8/11/2019 Sizing Securities Fraud Damages Constant Percentage

1/3

Over the last several years,a debate has quietly beenraging between plaintiffs anddefendants, and their respectiveeconometric experts, about

how to properly calculate damages in

securities fraud cases arising under10(b) of the Securities Exchange Act of1934.

Two of the principal methods for doingso are the constant percentage methodand the constant dollar method. As thenames suggest, the constant percentagemethod measures inflation caused byfraud as a constant percentage of stockprice throughout the class period, whilethe constant dollar approach measuresinflation as a constant dollar amount ofstock price.

The two approaches can result indramatically different damages calculations.

When the price of a companys stockis declining during the class periodascenario that is very typical in securitiesfraud casesthe constant percentagemethod will generally result in significantlygreater per share damages figures than theconstant dollar approach. Perhaps notsurprisingly, the constant percentage tendsto be the method of choice for plaintiffs,while defendants principally rely on theconstant dollar approach.

While plaintiffs and defendants joustedover these two methodologies, the courtsstood by largely silent on the issue.That remained the case until the U.S.

District Court for the Northern District ofOklahomaa district not necessarily knownfor being a hotbed for securities lawjurisprudencetackled the debate head-on inIn re Williams Securities Litigation.

In excluding the damages and losscausation report of plaintiffs expert, theWilliams court found that the constant

percentage method was in direct conflictwithDura Pharmaceuticals Inc. v. Broudo,the controlling Supreme Court precedenton loss causation. Securities litigators andtheir experts should pay heed to Williams.To the extent that this well-reasoneddecision starts a trend in the case law,

use of the constant percentage method insecurities fraud cases may become a thingof the past.

The Debate

The Constant Percentage Methodvs. the Constant Dollar Method

Damages in 10(b) cases are premisedon the notion that, during the class period,the companys stock is artificially inflateddue to a material misstatement or omission,and the inflation is then removed when thetruth is disclosed to the market and thestock price declines.

The idea is that investors who purchasedthe stock during the class period paidan inflated amount because the market

overvalued the shares due to false oromitted information, and then weredamaged when the truth was revealed andthe stock price drops.1The inflation onany given day in the class period is thedifference between the actual stock price

and the price the stock would have sold forabsent the misstatement or omission, anddamages for shareholders are calculated bymeasuring the difference between inflationat purchase and inflation at sale.2

There are two principal ap-proachesexperts have used to measure the inflationmeasuring it as a constant percentage ofstock price (the constant percentagemethod), or measuring it as a constantdollar amount of stock price (the constantdollar method).3The two approaches canlead to drastically different results.

An Example

The best way to understand thedifferences is through an example.Assume there are two investors, A and B.A purchases the stock of the company onDay 1 for $200. B purchases the stock ofthe company on Day 2 for $100 because theprice has declined for reasons unrelatedto the alleged fraud. On Day 3, the fraudis fully revealed by a corrective disclosureand the residual drop in the stock (in otherwords, the price drop after controllingfor market, industry and other nonfraudinfluences in the price movement) is $50Because the full truth has been revealedas a result of the corrective disclosure on

Day 3, the inflation at the end of Day 3 isnow considered to be zero.Using the constant dollar approach

to measure fraud-based inflation in thisscenario is relatively straightforwarditposits that the inflation is simply theamount of the residual price decline onDay 3, or $50 in this example.4Thus, bothAs damages and Bs damages would be$50 because that was the dollar declinecaused by the disclosure of the revelationof the truth.

The constant percentage approach, onthe other hand, would take the percentage

SERVING

THEBEN

CH

AND

BARSINCE

1888

VOLUME 241NO. 13 WEDNESDAY, JANUARY 21, 2009

Sizing Securities Fraud Damages:Constant Percentage on Way Out?

Outside Counsel Expert Analysis

JEFF G. HAMMEL and B. JOHN CASEY are partners in thesecurities litigation and professional liability groupof Latham & Watkins.

2009 INCISIVE MEDIA US PROPERTIES, LLCWWW. NYLJ.COM

The two key ways tomeasure the

difference between stock-price

inflation at purchase and inflation at

sale are as a constant percentage

of price or as a constant dollar

amount of price.

ByJeff G.Hammel

AndB. JohnCasey

8/11/2019 Sizing Securities Fraud Damages Constant Percentage

2/3

decline of the residual dropin percentageterms, the $50 residual decline on Day 3represents a 50 percent decline from theprevious days priceand apply thatpercentage to the stock price throughoutthe entire class period.

In our example, A and B would beconsidered to have purchased at thesame inflation percentage (50 percent),but they would recover different dollaramounts because their purchase priceswere different. As damages would be $100(50 percent of the $200 purchase price),while Bs would be $50 (50 percent of the$100 purchase price). So in this example,use of the constant percentage approachincreases As damages by $50.

