Frances Unger, William Patterson, Lead Gordon Ellis, Lead v. Amedisys Inc.

high, tоo had to reimburse REAVLEY, Before JONES and comply Medicare the difference. To with DENNIS, Judges. Circuit procedures, Amedisys pur- new PPS implemented computer chased and new JONES, Judge: EDITH H. Circuit software. case, to Fed. pursuant This on review willfully allege Amedisys Plaintiffs that 23(f), stan- implicates Rule Civ. PROC. manipulated program the PPS to inflate procedures dards and used district estimated costs for certain health ser- considering courts when certification of se- vices; thereby artificially fuelеd it dependent curities class actions on the and, ultimately, company earnings; Basic, theory. “fraud on the market” See Amedisys’s wrongfully enhanced actions Levinson, Inc. v. 485 U.S. 108 S.Ct. 13, 2001, Amedi- price. its stock On June (1988). 978, 99 L.Ed.2d 194 Like our statement, sys conceding a curative issued Third, Fourth, brethren Seventh revenues, that it had overstated but main- Circuits, and Ninth we hold that a careful were inad- taining that the overstatements inquiry required and find- certification vertently caused the new software used ings adequate must be mаde based on program. The stock with the PPS justify evidence to class certifi- admissible fell. cation. Because the district court errone- 21, 2001, August Unger filed ously applied proof too lax a standard of On Frances against Amedisys, alleging violations plaintiffs’ allega- fraud-on-the-market suit (5th 20(a) 10(b) garat, 378 F.3d of the Securi- of Sections Mann, v. Cir. and Rule 10b-5 U.S. Exchange Act of 1934 ties 1998). As is often the promulgated thereunder. case, lawyers potential solicited plaintiffs’ requires Rule 23 the claims of a over the Internet class members require class to meet several proposed advertisements. Sev- through newspaper can certified. ments before the class be consolidated with Un- eral other suits were party seeking certification bears the were chosen as ger’s ahd five individuals require of establishing burden that all certification was re- plaintiffs. lead Class ments of Rule 23 have been satisfied. persons “all and entities who quested for Berger Comрaq Computer Corp., Amedisys, common stock of purchased the First, F.3d through November Inc. between court must find what has been district Discovery 2001.” occurred [sic] June numerosity, commonality, typicali termed qualifications pro- to ascertain ty, representativeness.1 For class ac hearing, At a

posed representatives. class money seeking damages, tions like this the district court evaluated evidence one, the district court must make addi plaintiffs’ sketchy sup- аnd the evidence findings predominance supe tional port of the fraud-on-the-market basis for 23(b)(3). riority. predominance Rule The Amedisys’s reliance on presumed their requires finding element a that common misrepresentations. The district court “predominate issues of law or fact over 23(b)(3). certified the-class under Rule any questions only affecting individual requirement, members.” Id. al This timely Amedisys sought, defendants though commonality reminiscent of the granted, interlocutory court an this 23(a), requirement of Rule is “far more in appeal raising two issues embodied demanding” because it “tests whether adequacy of the class certification: proposed sufficiently classes are cohesive plaintiffs’ qualifications lead and the suffi- adjudication by to warrant representa ciency support plaintiffs’ of evidence to Prods., Windsor, tion.” Amchem Inc. v. presumption. fraud on the market 591, 623-24, 521 U.S. 117 S.Ct. (1997). 2249-50, Finally, 138 L.Ed.2d 689 DISCUSSION a superior class action must afford the The class cеrtification determina adju means to achieve “fair and efficient tion rests within the sound discretion of controversy.” dication of Rule Bernard, the'trial court. Oil Co. v. Gulf 23(b)(3). 89, 100, 2193, 2200, 452 U.S. 101 S.Ct. 68 (1981). discretion, L.Ed.2d 693 That how Recognizing important due ever, process must be exercised within the con concerns of plaintiffs both and de of A straints Rule 23. Id. district court fendants inherent in the certification deci premises legal analysis sion, that its Supreme requires on an erro Court district understanding governing neous rigorous analysis law courts to conduct a has prerequisites. abused its discretion. U.S. v. Insaul Rule 23 Gen’l Tel. Co. v. 23(a) class, (3) specific language 1. The of Rule is: common to the the claims or representative parties defenses of the are may One or members of a class more sue or typical of the claims or defenses of the representative parties be sued as on behalf class, (4) (1) only representative parties of all the class will if is so numerous joinder fаirly adequately protect impractica- of all members is the interests of ble, (2) questions there are of law or fact the class.

