Full text
UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF FLORIDA
CASE NO. 21-MD-2989-ALTONAGA/DAMIAN
In re:
JANUARY 2021 SHORT SQUEEZE
TRADING LITIGATION
_____________________________________/
This Document Relates to the Federal Securities Tranche
DEFENDANTS ROBINHOOD MARKETS, INC., ROBINHOOD FINANCIAL LLC AND
ROBINHOOD SECURITIES, LLC’S REPLY IN FURTHER SUPPORT OF THEIR
MOTION TO EXCLUDE THE OPINIONS OF DR. ADAM WERNER ON MARKET
EFFICIENCY
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 1 of 16
TABLE OF CONTENTS
TABLE OF AUTHORITIES .......................................................................................................... ii
PRELIMINARY STATEMENT .....................................................................................................1
ARGUMENT ...................................................................................................................................1
I.
Dr Werner’s Market Efficiency Opinions Are Inadmissible. ..................................1
II.
Dr. Werner’s Opinions Concerning Efficiency in His Opening Report Are
Unhelpful and Unreliable. ........................................................................................4
A.
Dr. Werner’s Analysis of the Year Preceding the Class Period Is
Unhelpful. ....................................................................................................4
B.
Dr. Werner’s Opening Report Ignores Value-Relevance and
Thereby Applies the Wrong Standard for Efficiency in His
News/No News Test. ...................................................................................6
C.
Dr. Werner Deliberately Departed from His Methodology in
Previous Cases, Apparently To Achieve a Desired Result. .........................7
III.
Dr. Werner’s Opinions Regarding Efficiency in His Rebuttal Report Are
Also Unhelpful and Unreliable. ...............................................................................8
IV.
Dr. Werner’s New Declaration Must Be Stricken. ..................................................9
CONCLUSION ..............................................................................................................................10
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 2 of 16
ii
TABLE OF AUTHORITIES
Page(s)
Cases
Basic Inc. v. Levinson,
485 U.S. 224 (1988) ...................................................................................................................4
Brokop v. Farmland Partners, Inc.,
18-cv-2104, 2021 WL 4916240 (D. Colo. July 27, 2023) .....................................................2, 3
Cohen v. State Farm Ins. Co.,
No. 3:07-CV-251, 2007 WL 4437223 (M.D. Pa. Dec. 14, 2007)..............................................3
Comcast Corp. v. Behrend,
569 U.S. 27 (2013) .....................................................................................................................4
Companhia Energetica Potiguar v. Caterpillar Inc.,
No. 14-CV-24277, 2016 WL 3102225 (S.D. Fla. June 2, 2016) .............................................10
Dyett v. N. Broward Hosp. Dist.,
No. 03-cv-60804, 2004 WL 5320630 (S.D. Fla. Jan. 21, 2004) ..............................................10
FindWhat Investor Grp. v. FindWhat.com,
658 F.3d 1282 (11th Cir. 2011) .................................................................................................2
George v. China Auto. Sys., Inc.,
No. 11-cv-7533, 2013 WL 3357170 (S.D.N.Y. July 3, 2013) ...............................................2, 3
Goodman-Gable-Gould Co. v. Tiara Condo. Ass’n, Inc.,
595 F.3d 1203 (11th Cir. 2010) .................................................................................................4
In re Alstom SA Sec. Litig.,
253 F.R.D. 266 (S.D.N.Y. 2008) ...............................................................................................8
In re PolyMedica Corp. Sec. Litig.,
453 F. Supp. 2d 260 (D. Mass. 2006) ........................................................................................9
Johnson v. Comm’r of Internal Revenue,
673 F.2d 262 (9th Cir. 1982) .....................................................................................................5
K & H Dev. Grp., Inc. v. Howard,
255 F.R.D. 562 (N.D. Fla. 2009) .............................................................................................10
Mitchell v. Ford Motor Co.,
318 F. App’x 821 (11th Cir. 2009) ..........................................................................................10
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 3 of 16
iii
Moore v. GNC, Holdings, Inc.,
No. 12-cv-61703, 2014 WL 12684287 (S.D. Fla. Jan. 24, 2014) ............................................10
Monroe Cnty. Emps.’ Ret. Sys. v. S. Co.,
332 F.R.D. 370 (N.D. Ga. 2019) ................................................................................................3
Simmons v. Ford Motor Co.,
No. 18-cv-81558, 2021 WL 6062633 (S.D. Fla. Dec. 22, 2021) .............................................10
Wallis Tractor Co. v. Comm’r,
3 B.T.A. 981 (1926) ...................................................................................................................5
Statutes & Rules
Fed. R. Civ. P. 26 ...........................................................................................................................10
Fed. R. Civ. P. 37 ...........................................................................................................................10
Other Authorities
Fischel, Daniel R., Efficient Capital Markets, the Crash, and the Fraud on the
Market Theory,
74 Cornell L. Rev. 907 ...............................................................................................................7
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 4 of 16
PRELIMINARY STATEMENT
In their Daubert opposition, Plaintiffs concede for the first time a fact they have
been dancing around for months: the markets for the Affected Stocks were inefficient during the
Class Period. (Opp. at 1.) That fact alone defeats class certification. It also highlights the
emptiness of their Daubert opposition: Dr. Werner analyzed market efficiency only for periods
before the Class Period, analyses which do not help Plaintiffs meet their burden for class
certification. Dr. Werner’s efficiency opinions are therefore unhelpful and inadmissible.