Common Pattern

Importantly, this example, while overlysimplistic, represents a common pattern in

securities fraud cases. Indeed, as we sawwith many of the securities fraud casesfiled in connection with the bursting ofthe Internet bubble in the early 2000s,there very often is a steady decline inthe companys stock before the truthof the fraud is disclosed through acorrective disclosure. When the price ofthe stock is on a downward trend duringthe class period, the percentage methodwill generally result in a much largerper share damages figure than will theconstant dollar method, as our exampleillustrates.

Proponents of the constant percentagemethod argue that, as a matter of

economic theory, the market reacts tonews in percentage, not dollar, terms.5Opponents, on the other hand, haveprincipally launched a legal attack againstthe percentage method. Specifically, theyhave argued that the constant percentagemethod is inconsistent with the conceptof loss causation because (i) it awardsdifferent damages to stockholders whosuffered the same loss, and (ii) it positsthat a shareholder may recover for aloss not associated with the fraud.

In our example, opponents of theconstant percentage method would pointout that not only does A get to recover

damages ($100) that are greater than Bsdamages ($50) even though they wereexposed to the identical fraud, butthat As damages are greater than theactual dollar drop associated with the fulldisclosure of the fraud ($50). Given thatdamages for a shareholder are calculatedas inflation on the day of purchase minusinflation on the day of sale, critics wouldalso point out the fact that had InvestorA sold the stock on Day 2the day beforethe truth of the fraud was revealedtheconstant percentage method would positthat A suffered a recoverable loss of $50because the dollar inflation on Day 1 (50

percent of $200, or $100) is $50 greater thanthe dollar inflation on Day 2 (50 percentof $100, or $50). Critics would argue thatthis result is contrary to principles of losscausation because a loss that occurredbefore the fraud was revealed could not

have been caused by the fraud.Duras Impact

In Dura Phar mace ut ical s, the U.S.Supreme Court addressed the pleadingrequirements for loss causation insecurities fraud cases. The specific issueinDura was whether merely pleading thatthe stocks purchase price was inflatedby fraud was sufficient to satisfy 10(b)srequirement of loss causation, i.e., thatthe alleged misrepresentation or omissioncaused the loss for which the plaintiffseeks to recover. 544 U.S. at 346 (citing 15U.S.C. 78u-4(b)(4)). The Supreme Court,

reversing the U.S. Court of Appeals for theNinth Circuits decision, found that merely

pleading an inflated purchase price wasnot enough to allege loss causation:

[A]n inflated purchase price will notitself constitute or proximately causethe relevant economic loss. Id. at 342.The Court reasoned that, at the time

an investor purchases a stock inflated byfraud, the investor has suffered no loss:the inflated purchase payment is offset byownership of a share that at that instant

possesses equivalent value. Id. If theinvestor sells the shares quickly beforethe relevant truth begins to leak out, themisrepresentation will not have led to anyloss. Id. Additionally, the Court noted that,even if a shareholder sold the stock at aloss after the relevant truth was revealed,the investor may not have suffered arecoverable loss under the securitieslaws because that lower price mayreflect, not the earlier misrepresentation,but changed economic circumstances,changed investor expectations, newindustry-specific or firm-specific facts,conditions, or other events, which taken

separately or together account for someor all of that lower price. Id. at 343.

The securities laws, held the High Court,were meant not to provide investors withbroad insurance against market losses, butto protect them against those economic

losses that misrepresentations actuallycause. Id. at 345. And a misrepresentationwill actually cause a loss only when therelevant truth of that misrepresentationhas been revealed, and where a plaintiffcan show that the revelation itselfasopposed to other factors unrelated to thedisclosureresulted in the loss. Id.

Although the decision did not directlyaddress damages calculations in securitiesfraud cases,6opponents of the percentagemethod seized on the Courts discussionof loss causation principles to argue thatthe percentage method is an inappropriatemethod of measuring damages. Specificallythey have argued that the percentageapproach is fundamentally inconsistentwith the holding inDura because it positsthat a shareholder may suffer (and recoverfor) a loss that occurred before therelevant truth is revealeda losswhich, by definition, could not have beenactually caused by the fraud.7

The Williams Decision

Although the opponents of the constantpercentage method viewedDura as theproverbial nail in the coffin for the constantpercentage method, courts still werereluctant to address the issue head-on.8

That changed in the Williams SecuritiesLit igation case. See 496 F.Supp.2d 1195(N.D. Okla. 2007). Williams was a 10(b)case brought by purchasers of securitiesissued by the Williams CommunicationsGroup Inc., a company in the fiber opticsbusiness. The price of the companysstock, like the stock of its competitors inthe telecommunications industry at thetime, was on a steady decline during theclass period (July 2000-April 2002). Indeedby the time of the first alleged correctivedisclosure (Jan. 29, 2002), the companysstock price was only $1.60.

