliberate over that information.

Pursuant to Section 13(d), the Williams Act requires the filing of a Schedule 13D within 10 days of the acquisition of 5% of a public company’s outstanding equity securities. The disclosure requires the acquirer of the stock to reveal certain information, such as the identity of the purchaser and the plans the purchaser has for the stock and the issuing company. Section 14(d) of the law requires the filing of a Schedule TO (originally was called Schedule 14D‐1) upon the initiation of a tender offer. Ironically, the act did not define what a tender offer is. This was done by a subsequent court decision where a tender offer was defined by the presence of eight factors – called the Eight Factor Test.⁵⁵ Simply stated, however, a tender offer is an offer made directly to the shareholders of a public company.

Schedule TO requires the disclosure of information similar to that required on Schedule 13D. In addition, the bidder has to reveal such relevant information as the terms of the offer as well as information on the bidder and its sources of financing. Because target companies tend to receive a takeover premium (a monetary amount in excess of the stock price prior to the announcement), stock prices rise on the announcement of takeovers.⁵⁶ For this reason, bidders may try to conceal their plans when making share purchases. In this way, they try to avoid paying some component of the takeover premium when they accumulate the target’s stock prior to the initiation of a formal bid. When such efforts lead to an evasion of the requirements of securities laws, the SEC or private parties may initiate a lawsuit.

Disgorgement in Merger‐Related Transactions

As noted, acquirers of stock in targets or potential targets of takeovers may want to avoid necessary filings, such as the filing of a Schedule 13D. With some exceptions, those who purchase shares in excess of 5% of the outstanding shares have to file a Schedule 13D with the Securities and Exchange Commission within 10 days of acquiring that holding.⁵⁷ Such filings, along with Schedule TO filings, which are required at the initiating of a tender offer, became required in 1968 pursuant to the passage of the Williams Act.

This allows the purchaser to buy the stock at a lower price than what he or she would have to pay if the market had the type of information available that disclosure pursuant to the Williams Act would provide. There is support in the finance literature for a general positive stock price effect associated with certain types of Schedule 13D announcements, such as ones that expose an increased likelihood of a takeover. Such purchases were not for the purposes of takeovers.⁵⁸ Additional purchases require the filing of an updated Schedule 13D. These filings provide great detail on who the purchaser is and what his or her intentions are. In the 1980s, certain individuals pursued various schemes including stock parking – the placing of purchased shares in names other than the true purchaser to avoid the requirements of Schedule 13D. Various actions were brought against them, including some high‐profile cases in which the purchasers were found to be criminally liable.

A typical illegal scenario is where a purchaser acquires shares but chooses not to make a timely filing as, if the market was aware of an intention of the purchaser to make a bid for the entire company, it would have to pay a higher market price. This is because most takeover bids come with the offer of a takeover premium that may be in the range of 30%–40%, depending on market conditions as well as the specifics of the target company and the industry.⁵⁹ Thus, there is a financial incentive to avoid the requirements of the law.

A methodology similar to what is used in the event study approach can be applied to determine the purchaser’s gains made through the evasion of the disclosure requirements of the Williams Act. The expert needs to compute the full information price – the price that the stock would have traded at had the market been aware of the purchaser’s acquisition of the shares. The amount to be disgorged is the difference between the purchase price and the full information price times the number of shares involved.

Another type of M&A‐related action in which an expert might become involved is insider trading cases. A common scenario might be where an individual learns of an impending deal and he or she, or other associates, might either purchase the target’s shares or trade in options. The expectation would be that the shares that are purchased might rise in value once the deal was announced. The SEC tries to police such transactions but research studies have showed that stock prices tend to rise in advance of announcements of deals, leading to a suspicion that many illegal trades are not detected or pursued by the SEC.

Another type of M&A‐related action in which an economic expert might become involved is insider trading cases. A common scenario might be where an individual learns of an impending deal and he or she, or other associates, might either purchase the target’s shares or trade in options. The expectation would be that the shares that are purchased might rise in value once the deal was announced. The SEC tries to police such transactions but research studies have shown that stock prices tend to rise in advance of announcements of deals, leading to a suspicion that many illegal trades are not detected or pursued by the SEC.

Deal Cancellation and Material Adverse Change Clauses

M&A deal agreements often include a material adverse change clause. These clauses are generally written very broadly so that the buyer can walk away from the deal if major unexpected events occur. Sometimes the possible cancellation may be quite possible. For example, the deal participants may be concerned about getting antitrust or other regulatory approval. Often the buyer wants the seller to agree to a termination fee in case the deal is cancelled. For example, in the wake of the subprime crisis, an event that lowered takeover premiums while also increasing the risk premiums built into DCF valuations, many buyers wanted to walk away from deal they had previously agreed to but had not yet closed. This occurred in 2007 when the private equity firm Cerberus Capital cancelled its acquisition of United Rentals.