Dr. Werner also made serious methodological errors in his analysis, including his
failures to exclude articles about stock-price movements and trading volume in his “news/no
news” test (which renders the test circular) and to compare news days to non-news days (which
means he is not testing for cause and effect). Plaintiffs’ attempts to justify Dr. Werner’s
methodology fall short. Doubtless realizing that, Plaintiffs attempt to remedy those errors by
introducing a new, never-before-disclosed declaration from Dr. Werner. Not only is that
declaration impermissible, but it also fails to address the very defects it purports to rectify.1 The
Court should exclude Dr. Werner’s opinions regarding efficiency in their entirety.
ARGUMENT
I.
DR WERNER’S MARKET EFFICIENCY OPINIONS ARE INADMISSIBLE.
As a threshold matter, Plaintiffs argue that Robinhood’s motion is moot because,
they claim, it challenges only Dr. Werner’s analyses of Cammer factor 5. (Opp. at 4.) Plaintiffs
assert that the Court can find efficiency without evidence of a causal relationship between news
and price movements (i.e., without Cammer factor 5 being satisfied),2 and so should not consider
the Daubert motion. (Id. at 4-5.) They are wrong on both counts. Dr. Werner’s misconception
of efficiency infects not only Cammer factor 5, but also conclusions on efficiency as a whole.
And on the facts of this case, Plaintiffs cannot establish efficiency by relying only on the
structural prerequisites to efficiency (i.e., the other Cammer factors). Cammer factor 5 is thus
critical to efficiency and class certification. For each of these reasons, the admissibility of
Dr. Werner’s efficiency opinions lies at the heart of whether class certification is appropriate, the
1 Robinhood has provided a sur-rebuttal declaration by Prof. Grenadier to respond to
Dr. Werner’s claims in his impermissible sur-rebuttal declaration. (Reply Ex. 1.) If the Court
does not consider Dr. Werner’s declaration, it need not consider Prof. Grenadier’s declaration.
2 As Robinhood explained in its Daubert motion, Cammer factor 5 asks whether a
company’s stock prices react to news about the value of the company. (Mot. at 4-5.)
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 5 of 16
2
Daubert motion is not moot, and the Court should exclude his opinions.
First, Dr. Werner’s efficiency opinions should be excluded in toto because his
fundamentally misguided understanding of efficiency infects his entire report. In Dr. Werner’s
view, not only is value irrelevant to efficiency (Ex. 1, Werner Dep. 160:14-161:11), but dramatic
price movements in the absence of news are consistent with efficiency (id. 84:12-86:10). As
explained in Robinhood’s Daubert motion (Mot. at 10-15, 18-20), its opposition to class
certification (Class Cert. Opp. at 18-21, 28), and section III, infra, that view is squarely at odds
with the understanding of efficiency that economists use in their academic work and that courts
have adopted in binding case law. See FindWhat Investor Grp. v. FindWhat.com, 658 F.3d 1282,
1310 (11th Cir. 2011) (in an efficient market, lack of new information “will not cause a change
in the stock price”). That methodological error is not limited to Cammer factor 5. Dr. Werner
testified that his overall opinion that the markets were efficient prior to the Class Period did not
apply the principle that prices should move in response to value-relevant news, and remain
relatively stable in its absence. (Ex. 1, Werner Dep. 82:4-86:10, 160:14-161:11.) His opinions
on market efficiency are based in their entirety on this misunderstanding.