321 Falcon, 147, 161, clearly court was neither in 457 U.S. S.Ct. erroneous its (1982). District 72 L.Ed.2d 740 in factfindings legally. nor error To ad- required to take a “close look” argument courts are dress this point- further would in mak parties’ at the claims and evidence lessly ‍‌‌‌​‌‌​‌​‌​‌‌​‌​‌‌‌​​​​​‌​‌‌‌‌​​‌​​​​​‌‌​​​​‌​‌​‍require us to recount case-spe- Amchem, 521 ing its Rule 23 decision. cific evidence. at 2246. Class certi U.S. S.Ct. The crux of appeal legal this lies in the not hearings

fication should be mini-trials sufficiency basis for and of evidence sup class or individual on the merits porting finding pre the district court’s Jacquelin, claims. Eisen v. Carlisle & 23(b)(3). dominance under Rule The dis 156, 177-78, 2140, 2152-53, U.S. 94 S.Ct. trict court expressed skepticism here (1974). time, At the same 40 L.Ed.2d Castaño, which discussed fraud and other however, “[g]oing beyond pleadings by putative claims raised nationwide necessary, as a court must understand the smokers, govern class of tobаcco should claims, defenses, facts, appli relevant securities fraud class skepti actions. Its cable substantive law order to make cism was logi unfounded. Castaño is not meaningful determination of the certifica limited, cally so reasoning and its has been Co., tion v. Am. Tobacco issues.” Castano approved the securities fraud context 734, 744 To assist 84 F.3d other circuit courts as well as district may it sanction process the court Gariety courts this circuit. v. Grant disсovery at controlled the certification LLP, Thornton 368 F.3d Advisory stage. See Fed.R.Civ.P. Com Cir.2004); Lynch, Newton v. Merrill amendments. The mittee’s Note to 2003 Pierce, Smith, Fenner & plain requires text of Rule the court to (3d see also Johnston assume, “find,” merely not the facts favor Inc., Management., HBO Film 265 F.3d 23(b)(3). ing class certification. Rule (3d 178, 186-88 Bridge Szabo v. Appellants challenge quali first Mach., (7th Inc., port 249 F.3d 676-77 representatives fications of the class under Cir.2001); Techs., Inc., Lehocky v. Tidel 23(a)(4). require Rule To meet Rule 23 (S.D.Tex.2004); 220 F.R.D. Krog ments, rep court must that class find Sterritt, man v. 202 F.R.D. resentatives, counsel, their and the rela (N.D.Tex.2001); K Intelligent v. G Griffin tionship adequate between the two are to Inc., Sys., 196 F.R.D. protect of class mem the interests absent (S.D.Tex.2000). Corp., bers. v. Exxon 280 F.3d Stirman representa Class emphasized by One of the lessons satisfy they, that tives must the court Castaño and related cases is that a district counsel, directing litigation. not are analysis perform court must sufficient this, representatives To do class must dеtermine that class members’ fraud sufficiently show themselves informed predicated proving claims are not on indi litigation litiga manage about the vidual reliance. If the circumstances sur Berger, tion effort. 257 F.3d at 479. rounding plaintiffs alleged each reliance differ, Nothing representations in the indicates that the on fraudulent then record an that will have to be district court abused its discretion with reliance is issue 23(a)(4) by plaintiff, proposed and the regard requirement. proven to the Rule each requirement. 23(b)(3)’s fully predominance class Rule The district court evaluated the evi- fails ano, dence, depositions and 84 F.3d at see which included tes- Cast Pierce, Lynch, Fen timony representatives. of the class also Simon v. Merrill alia, Smith, Inc., security at issue is 482 F.2d inter ner & in an Id. at traded “efficient market.” Cir.1973). 248-49, In many at 992-93. S.Ct. invoking the fraud on the Only by cases, heavily-traded where or well known plaintiffs estab theory can these market suits, target stocks are the market effi рresumption rebuttable a classwide lish But ciency will not even be an issue. Amedisys’s alleged misrepres reliance on where, here, small-cap as the suit involves Basic, Inc., Supreme In entations.2 markets, a less-organized stocks traded in may presumed, that reliance be held Court demonstration of an efficient market is a proceed, actions to enabling 10b-5 class an prerequisite for certification.4 Without misrepresentation or a fraudulent “when efficiеncy, initial demonstration of market security of a impairs the value omission no there is assurance the available market.”3 As the in an efficient traded concerning information the stock material explained, Court an translates into effect on the market price supports presump a classwide theory is based The fraud on the tion of reliance. Absent an efficient mar that, open in an hypothesis on the ket, plaintiff individual reliance each mаrket, developed securities proven, proposed must be and the class is determined company’s stock predominance requirement. will fail the regarding material information available Castano, 84 F.3d at 745. Because this and its business .... Mis- Cf. inquiry prove can decisive for class certifi will leading statements therefore de- cation, because, given the realities of of stock even if the purchasers fraud costs, litigation compel certification can directly not purchasers rely do on the trial, settlements without ‍‌‌‌​‌‌​‌​‌​‌‌​‌​‌‌‌​​​​​‌​‌‌‌‌​​‌​​​​​‌‌​​​​‌​‌​‍courts have fre .... The causal connec- misstatements quently applied rigorous, though prelimi tion the defendants’ fraud and between nary, proof standards of to the market purchase of plaintiffs’ stock such See, efficiency e.g., determination. Garie significant no less than in a case case is 368-70; Newton, ty, 368 F.3d at 259 F.3d misrepresentations. of direct reliance on 167-69; Szabo, 675-77; at 249 F.3d at Basic, 241-42, U.S. 108 S.Ct. 1059, 1064-65 Gillespie, Binder v. (internal omitted) (empha- at 989 citations (9th Cir.1999); Seagate In re Tech. II Sec. added). sis 1341, Litig., F.Supp. 843 1354-55 support (N.D.Cal.1994); To pre 473; this rebuttable Krogman, 202 F.R.D. at Grif a sumption, plaintiff prove, fin, securities must 196 F.R.D. at 303-05. have Courts claim, prevail plaintiff 2. a To on 10b-5 a must Markets” out the Fraud on the Efficient (1) prove misrepresentation a material or MarlcetDoctrine the Private Lit- Securities after defendant, (2) omission the scienter on the Act, igation 97 Nw. U.L.Rev. 995 Reform defendant, reliance, (3) (4) part of the due (2003). The author contends that "what de- diligence by plaintiff pursue his or her justified termines whether investors were good own with care interest faith. Ste relying integrity price on the of the market is Curtis, phenson v. Paine Webber Jackson & efficiency not of the relevant market but Inc., 1095, F.2d 1098 839 rather whether a misstatement distorted the security.” of the affected 97 Nw. L.Rev. Newton, 3. at 259 F.3d 175. persuasively argued, at 1035. The article is 4. A recent law review article Supreme criticizes job but it is the Court’s to overrule theory adopted Basic, efficient market in Basic as outright in the absence of conflict with step analysis out of with current economic Act, Litigation the Private Securities Reform and inconsistent with the thrust of recent 104-67, (1995). Pub.L. No. 109 Stat. 737 Oldham, legislation. Jeffrey Taking See L.