Damages Involving Smaller Closely Held Deals

As we noted above, participants in smaller M&As may claim different types of damages than their public counterparts. For example, in smaller deals there may be post closing purchase price adjustments. These adjustments are more short‐term oriented and often focus on post closing financial statements. With such disputes, the parties may employ accounts to review and analyze the financial statements produced, as well as the underlying financial data. The parties may also disagree on the accounting methods used.

Disputes related to earn‐outs may arise in M&As. A buyer of a closely held company may make some of the purchase price a function of the achievement of certain longer‐term financial targets, such as specific revenue, profit, or EBITDA levels. The buyer may require the prior owners to work at the firm for a certain time period and may use earn‐out clauses as a way of trying to ensure that the management will work diligently to ensure that the company continues to perform well. The longer the earn‐out period, the greater the opportunity for a dispute.

When buyers provide earn‐out clauses, they often want not just certain revenue levels to be achieved but they want cost levels controlled as well. Revenues that yield much lower post acquisition profitability before the deal are less valuable. One of the problems that can arise is when we go further down the income statement and focus on various different costs, there may be more opportunity to have a disagreement.

The measure of performance can become convoluted when the target is merged into the buyer company. Often the seller tries to maintain some separate financial records but the more the target is integrated into the buyer, the harder it is to determine what the seller’s stand along performance really was. The greater the degree of integration, the more likely the seller may argue that any failure to achieve certain performance targets was due to the buyer’s actions.

Measuring Damages

When measuring damages in broker‐client cases, the claimant tries to show what the value of its investment would be “but for” the improper actions of the broker. For example, if the broker invested the claimant’s portfolio in investments that the claimant asserts were unsuitable and that lost value over the alleged loss period, then the claimant will try to show what the value would be if the portfolio had been invested properly. This may involve stipulating what a more reasonable portfolio would have been and would have earned over that time period. Such an assessment may be done through using broad categories of investments, such as a diversified portfolio of bonds, large company stocks, and so on. Statistics are available through sources such as Morningstar to show how these broad categories have performed over different historical time periods. In response, brokers raise a variety of defenses (which are too numerous to cover here). We discuss damages further in the next section.

Elements of a Churning Claim

There are four elements of a churning claim:

1. Control
2. Suitability
3. Scienter
4. Excessive trading

The term “control” refers to the extent of the customer’s involvement in the trading decisions. Control is clear when the customer had no involvement with the trading decisions and when the broker did not have to consult the customer in order to make trades. Brokers or registered representatives often attempt to acquire such control through an explicit written agreement rather than through oral authority. The situation becomes less clear, however, when the customer provides such written authority but also orally instructs the broker to consider this authority limited to certain trades or types of securities. The fact that many customers fail to carefully read a written agreement and place greater weight on the oral representations of brokers makes some churning disputes problematic for arbitrators.

The term “suitability” refers to the recommendation of certain investments that are suitable for the customer in light of his or her personal financial circumstances or investment objectives. Broker recommendations to purchase speculative securities for customers who lacked a background in finance or securities trading and who had limited financial resources have been found to be unsuitable. It may be the case that margin trading itself is indicative of unsuitability. In such cases, it may be necessary to present expert testimony on the degree of financial sophistication necessary to fully comprehend the risks of particular types of transactions. For example, such expertise is useful in demonstrating why buying uncovered call options is a speculative transaction. Expert testimony could be used to explain why a naive investor might have difficulty comprehending the risks and potential for gain in complex option trading strategies, such as the use of butterfly spreads or straddles.

In proving a churning claim, the plaintiff may need to demonstrate scienter. This refers to the assertion that the defendant acted either with the intent to defraud the customer or with careless disregard for the financial interests of the client. In 1976, scienter became necessary in order to establish a claim of churning under Section 10(b) of the Securities and Exchange Act of 1934. Used in the context of churning, “scienter” refers to a mental state embracing the intent to deceive, manipulate, or defraud. While scienter is an integral part of the churning claim, the financial expert may not play an active role in proving this allegation.

The final necessary aspect of a successful churning claim is the demonstration of excessive trading in light of the objectives of the account. Such a demonstration attempts to show that the volume of trading was so excessive that the goal of the trading activity was to generate profits for the broker rather than gains for the customer. It is here that certain quantitative measures of trading activity can be employed to relate the pattern of trading in the accounts to accepted legal standards of churning.

Quantitative Measures of Churning

There are three main quantitative measures of churning:

1. Turnover ratios
2. Average holding periods
3. Cost of investment services

A successful churning claim may require the combined use of all these measures. High turnover is often associated with a low holding period, which implies high trading costs. These high costs may mean that the trading strategy is required to yield a high rate of return simply to break even.

TURNOVER RATIOS

The most common formulation of turnover ratios is the Looper formula. The Looper formula is defined as follows:

(10.12)

Average equity is defined as:

(10.13)

For example, an account that had an average equity of $100,000 and purchases of $850,000 in a given year has a turnover ratio of 8.5.