Second, Plaintiffs are also wrong in asserting that they can ignore Cammer
factor 5 and rely on structural factors alone. There is overwhelming direct evidence that the
unprecedented price movements of the Affected Stocks bore no relation to news immediately
before and during the Class Period. Because Robinhood has proffered direct, affirmative
evidence of a temporary deviation from efficiency, Plaintiffs cannot establish efficiency by
relying on the structural prerequisites to efficiency alone. See Brokop v. Farmland Partners,
Inc., 18-cv-2104, 2021 WL 4916240, at *18 (D. Colo. July 23, 2021) (refusing to find efficiency
for the period for which there was no direct evidence of efficiency); George v. China Auto. Sys.,
Inc., No. 11-cv-7533, 2013 WL 3357170, at *9 (S.D.N.Y. July 3, 2013) (finding the market
inefficient where only Cammer factor 5 was disputed).
This case involves a temporary deviation from the normal state of affairs for well-
capitalized stocks with high levels of trading volume on a U.S. public exchange and robust
analyst coverage.3 While it was therefore true that the structural conditions for efficiency were
in place, Robinhood’s experts have demonstrated that, for the few weeks before the Class Period,
3 Seven of the Affected Stocks (AMC, BB, BBBY, EXPR, GME, NOK and TRVG) may
have fallen into this category in normal times before the “meme stock” events of January 2021.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 6 of 16
3
the price movements of the Affected Stocks became temporarily disjointed from any value-
relevant news. That is, Robinhood has presented evidence that in January 2021, the prices of the
Affected Stocks moved dramatically in the absence of value-relevant news. (Ex. 5, Grenadier
Report ¶¶ 97-105; Reply Ex. 3, Fischel Report ¶ 36.) That disconnect is affirmative evidence of
the absence of a causal relationship between news and prices. As Dr. Werner himself has
explained, that causal relationship—entirely absent here—is “the essence of market efficiency.”
(Reply Ex. 7, FXCM Report ¶ 106; see also Ex. 1, Werner Dep. 141:13-142:3 (a cause-and-
effect relationship is the only direct test of market efficiency).)
The evidence that this “essential” causal relationship was absent in the few short
weeks before the Class Period makes this case entirely unlike the cases cited by Plaintiffs. In
those cases, the defendants argued that, even though the structural conditions for efficiency were
present, the plaintiffs had not shown that the markets were efficient as a general matter. See,
e.g., Monroe Cnty. Emps.’ Ret. Sys. v. S. Co., 332 F.R.D. 370, 384-88 (N.D. Ga. 2019). The
courts in those cases found that the conflicting evidence the parties had proffered as to Cammer
factor 5 did not suffice to rebut the conclusion that, since the structural conditions for efficiency
were in place, the markets were probably efficient as a general matter—even though there was
conflicting direct evidence of efficiency. See, e.g., id.; (Ex. 1, Werner Dep. 141:13-142:3).
Robinhood does not dispute that because the structural conditions were in place,
the markets for many of the Affected Stocks may be efficient in normal times. (See supra at 2
n.3.) However, there was a temporary deviation from efficiency in the weeks before the Class
Period due to the extraordinary social media coordination and attempted short squeezes. None of
the cases Plaintiffs cite holds that a temporary deviation from efficiency cannot arise despite the
presence of structural factors. Indeed, Plaintiffs concede this can happen, arguing themselves
that the markets were temporarily inefficient without any change in the underlying structural
factors during the Class Period. (Opp. at 1); see Brokop, 2021 WL 4916240, at *18. Because
Robinhood has proffered direct evidence of temporary inefficiency, the Court cannot conclude
that the markets were efficient at the relevant time without direct rebuttal evidence from
Plaintiffs of a causal relationship between news and price movements. Cohen v. State Farm Ins.