Plaintiffs alleged that management ofthe company and the companys auditors

misrepresented the financial condition ofthe company. In support of their claimsplaintiffs submitted expert testimony onthe issues of materiality, loss causationand damages. Defendants moved toexclude that testimony on the groundsthat it did not pass muster underDaubertv. Merrell Dow Pharmaceuticals Inc., 509U.S. 579 (1993). One of the damagesscenarios (Scenario 2) set forth byplaintiffs expert was premised on theconstant percentage method. Due to thefact that the stock was decliningindeedwas in a free fallduring the class period,the experts constant percentage model

WEDNESDAY, JANUARY 21, 2009

The Williams court found that

the constant percentage method

was in direct conflict with Dura

Pharmaceuticals, the controlling

Supreme Court precedent on loss

causation. Securities litigators

should pay heed to Williams. [If it]

starts a trend in the case law, use of

the constant percentage method insecurities fraud cases may become a

thing of the past.

8/11/2019 Sizing Securities Fraud Damages Constant Percentage

3/3

showed shareholders being damagedprior to the first alleged correctivedisclosure.

Recognizing that this result wouldbe in conflict with Duras holding thatdamages are not recoverable prior to the

disclosure of the relevant truth, plaintiffsexpert modified his constant percentageapproach by only allowing damages toshareholders who held stock past thefirst corrective disclosure (identified byplaintiff as Jan. 29, 2002). 496 F.Supp.2dat 1260. The court noted that, althoughthe modification attempted to make themodelDura-compliant, it resulted in aninternal inconsistency:

[Plaintiffs expert] acknowledgedthat he could give no economic orlogical reason why a shareholderwho sold on Jan. 29 would have aclaim and a shareholder who sold

on Jan. 28 would not have a claim,explaining that [his damages] scenario2 was produced to meet some sortof legal definition of damages. Id.Furthermore, the court noted that the

constant percentage method, even with theplaintiffs experts proposed adjustment,resulted in a situation where an investorwho sold her stock after the first correctivedisclosure was recovering not just forthe loss caused by that correctivedisclosure, but for losses suffered priorthereto (a period during which the entiretelecommunications sector was sufferingmassive losses):

The result [of the percentagemethod]is that a small loss in sharevalue (on or after Jan. 29) in dollarterms translates into a large dollarrecovery per share for shares boughtearly in the class perioda recoveryin dollars per share that is attributablemostly to share price declines thatoccurred beforethe first correctivedisclosure. Id.; see also id. at 1269([U]nder [Damages] Scenario 2, aninvestor who sold on Jan. 28, 2002gets no recovery, while an investorwho sold on Jan. 29 recovers for allof his loss of per-share value for theentire class period (nearly $25 per

share for an investor who bought atthe beginning of the class period),even though the price of the stock hadalready declined 94 percent, to $1.60before (as postulated in Scenario 2)the relevant truth emerged. (emphasisin original).The court held that this result was

incompatible withDura because a loss invalue occurring before the first correctivedisclosure cannot be considered causallyrelated to [defendants] fraudulent[conduct]. Id. at 1269-70 (citationomitted). In the courts view, theapplication of the constant percentage

inflation approach would give the equityinvestor the partial downside insurancepolicy which Dura counsels that thesecurities law should not provide. Id. at1270.9

Although the court noted in a footnote

that it could conceive of a case involvinga short class in which the constantpercentage approach would at leastclear theDaubert hurdle, it offered noreal guidance as to when that approachwould ever be appropriate. Id. at 1270,n. 54. Indeed, the fundamental premiseof the courts rulingthat the constantpercentage method is contrary toDurabecause it allows for the recovery ofdamages not tied to the disclosure of thefraud seems totally at odds with anyapplication of the constant percentagemethod.10

Conclusion

For critics of the constant percentagemethod, the Williams case did explicitlywhat the Dura case did implicitlycondemn the constant percentage methodas being inconsistent with the principlesof loss causation in a securities fraud case.Whether Williams starts a trend in the caselaw remains to be seen, but it certainly hasthe potential to be a watershed case onhow damages are calculated in securitiesfraud cases.

1. See David H. Topol, Attacking Plaintiffs-StyleDamages During Mediation of Securities Cases,PROFESSIONAL LIABILITY UNDERWRITING SOCIETYPLUS, May-June 2004.

2. As discussed below, Dura now prohibits therecovery of damages for shares sold before therelevant truth is disclosed about the fraud. See

Dura Pharmaceuticals Inc. v. Broudo , 544 U.S. 336,342-46 (2005). Additionally, damages in 10(b) casesare subject to the 90-day look back provision ofthe Private Securities Litigation Reform Act, whichprovides that the award of damages to the plaintiffshall not exceed the difference between the purchaseor sale price paid or received, as appropriate, by theplaintiff for the subject security and the mean tradingprice of that security during the 90-day periodbeginning on the date on which the informationcorrecting the misstatement or omission that is thebasis for the action is disseminated to the market.15 U.S.C. 78u-4(e).