323 degree proof required to the 202 F.R.D. at 477-78 (using likened the the last three injunction factors). in preliminary used standards gauging These tools for market 366, or hearings, Gariety, see 368 F.3d at efficiency have by many been used courts 12(b)(1) in Fed. Rule Civ. Proo. and throughout country and within this 12(b)(2) Szabo, contests, jurisdiction 249 See, circuit. e.g., Gariety, 368 at F.3d Although F.3d at 676. the court’s determi Binder, 1064-65; 184 Hayes F.3d at purposes may nation for class certification Gross, (3d Cir.1992); (or rejected) wholly be revised the ulti Horwath, Freeman v. Laventhol & factfinder, may simply mate the court not F.2d 198-99 Lehocky, facts in favor of an efficient presume the 220 F.R.D. at 505-09. market. Although represent this does not an ex Courts have relied on several fac list, haustive and in some cases one of the a tors to dеtermine whether stock traded may above factors unnecessary, be once a (1) average in an “efficient market”: factors, court apply endeavors to these weekly trading expressed per volume as a they weighed must be analytically, not (2) shares; centage outstanding of total counted, merely repre as each of them analysts following the number of securities sents ‍‌‌‌​‌‌​‌​‌​‌‌​‌​‌‌‌​​​​​‌​‌‌‌‌​​‌​​​​​‌‌​​​​‌​‌​‍a distinct facet of efficiency. market (3) stock; reporting on the the extent Some courts have concluded that there arbitrageurs market makers and which not an efficient as a market matter of law (4) stock; eligi company’s trade for stocks trading the over-the-counter bility registration to file SEC Form S-3 market. See In re Data Sys. Access Sec. (as (5) S-2);5 opposed to Form or S-l Litig., (D.N.J.1984), 103 F.R.D. empirical “showing facts a existence grounds rev’d on other by 843 F.2d 1537 un relationship cause and effect between (3d Cir.1988); Epstein v. Am. Reserve expectеd corporate events or financial re Corp., No. 79 C 1988 WL 40500 response an immediate leases and (N.D.Ill. Apr.21, 1988). go need not so We (6) price”; company’s stock here, holdings far but such (7) are indicative capitalization; spread the bid-ask for gulf (8) type wide between the of mar sales; float, stock the stock’s trad ket for stocks that trade millions of shares ing counting volume without insider-owned Basic, Bloom, daily, e.g., 243-44, 485 U.S. F.Supp. stock. See Cammer v. (D.N.J.1989) (using and S.Ct. at and the much less aсtive dis factors); cussing Krogman, Amedisys.6 the first five market for stocks like determination, companies may 5. Form S-3 is reserved for whose which often benefit statistical, economic, actively widely stock is traded and followed. from and mathematical See, S-3, analysis. e.g., Bell v. Ascendant Solu To file Form must have tions, -CV-0166-N, No. Civ. A. reports filed SEC for twelve consecutive (N.D.Tex. 2004); July WL possess seventy-five Le months and a million dol- hocky, Krogman, 220 F.R.D. at capitalization lar market level. See 17 C.F.R. сontrast, Although F.R.D. at 467. courts are not to By § 239.13. there is no minimum upon experts” a insist “battle of the at the capitalization requirement to file either Form Further, stage, Complex see for certification Manual S-l or S-2. a need not 21.21, ed.2004) Litigation § one court requirements reporting spelled even meet the explained that § out in 239.13 to file a Form S-l. See 17 cases, §§ C.F.R. 239.11-239.12. many In it makes sense to consider admissibility testimony the of the of an ex- requirement expert pert profferеd 6. There is no for testimo to establish one of the Rule ny efficiency, many on the issue of market but 23 elements in the context of a motion to addressing prior considering courts have considered it when strike to class certifica- of a Amedisys weighed court for stock favor Unfortunately, the district finding efficiency. support insufficient attention of market To devoted in this case conclusion, efficiency single factors. a market the court relied on evaluating the to that, during the printout, coupled Internet with affidavits determination The court’s in. stock traded that were