Co., No. 3:07-CV-251, 2007 WL 4437223, at *2 (M.D. Pa. Dec. 14, 2007). Yet “Plaintiff[s],
despite bearing the burden of establishing the prerequisites for class certification, ha[ve]
presented no [such] contrary evidence.” Id.; see also George, 2013 WL 3357170, at *13.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 7 of 16
4
Accordingly, Robinhood’s Daubert motion strikes at the core of the class certification decision.
II.
DR. WERNER’S OPINIONS CONCERNING EFFICIENCY IN HIS OPENING
REPORT ARE UNHELPFUL AND UNRELIABLE.
A.
Dr. Werner’s Analysis of the Year Preceding the Class Period Is Unhelpful.
Plaintiffs concede that Dr. Werner looked only at the year preceding the class
period in his opening report, but fail to identify any issue at class certification as to which
efficiency for the full year before the Class Period is relevant. (See Opp. at 5.) Instead, they
assert that their “theory of the case is that Robinhood’s manipulation rendered the markets
inefficient during the Class Period.” (Id. at 1.) But that does nothing to explain how efficiency
before the Class Period helps Plaintiffs establish reliance or damages on a class-wide basis.
First, efficiency before the Class Period cannot establish class-wide reliance
under Basic Inc. v. Levinson, 485 U.S. 224 (1988). As Plaintiffs concede, reliance at the time of
purchase is not an element of their claim: only reliance at the time of sale, i.e., during the Class
Period, is. (Mot. to Certify at 4, 15.) Moreover, neither Dr. Werner nor Plaintiffs contend that
Dr. Werner’s findings of efficiency can be extrapolated to the Class Period. (Ex. 8, Werner
Report ¶ 1 n.1; Opp. at 1.) Efficiency before the Class Period, therefore, has no bearing on
reliance in this case, and does not establish that reliance can be proven class-wide.4
Second, efficiency in the year before the Class Period cannot help Plaintiffs
calculate damages class-wide, as they must under Comcast Corp. v. Behrend, 569 U.S. 27
(2013). The only supposed class-wide damages methodology Plaintiffs have proposed—and the
only one on which they may rely, see Goodman-Gable-Gould Co. v. Tiara Condo. Ass’n, Inc.,
595 F.3d 1203, 1211 (11th Cir. 2010)—is to use the January 27, 2021 closing price as the but-for
price. (Opp. at 17.) This is not a viable methodology (see Class Cert. Opp. at 26-29), but in any
event, efficiency for the full year before the Class Period has no bearing on that methodology.
The January 27, 2021, closing price cannot reliably serve as a but-for price unless
it resulted from an efficient market. (See Class Cert. Opp. at 26-29.) If it did not, there is no
reason to believe the price would have remained at the January 27, 2021, level over the Class
4 By contrast, Robinhood’s experts analyze the weeks before the Class Period because
Robinhood does contend that inefficiency can be extrapolated into the Class Period. Because
these markets were inefficient right before Robinhood’s actions, they would have remained
inefficient during the short Class Period regardless of those actions. Robinhood has also
presented direct evidence that the markets were inefficient during the Class Period.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 8 of 16
5
Period in the but-for world without trading restrictions. Plaintiffs concede this point in their
motion for class certification (Mot. to Certify at 25 (citing Profs. Fama’s and Fischel’s
statements that price is the best estimate of value in efficient markets)) and again in their Daubert
opposition, citing a case that used the market price in an efficient market as a proxy for value.
(Opp. at 18 (citing Johnson v. Comm’r of Internal Revenue, 673 F.2d 262 (9th Cir. 1982)).)
Notably, Johnson distinguished cases that did not use market price as a proxy for value because
the markets in those cases were inefficient. Johnson, 673 F.2d at 265 (citing Wallis Tractor Co.
v. Comm’r, 3 B.T.A. 981, 1003 (1926)). Thus, to satisfy Comcast, Plaintiffs must show that the
January 27, 2021, price resulted from an efficient market.