3. For a more expansive discussion of themethodologies for calculating damages in securitiesfraud cases (and an analysis of the impact of Dura

and Williams), see David Tabak, Inflation andDamages in a Post-Dura World, (Sept. 25, 2007),http://www.nera.com/Publication.asp?p_ID=3287.

4. This would be the commonly used back-casting method of measuring inflation, whichdetermines inflation by measuring the pricedecline caused by the corrective disclosure, andapplying that price decline, in dollar or percentageterms, throughout the entire class period. Theless common forward casting approach, onthe other hand, measures inflation by examiningthe stock price increase when the allegedmisstatements/omissions were originally made,not the decline when the truth was revealedthrough a corrective disclosure. See generallyALLEN FERRELL & ATANU SAHA, The LossCausation Requirement for Rule 10b-5 Causes ofAction: The Implication of Dura Pharmaceuticalsv. Broudo, The Harvard John M. OlinDiscussion Paper Series 596 (August 2007), http://

www.law.harvard.edu/programs/olin_center/Ferrell_et percent20al_596.pdf.

5. See Daniel P. Lefler & Allan W. Kleidon, Just HowMuch Damage Did Those Misrepresentations ActuallyCause and to Whom?: Damages Measurement inFraud on the Market Securities Class Actions, 1505PLI/Corp. 285, 295-296 (2005).

6. The constant dollar versus constant percentage

issue was briefed in Dura, but the Court did nogive an opinion on the issue. See, e.g., Brief oBroadcom Corporation as Amicus Curiae in Supporof Petitioners, 2004 WL 2075751, at *10 (Lossescaused by a misrepresentation should be measuredby the dollar decline when the market learns of amisrepresentation. Using the percent decline asopposed to the dollar decline leads to perverseresults. For example, the percent method can leadboth to awards of damages to stockholders whosuffered no losses caused by a misrepresentationand to different treatment of stockholders whoseactual losses caused by a misrepresentation are thesame.).

7. See, e.g., Lefler & Kleidon, supra n. 98. A number of post-Dura cases have held tha

damages may not be awarded for losses sufferedprior to the revelation of the truth of the fraudwhich can be interpreted as implicitly barring theconstant percentage approach, but none has directly

addressed the constant percentage versus constandollar debate. See generallyIn re Daou Systems IncSecurities Litigation, 411 F.3d 1006, 1026-1027 (9thCir. 2005) (We note that, as the [Third AmendedComplaint] currently reads, at the time when Daoubegan to reveal its true financial health in Augus1998, its stock was trading at $18.50 per share andnot at the class high of $34.37. The [Complaint] doesnot allege any revelation of Daous true financiahealth prior to August 1998. Thus, as the [Complaintreads now, any losses suffered between $34.37 and$18.50 cannot be causally related to Daous allegedlyfraudulent accounting methods because before therevelations began in August 1998, the true natureof Daous financial condition had not yet beendisclosed.); In re Redback Networks Inc., Securitie

Litigation , 2007 WL 963958, at *6 (N.D. Cal. March 302007) ([T]he stock price began to fall in June 2001when the truth began coming out. However, at thattime the stock already had fallen to less than $12 peshare. Based on Plaintiffs own allegations, the fal

from $150 to $12 cannot be attributed to the allegedfraud.) (emphasis in original).

9. The court found that the experts other damagesscenarios did not square with the law of loss causationapplied through the prism ofDauber tand Rule 702.Id. at 1294. Finding no evidence of loss causation odamages, the court granted defendants summaryjudgment. Id. at 1295 ([T]he summary judgmenrecord provides no basis upon which a finder ofact could reasonably determine, consistent withloss causation doctrine established by Dura andits progeny, that plaintiffs claimed losses werecaused by defendants alleged misrepresentationsor omissions and not by changed economiccircumstances, changed investor expectations, newindustry-specific or firm-specific facts, conditionsor other events) (citingDura,544 U.S. at 342-43).

10. Tabak, supra n.3, notes that the constantpercentage method may be adjusted to make theapproach compliant with Dura, but notably, the

suggested adjustments result in damages thaare always less than or equal to what would be foundunder the constant dollar method.

Reprinted with permission from the January 21, 2009 editionof the NEW YORK LAW JOURNAL 2009 Incisive USProperties, LLC. All rights reserved. Further duplication

without permission is prohibited. For information, contac877-257-3382 or reprintscustomerservice@incisivemediacom. # 070-01-09-38

WEDNESDAY, JANUARY 21, 2009