admitted question,-Amedisys plaintiffs’ witnesses time market, predicated on its was opportunity an efficient without for cross-examination. high trading stock Moreover, factors: finding acknowledge of three the court failed to stock, volume, trading makers growing market concern that the mere number of relationship be makers, and a cause-and-effect analysis, without further move price events corporate tween efficiency. has little to do with market (not- See, ment. e.g., Krogman, 202 F.R.D. at 476 ing that the “number of market makers” trading sug- volume weekly A stock high an practice proven factor has unreliable active, in- informed presence gests efficiency measure of market unless tied to evaluating trading the stock In vestors. price); Griffin, trade volume and however, the district court never volume, 304; Serfaty F.R.D. at v. Int’l Automated plaintiffs never asсertained —and (D.Utah 418, 422 Sys., 180 F.R.D. Amedisys number of proved actual —the 1998); Appel O’Neil v. 165 F.R.D. regularly Accepting traded. being shares (W.D.Mich.1996)(“The economic literature analyt- naked claim as to this plaintiffs’ upon has criticized reliance the number of yield point cannot a reliable starting ical market makers as an indicator of efficien- “found” that court first result. The Barber, al., cy.”); see M. also Brad et trading 3.9% weekly volume was average Theory Fraud-on-the-Market and Indica- shares, but then con- outstanding Efficiency, tors Common Stock’s J. could be cut half. figure ceded that the (1994). CoRP. L. The district court appears court to have based Because ques- erred when it did not consider the only printouts its on two determination finding. tionable relevance of this Internet, not deter- from the the court did mathematically average correct mine the The district court also found causal weekly trading volume. As commentators Amedisys corporate connection between observe, however, trade volume can be and the movement of the events stock exchanges on some grossly exaggerated price, not take into account but did double-counting, through sometimes many other factors that could affect the Barclay & F. Tor- fifty percent. over M. Amedisys price of stock. The court cor- chio, Comparison Trading A Models identified, rectly causal connection as Damages Calculating Aggregate Used for important one of the most market-efficien- Litigation, 64 Law & Con- in Securities cy It goes factors. the heart of the (Summer 2001). temp. At 105, 106 Probs. theory: “fraud on the In an market” effi- stage, reliance on unverifi- certification market, nearly cient where information is hardly relying able is better than evidence perfect, material misstatements alter a allegations. on bare immediately. stock’s almost In such circumstances, easy injury “it The district court also found see how twenty-two person “market makers” can who is unaware of the presence of befall testimony supporting mo- class is tion. In order to consider Plaintiffs' certification appro- reliable. tion for class certification with the Bell, (citations priate scrutiny, the Court must 2004 WL at *3-*4 amount omitted). expert first determine whether Plaintiffs' of, utility v. Balcor Film In rather scant example, deceit.” See Eckstein for vestors, Cir. Further, number of “market makers.” 1993). Demonstrating that market reac court not did address the effect on the by company press caused tions are releas efficiency market determination of Amedi- not, however, in es should ‍‌‌‌​‌‌​‌​‌​‌‌​‌​‌‌‌​​​​​‌​‌‌‌‌​​‌​​​​​‌‌​​​​‌​‌​‍be an exercise sys’s ineligibility to file an Form SEC S-3 hoc, Many post propter logic. hoc varia (the at in question the time other factor potential to do bles have the affect case).7 in absent this Because Rule 23 price daily average; stock market na complete mandаtes a analysis of “fraud on —the tional, industry-specific local and economic indicators, the market” district courts news; activities; competitors’ and on and must and weigh factors both for address volatility on. The overall of the stock against efficiency. market price speed of its reaction to com See,