Dr. Werner’s analysis does nothing to establish that the closing price on January
27 was the result of an efficient market. He monolithically analyzed an entire year, ignoring the
demonstrably unusual weeks at the end of that year. (See Mot. at 8-10.) Plaintiffs argue that this
was reasonable because—they claim—those weeks were no different from the rest of the
preceding year. (Opp. at 6.) That assertion is remarkable. As the Court is well aware, the weeks
immediately preceding the Class Period reflected extreme price volatility and trading volume as
the social-media driven fad drove the stock market nearly to a breaking point. (MTD Order at 1,
5-6.) Dr. Werner concedes that January 27, 2021, was anomalous by excluding that date from
his regression. As Plaintiffs note, because “the purpose of a regression model is to estimate
typical stock price behavior,” Dr. Werner “excluded [January 27, 2021] to avoid biasing the
results of his regression analysis by removing potentially abnormal price movements.” (Opp. at
7 (emphases added).) He excluded January 27, 2021, precisely because it was a clear outlier.5
Moreover, Robinhood proffered direct evidence that the Affected Stocks behaved
aberrantly immediately before the Class Period. For example, Prof. Grenadier showed extreme
price movements and a structural break in the Affected Stocks’ behavior relative to the market in
January 2021. (Ex. 5, Grenadier Report ¶¶ 71-91 & n.140 (finding that the January 21-February
4, 2021, period differed significantly from the last six months of 2020 in terms both of volatility
and of the relationship between individual stock returns and market or industry returns).)
Plaintiffs ignore this evidence altogether, and instead proffer—in an untimely sur-
5 Plaintiffs’ defense of Dr. Werner’s use of a dummy variable (Opp. at 6-7) misses the point.
Robinhood does not claim that excluding that date from the regression was inappropriate—to the
contrary, it was appropriate, because the price movements on that date were abnormal.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 9 of 16
6
rebuttal report (see infra Part IV)—an irrelevant test they claim establishes that conditions were
constant over time: the Breusch-Pagan test of heteroskedasticity. (Opp. at 6; id. Ex. 3 ¶ 16.)
Plaintiffs are correct in asserting that “there are scientific tests to determine whether there are in
fact substantial variations within the one-year period.” (Opp. at 6.) The problem for Plaintiffs is
that even in his improper sur-rebuttal report, Dr. Werner does not actually conduct any such test.
As Dr. Werner explained in his opening report, his regression “determine[d] how
the Affected Companies’ stock prices [the dependent variable] typically behaved in relation to
the overall stock market and industry-specific factors [the independent variables].” (Ex. 8,
Werner Report ¶ 56.) Once the regression has estimated that typical relationship, it generates an
expected return for the Affected Stock based on the overall market and industry-specific return
on a given day. (Id. ¶¶ 56-57.) That expected return is then compared to the actual return for
that day. (Id.) The difference between the expected and actual returns is the “residual.” (Id.)
In his sur-rebuttal declaration, Dr. Werner purports to establish that these
residuals did not vary materially over time in the year preceding the class period. (Opp. Ex. 3
¶ 15.) But his analysis does not actually answer that question. Instead, he analyzes whether the
variance of the residuals depends on the size of the market and industry return. Thus,
Dr. Werner demonstrates that when the industry return is (say) 0.5%, his regression’s prediction
of a given stock’s return is just as close to the actual return as when the industry return is
(say) -8%. This demonstrates only that his regression is equally good no matter the industry and
market returns. It says nothing about whether those residuals during the few weeks before the
class period were dramatically different—and dramatically more volatile—than during the 11
months preceding the extraordinary trading frenzy. (Reply Ex. 1, Grenadier Decl. ¶¶ 7-10.) In
contrast, Prof. Grenadier has actually tested the behavior of the Affected Stocks over time (which
is what Plaintiffs concede is relevant), and those tests confirm the (obvious) fact that the
residuals were dramatically more variable starting the week before the Class Period—and that,
moreover, the relationship between individual-stock and market-wide or industry-specific returns
was measurably different—than they were during the last six months of 2020, before the trading
frenzy. Because the weeks before the Class Period were measurably different from the rest of
the preceding year, Dr. Werner’s analysis of that year inadmissibly commingles data sets.
B.
Dr. Werner’s Opening Report Ignores Value-Relevance and Thereby Applies
the Wrong Standard for Efficiency in His News/No News Test.