Two of the most commonly used periods are monthly and annual periods. Monthly periods can be particularly useful for higher volumes of trading. In the case of margin accounts, the debit balances must be deducted from the equity measure. In addition, it is generally better to include all accounts in the computation of average equity. For example, if the investor has both a cash and a securities account, both should be combined to arrive at total average equity. However, it may be useful to segregate the trading of certain accounts, such as options accounts. If this is the goal of the analyst, a separate turnover analysis can be computed that excludes the options account. Given the inherently short‐term nature of options, it may be useful to present a separate analysis in which these trades are excluded. Another ratio, the contract index – which is the ratio of the number of option contracts bought or sold within a given period divided by the average equity – is sometimes used as an alternative to the traditional Looper turnover ratio.

Given the fact that both purchases and sales generate commissions for the broker, it might also be useful to include in the Looper formula both types of transactions in the numerator, as in Equation 10.14.

(10.14)

While Equation 10.16 is more revealing about the total value of trading and the resulting commission income proportional to the trading volume, it may be counting the same trade twice. If a broker sells one security in order to buy another, this can be considered one transaction. The second security has replaced the first. However, as far as commission accounting is concerned, each is a separate transaction that generates a commission charge. The debate becomes moot, however, in light of the reasons for the computation of the turnover ratios. The value of the specific turnover ratio utilized is to derive a quantitative measure of trading activity that can be compared with other quantitative measures that have been indicative of churning in prior litigation. Therefore, the standard that is ordinarily relied on is the traditional Looper formula (Equation 10.14), which includes only purchases in the numerator. Although the analyst might find it enlightening to compute both measures, the standard Looper formula is more useful for comparison purchases.

TREND IN AVERAGE HOLDING PERIODS

There have been many case decisions from years ago that relied on reflected trading data from different market conditions. However, it is important to bear in mind that the average equity holding periods have declined over time. This can be seen from a review of Exhibit 10.9. Thus, an evaluation of holding periods in a given case needs to take this into account. It is also important to bear in mind that given that institutions account for the bulk of stock ownership, these changes in holding periods are greatly influenced by institutional investment objectives – objectives that many differ from the investment mandate set forth by a plaintiff in a churning case.

Holding Periods	Excessive
Percent held for less than 6 months	75% or more
Percent held for less than 3 months	50% or more
Percent held for less than 1 month	25% or more

Measuring Damages

The typical churning case involves a claimant who entrusted funds to a broker with the expectation of achieving rates of return that are consistent with the historical experience in the market. This may involve a diversified portfolio of stocks, bonds, and money market investments. If this is the case, the expected value of the portfolio can be measured using a historical weighted average rate of return based on readily available annual return statistics. However, the actual rate of return over the loss period is more relevant. This can be computed in the same manner using the actual rates that prevailed for each of the broad categories of stocks, corporate bonds, and Treasuries; these are combined in a manner that is similar to the composition of the portfolio. The end of the loss period is either the date of the litigation or the date that the claimant ended the relationship with the broker and withdrew the funds. The actual value of the portfolio at the end of this time period can then be deducted from the expected value to determine the damages.

The argument sometimes is made that broad market statistics are inherently different since they are based on a larger portfolio of securities. However, this argument is opposed by the abundant research in the field of finance, which shows that as portfolios are diversified, they approach the rate of return for the market. That is, as portfolios become increasingly diversified, the unique, unsystematic risk of the portfolio decreases and approaches that of the market. If the investment objective is simply to maintain an original portfolio, or if the broker uses his or her own judgment to purchase a specific portfolio that differs in performance compared to the market, a specific rate of return for the portfolio in question can be computed using the CRSP monthly return database discussed earlier in this chapter.

The loss computation can be complicated by the fact that there may be some withdrawals from or deposits into the account. Deposits can be handled individually using the same rate‐of‐return computation that is applied to the original portfolio but with a different starting date. Withdrawals also are handled using the same method, but they are given negative values to offset the larger portfolio compounded values.

Conclusion of Churning Analysis Model

The model and approach described in this chapter are methodologically straightforward, and the computations involve basic mathematics. However, the computations can be a time‐consuming and arduous task. The analyst should therefore be able to rely on good spreadsheet or database software to generate the large volume of calculations that are often necessary in active brokerage accounts. While the calculations themselves are basic, their value in litigation can be substantial. Without definitive quantitative evidence, the claimant would have to rely simply on the impressions of the industry expert and the arbitration panel. However, with detailed quantitative analysis, the extent of the trading and its related costs can be precisely demonstrated. Such evidence can also enable the claimant to counter the internally generated analysis that is often presented by the defendants. In addition, a clear model such as the one presented in this chapter can be invaluable to defendants in actions where churning has not occurred. The same analysis that a plaintiff might present in cases where there was excessive trading can also be used by defendants to show that the trading was not in excess of the legal standards.