pany may significant. news also be CONCLUSION e.g., Krogman, 202 F.R.D. at 477-78. To Although we owe considerable deference end, testimony may expert helpful this be tо district courts in reviewing certification utility because of of statistical event decisions, we cannot affirm the order itas analysis for n. inquiry. supra, See 6. is presently supported. After a more recognizing complexity Instead of however, thorough inquiry, certification factor, court this cause-and-effect re- may ultimately prove correct. When a showing lied on a that on March 1 and court considers class certification based on 1, 2001, May the stock rose following the theory, fraud on the market it must positive announcements issued Amedi- engage in thorough analysis, weigh the sys days, price dropped on those and the factors, require parties relevant both day announced that its justify allegations, ruling their and base its earnings restated. This would be evidence Questions on evidence. admissible of mar- worthwhile, alone, standing is no doubt but efficiency differently ket cannot be treated determine, insufficiently it probative from preliminary other certification issues. facts,” Cammer, on “empirical based see Courts cannot make an informed decision at that a F.Supp. causal connec- allegations, based on bare one-sided affida- short, tion In incorrectly exists. the court vits, unexplained printouts. Internet all in used three factors it found favor of reasons, foregoing For the the class cer- efficiency market as a checklist rather tification order is VACATED and RE- analytical than an tool. for proceedings MANDED further consis- Similarly, the court failed to evaluate the tent herewith. significance market-efficiency fac- VACATED AND REMANDED. lacking tors the instant case. For in- stance, analysts the number of securities DENNIS, Judge, specially Circuit following important the stock is an factor. concurring: See, e.g., Krogman, 202 F.R.D. at Cammer, outcome, Hence, I I Although concur F.Supp. 1286-87. analyst disagree majority opinion’s the fact that with the state- reporting no was on “[cjourts Amedisys question stock at the time ment that have likened the de- weighed against gree determining should have proof required been [in stock, Amedisys's 7. The court also failed to refer to its and the float. capitalization, spread the bid-ask 326 only in “standards” that have ever the standards used efficiency] to in hearings required ... and class certifications are injunction been