As explained in Robinhood’s Daubert motion and opposition to class
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 10 of 16
7
certification, efficiency requires that prices move in response to value-relevant news. (Mot. at
10-12; Class Cert. Opp. at 18 n.16.) Yet Dr. Werner entirely ignored value-relevance in
selecting news days (Ex. 1, Werner Dep. 154:21-157:23, 160:14-161:11), and thus swept in
articles about unusual stock-price movements and trading volumes—and even articles that
expressly noted the absence of value-relevant news.6 Plaintiffs defend that approach by arguing
that Dr. Werner’s definition of efficiency is appropriate. (Opp. at 7-13.) They are wrong.
Rather than testing only whether stock prices move in response to news relevant
to cash flows (a cause-and-effect relationship)—as prices should in an efficient market—
Dr. Werner tests whether news is published about price movements or trading volume the same
day those price movements and trading volume occur (which does not help determine whether
new (let alone value-relevant) information caused the price movements at all). While Plaintiffs
suggest that this error goes only to weight, and not admissibility, they are wrong. Dr. Werner is
asking the wrong question, and his answer is therefore irrelevant to efficiency. (Mot. at 12-16.)
Plaintiffs also suggest that Robinhood sets the bar too high by requiring
Dr. Werner to show “fundamental” efficiency. (Opp. 7-9.) But that is not the case. Plaintiffs
conflate the requirement that price movements be in response to value-relevant news
(informational efficiency) with a requirement that prices always reflect the present value of
future cash flows (fundamental efficiency)—despite quoting the very language highlighting that
distinction. (Id. at 8 (quoting Fischel, Efficient Capital Markets).)7 Because he ignores whether
news is value-relevant, Dr. Werner has not analyzed even informational efficiency.
C.
Dr. Werner Deliberately Departed from His Methodology in Previous Cases,
Apparently To Achieve a Desired Result.
Plaintiffs admit that Dr. Werner changed his methodology from prior cases “to
use a methodology he personally had not used before.” (Opp. at 13.) Plaintiffs also admit that
Dr. Werner used a “broader” definition of news than he had in prior cases (id.), sweeping in
every news article that hit on search terms, rather than limiting his study to the types of news
likely to contain value-relevant information. (See Ex. 6, Grenadier Rebuttal ¶¶ 52-54 & Ex. 1.)
Plaintiffs claim that Dr. Werner’s new definition of “news days” was merely one
of many permissible choices. They claim Dr. Werner was being “objective” by not screening for
6 Plaintiffs purport to remedy this defect in Dr. Werner’s new declaration. That declaration
is impermissible for the reasons explained below. (See section IV, infra.)
7 If the Court grants a hearing, Prof. Fischel would explain this point in greater detail.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 11 of 16
8
value-relevance. (Opp. 9-14.) But that does not explain why Dr. Werner did not use one of the
equally objective definitions of news he used in prior reports (e.g., all earnings releases, all Form
8-Ks or all press release). (See Ex. 6, Grenadier Rebuttal Ex. 1.)8 Nor can that decision be
explained by the over-inclusiveness of those past definitions (cf. Opp. Ex. 3 ¶¶ 19-28), given the
glaring examples of value-irrelevant articles that litter Dr. Werner’s data set here. (See Mot. 12-
16.) Dr. Werner’s choice can be explained only by the failure of his past definitions to yield the
desired results. (See Ex. 6, Grenadier Rebuttal Ex. 1.) While his new definition of news may be
objective, his decision to use that definition certainly was not.
Nor do Plaintiffs show that “other experts” have used this approach. (Opp. at 9
n.7, 13.) The experts in Angley and McIntire, like Profs. Grenadier and Fischel, employed
definitions of news days that required them to apply their judgment as economists to determine
whether an item was value-relevant. (Reply Ex. 6, Angley Decl. Ex. 12; Reply Ex. 5, McIntire
Decl. ¶ 49.) There is no evidence that the expert in In re Alstom SA Sec. Litig., 253 F.R.D. 266
(S.D.N.Y. 2008), used a definition of news days akin to Dr. Werner’s. Id. at 280. The expert in
Netbank did not test for a causal relationship at all. (Reply Ex. 4, Netbank Decl. ¶ 23.) And
Dr. Tabak has removed articles “solely reporting on [] price movements or volume” in all his
reports of which Robinhood is aware. (Mot. at 14.) Dr. Werner did not do any of this.
III.
DR. WERNER’S OPINIONS REGARDING EFFICIENCY IN HIS REBUTTAL
REPORT ARE ALSO UNHELPFUL AND UNRELIABLE.
As noted above, efficiency requires a causal relationship between value-relevant
news and price movements. In their opposition, Plaintiffs do not dispute that showing a cause-
and-effect relationship requires analysis of whether the effect is present both with and without
the alleged cause. Here, if (as Dr. Werner contends) news caused stock-price movements in the
weeks before the Class Period, then price movements should generally not be found absent new
information. (See Mot. at 19-20.) While Plaintiffs maintain that Dr. Werner’s rebuttal report
reliably answers that question (Opp. at 14-16), Dr. Werner’s testimony makes clear that this is
not part of his understanding of efficiency, and indeed, the test he employed in his rebuttal report
does not test that cause-and-effect relationship.
8 In any event, objectivity is not enough. Limiting news days to “Mondays” would be as
objective as it is uninformative. (See Reply Ex. 2, FXCM Dep. 181:7-182:14.) That is why Dr.
Werner, retained by the Rosen Law Firm, has previously used news day definitions that require
judgment (see, e.g., id. at 124:9-127:8)—like all the other experts he has tried to point to.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 12 of 16
9
As explained in Robinhood’s Daubert motion, Dr. Werner’s “binomial” test does
not test whether news causes price movements, because it does not compare days with news to
days without news. (Mot. at 18-20); In re PolyMedica Corp. Sec. Litig., 453 F. Supp. 2d 260,
269-70 (D. Mass. 2006). Plaintiffs’ denials that Dr. Werner admitted this at deposition (Opp. at
15) are contrived and contradicted by identical testimony ten pages earlier. (See Ex. 1, Werner
Dep. 324:17-25 (Q. “[I]n your binomial test as set forth in your rebuttal report, there is no
comparison between the statistical significance of the returns on the news dates on the one hand
and the statistical significance of the returns on the non-news dates on the other hand, correct? A.
I believe that is correct.”).) These admissions are dispositive.
Plaintiffs argue that a binomial test is reliable because some cases have accepted
it (Opp. at 15), but that is beside the point: it is a reliable test of something else altogether. A
binomial test detects only whether a set of events is too unlikely to be random. Thus, Dr. Werner
showed that the number of news days with statistically significant residual returns was too high
to be random. (Ex. 9, Werner Rebuttal ¶¶ 49-51.) But the number of non-news days with
statistically significant residual returns was also too high to be random. (Ex. 13, Dep. Ex. 288,
Analysis in Response to Werner Table 7.) The cause of the improbably high returns was
therefore not the news, since there were improbably high returns whether or not there was news.
Rather, it was the coordinated trading that characterized the week before the Class Period.
Because Dr. Werner’s rebuttal report does not test for a difference between news
and non-news days, the conclusion of efficiency he reaches is based on a fundamental
misunderstanding of the causal relationship required to show efficiency.
IV.
DR. WERNER’S NEW DECLARATION MUST BE STRICKEN.
Plaintiffs submitted with their Daubert opposition a new report by Dr. Werner,
never previously disclosed. The 25-page declaration (ECF No. 528-4) is a transparent attempt to
circumvent deadlines for expert reports and deprive Robinhood of the opportunity to ask
Dr. Werner about the new analyses or to respond to them. This declaration should be excluded.
Neither the Rules nor the Court’s scheduling order provides for a sur-rebuttal
report, much less one filed after Robinhood’s opposition to class certification. (ECF No. 517);
Fed. R. Civ. P. 26(a)(2)(D)(ii). The deadline to file a rebuttal report was March 28, 2023. (ECF
No. 517.) Robinhood has presented no expert evidence since the deposition of Prof. Grenadier
on April 19, 2023. Class certification discovery closed on April 28, 2023. (Id.) Dr. Werner’s
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 13 of 16
10
new declaration blows past all of these dates by two months. It violates Rule 26(a) and must be
excluded. See Fed. R. Civ. P. 26(a)(2)(D) (“A party must make [expert] disclosures at the times
and in the sequence that the court orders.”); Simmons v. Ford Motor Co., No. 18-cv-81558, 2021
WL 6062633, at *4 (S.D. Fla. Dec. 22, 2021) (same); Dyett v. N. Broward Hosp. Dist., No. 03-
cv-60804, 2004 WL 5320630, at *2 (S.D. Fla. Jan. 21, 2004) (exclusion under Fed. R. Civ. P.