preliminary look,” 12(b)(2) e.g., “close jurisdictional open contests.”1 Con- more textured: Products, Windsor, reading, Fourth mаjority’s Amchem Inc. v. 521 trary to the Gariety 591, 615, 2231, in v. Grant 117 138 L.Ed.2d opinion U.S. S.Ct. Circuit’s (4th Thornton, LLP, 356, (1997); Falcon, “rigorous analysis,” 368 F.3d 366 689 Cir.2004), 161, opin- Spence and the Seventh Circuit’s 457 at v. U.S. 102 S.Ct. Machines, Inc., (5th Glock, Ges.m.b.H, Bridgeport v. 227 ion in Szabo F.3d 308 (7th 672, Cir.2001), Co., do not 676 v. American Tobacco 249 F.3d tano Cas Cir.1996). 734, compare degrees standards or 740 liken or 84 F.3d On hand, In- proceedings other at all. in proof Supreme from other Court Eisen v.

stead, the Fourth and Seventh Circuits Jacquelin, 417 U.S. 177- Carlisle (1974), mod- inquiries referred to those as simply 94 S.Ct. 40 L.Ed.2d 732 of how courts can analogs likely or district that a “more than els admonished resolving in pleadings behind the in a “probe prevail” inappropriate standard is fact, certifications”2 without diso- analysis. class action certification In we Rule 23 Supreme beying recently Court’s admonish- a held that court must conduct an ... against “expanding in Eisen investigation” ment “intense factual while at the analysis to include consider- “takfing] inquire certification same time care to into proposed аtion of whether the class is like- substance and structure of the under ly prevail ultimately on the merits.”3 passing claims without lying on their mer example, Gariety simply says: For its.” Robinson Texas Auto. Dealers Ass’n, 387 F.3d 416 process A model for [the certification] can be observed the context of the Thus, majori- I although agree with the injunction preliminary practice. Courts ty’s holding the district court did not findings determining factual make weigh the for and adequately factors preliminary injunction whether should against finding efficiency, of market I issue, findings but those do not bind the majority’s strongly disagree with the read- ..., jury jury’s ‍‌‌‌​‌‌​‌​‌​‌‌​‌​‌‌‌​​​​​‌​‌‌‌‌​​‌​​​​​‌‌​​​​‌​‌​‍findings on the ing Gariety Contrary to the Szabo. govern judgment merits to be en- majority’s impression, these cases do not tered in the case.4 support suggest adoption applica- or or And vein Szabo the same observes degrees proof tion of or standards of “[cjourts inquiries routinely make similar efficient market determinations for the deciding ... pos- before whether [courts] purposes of class certification. jurisdiction subject

sess over the matter of persons

the case and the of the defen-

dants, venue, proper the location of the conveniens,

application of forum non preliminary

other issues.”5 (citing 4. Op. Pg. Id. at 1. 322-23. Univ. Texas v. Camen 390, 395, isch, U.S. S.Ct. (quoting

2. 368 F.3d at 366 Gariety, General (1981)). L.Ed.2d Falcon, Telephone Southwest v. Co. of 147, 160, U.S. 102 S.Ct. 72 L.Ed.2d 740 Szabo, 5. 249 F.3d 676. (1982)). Co., (citing

3. Id. Castano v. American Tobacco (5th Cir.1996)).

Case Details

Case Name: Frances Unger, William Patterson, Lead Gordon Ellis, Lead v. Amedisys Inc.

Court Name: Court of Appeals for the Fifth Circuit

Date Published: Feb 17, 2005

Citation: 401 F.3d 316

Docket Number: 03-30965

Court Abbreviation: 5th Cir.