37(c)(1) is “automatic” unless party can show violation was justified and harmless).
Compounding that failure, Dr. Werner has given Robinhood no ability to assess
his results. Contrary to Rule 26(a)(2)(B)(ii), he has not provided the “facts or data” underlying
his new analyses. For example, Dr. Werner has not provided the data that would be needed to
analyze his revised “news/no news” test. (Reply Ex. 1, Grenadier Decl. ¶¶ 12-14.) Under Rule
37, this also warrants exclusion. See Moore v. GNC, Holdings, Inc., No. 12-cv-61703, 2014 WL
12684287, at *3-4 (S.D. Fla. Jan. 24, 2014) (excluding report).
Plaintiffs bear the burden of explaining why their declaration is justified and
harmless. Mitchell v. Ford Motor Co., 318 F. App’x 821, 824-25 (11th Cir. 2009). They have
not done so, nor can they. Dr. Werner could have presented these analyses in his rebuttal report
on March 28, 2023. He was in possession of all expert materials proffered by Robinhood in this
case for over two months—since April 19, 2023—before filing his unauthorized sur-rebuttal.
Plaintiffs have not proffered any justification for that delay. Companhia Energetica Potiguar v.
Caterpillar Inc., No. 14-CV-24277, 2016 WL 3102225, at *9 (S.D. Fla. June 2, 2016); K & H
Dev. Grp., Inc. v. Howard, 255 F.R.D. 562, 567 (N.D. Fla. 2009). The prejudice to Robinhood is
substantial. Without any underlying data, it is impossible for Robinhood to assess Dr. Werner’s
new news/no-news test. Moore, 2014 WL 12684287, at *3. Robinhood has no way of knowing
which articles Dr. Werner has removed or retained and so no way of evaluating those choices.
Robinhood also had no opportunity to examine Dr. Werner on his new test for heteroskedasticity
(see supra at 5-6). Dr. Werner’s new declaration is unjustified and prejudicial and must be
stricken. Simmons, 2021 WL 6062633, at *4.
CONCLUSION
For these reasons and the reasons explained in Robinhood’s Daubert motion, Dr.
Werner’s opinions regarding efficiency, including his declaration dated June 21, 2023, should be
excluded. Plaintiffs’ opposition—along with the untimely sur-rebuttal report—also underscore
the importance of a hearing on the issues related to their motion for class certification.
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 14 of 16
11
Dated: June 28, 2023
/s/ Samuel A. Danon
HUNTON ANDREWS KURTH LLP
Samuel A. Danon (FBN 892671)
Tom K. Schulte (FBN 1025692)
María Castellanos Alvarado (FBN 116545)
333 S.E. 2 Avenue, Suite 2400
Miami, FL 33131
Telephone: (305) 810-2500
Facsimile: (305) 810-2460
sdanon@huntonak.com
tschulte@huntonak.com
mcastellanos@hunton.com
CRAVATH, SWAINE & MOORE LLP
Antony L. Ryan
Kevin J. Orsini
Brittany L. Sukiennik
825 Eighth Avenue
New York, NY 10019
Telephone: (212) 474-1000
Facsimile: (212) 474-3700
aryan@cravath.com
korsini@cravath.com
bsukiennik@cravath.com
Counsel for Defendants Robinhood Markets,
Inc., Robinhood Financial LLC and
Robinhood Securities, LLC
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 15 of 16
12
CERTIFICATE OF SERVICE
I HEREBY CERTIFY that on June 28, 2023, I electronically filed the foregoing
document with the Clerk of the Court using CM/ECF. I further certify that the foregoing
document is being served this day on all counsel of record via transmission of Notices of
Electronic Filing generated by CM/ECF or in some other authorized manner for those counsel or
parties who are not authorized to receive Notices of Electronic Filing.
Dated: June 28, 2023
/s/ Samuel A. Danon
Samuel A. Danon (FBN 892671)
Case 1:21-md-02989-CMA Document 583 Entered on FLSD Docket 06/28/2023 Page 16 of 16