Audet et al v. Garza et al

Filing 206

ORDER. For the reasons set forth in the attached, Fraser's motion to decertify the class as to damages is DENIED without prejudice. This case will proceed on a bifurcated basis, and the trial currently scheduled for August will be focused on l iability.Significant questions remain regarding the scope of the liability trial and the extent to which it should include some issues related to damages that might be resolvable on a class-wide basis. On these questions the Court would bene fit from further briefing. By May 25, 2020, the parties are directed to submit supplemental briefs of no more than twenty-five pages addressing the following issues:(1) Should any discrete issues pertaining to damages be determ ined by the jury in the class-wide liability trial? For example, should the reliability of the ZenCloud database be determined by the jury? Which other damages-related issues, if any, should be determined in the class-wide liability trial? The par ties should also address whether trying any such issues during the liability phase raises concerns under the Reexamination Clause of the Seventh Amendment.(2) What questions should the Court decide as a matter of law during the liability phas e of this case? Of these, which questions should the Court decide prior to trial, and which questions should it decide after hearing the evidence but prior to approving the final jury instructions? Each party should address the merits of any questi ons it would have the Court decide prior to trial.Reply briefs of no more than ten pages will be due by June 8, 2020.Finally, as the Court discussed with the parties at the conclusion of oral argument, although the trial remains scheduled for August 5, 2020, and the Court has set other trial-related deadlines, there is some uncertainty about these dates at this time because of the current pandemic. In addition, the Court recognizes that the briefing schedule set fo rth above will require the parties to incur additional time and expense in litigating this case. Therefore, if the parties would prefer to use the next month or two to engage in any settlement discussions, the Court is prepared to accommodate such a request, and would even consider moving the trial date if that is necessary. Should the parties wish to make such a request, they must file, within 14 days of this ruling, a joint statement certifying that (1) counsel have conferred with the ir clients and each other, (2) the parties wish to proceed to mediation, (3) the parties are willing to participate in settlement efforts at such mediation in good faith, and (4) counsel believe that a mediation stands at least a reasonable chance of resolving the case without trial. Any such statement should also indicate whether the parties seek the Court's assistance in arranging for a mediator (i.e., a U.S. Magistrate Judge or Court-appointed Parajudicial Officer) or whether they wish to pursue private mediation on their own.Signed by Judge Michael P. Shea on 5/4/2020. (Karpman, Michael)

UNITED STATES DISTRICT COURT
DISTRICT OF CONNECTICUT
DENIS MARC AUDET et al.,
Plaintiffs,

No. 3:16-cv-00940 (MPS)

v.
STUART A. FRASER et al.,
Defendants.
RULING ON MOTION TO DECERTIFY/BIFURCATION
I.

INTRODUCTION
This securities class action stems from the collapse of an expansive and allegedly

fraudulent cryptocurrency enterprise. Defendant Stuart A. Fraser has moved to decertify the
class as to damages. As set forth below, because there does not appear to be a method by which
a jury could determine aggregate damages with reasonable accuracy, and because bifurcation
serves the interests of judicial economy, I find that bifurcation is the most appropriate course of
action. I will determine how best to proceed with determining damages—including whether
decertification as to damages is warranted—if and when the question of liability is resolved in
the Plaintiffs’ favor. Accordingly, Fraser’s motion to decertify is DENIED without prejudice.
II.

BACKGROUND
I assume familiarity with the history of this case and the Court’s prior rulings, including

the Ruling on Class Certification, in which I granted the Plaintiffs’ motion for class certification
but authorized further discovery into individual damages issues. ECF No. 141. Upon the
completion of discovery, Defendant Stuart A. Fraser moved to decertify the class as to damages.
ECF No. 179. Fraser argues that calculating each class member’s damages requires a highly
individualized inquiry. Specifically, Fraser points to multiple offsets he argues complicate the
damages inquiry: credit card chargebacks, reseller sales, Paycoin sales, Paycoin-to-Ion
1

conversions, account sales, sales on the GAW Miners Marketplace, and netting gains and losses
for individuals with multiple accounts. Each of these is described in turn in the sections that
follow.
A.

Credit Card Chargebacks
Chargebacks are refunds that some class members received through their credit card

companies. While it is unclear how many class members received chargebacks, the evidence
before the Court indicates that this number is not insubstantial. Three of the seven class
members deposed by Fraser indicated that they received chargebacks. For example, class
member John Tuberosi indicated that he received chargebacks from all but two of the credit
cards he used to purchase products from GAW Miners LLC (“GAW”). ECF No. 179-2 at 94.
Similarly, Teresa Crivello indicated that she received approximately $100,000 in chargebacks.
ECF No. 179-2 at 98. 1 As the Plaintiffs point out, however, Fraser chose which class members
to depose, so these seven class members cannot be presumed to be representative of the class as a
whole.
A spreadsheet compiled by Named Plaintiff Dean Allen Shinners (“the Shinners
Spreadsheet”), which includes data for 490 class members, indicates that 40 received
chargebacks—approximately 8%. ECF No. 191 at 12 (citing ECF No. 191-3). But this data was
self-reported by the class members and appears to be inaccurate as to at least some class
members. For example, Teresa Crivello, one of the seven class members Fraser deposed,
admitted to having received substantial chargebacks, even though the Shinners Spreadsheet
indicates she did not receive any “refunds or chargebacks.” ECF No. 179-2 at 21-29. Class

1

Ms. Crivello indicated that her losses, net of any chargebacks she received, were between $200,000 and $215,000.
ECF No. 179-2 at 103.

2

member Daniel Simpson likewise testified to receiving chargebacks “after fighting tooth and nail
for over a year,” even though the Shinners Spreadsheet indicates he did not receive any “refunds
or chargebacks.” ECF No. 179-2 at 111; ECF No. 191-3 at 3. The Spreadsheet does, however,
indicate that Mr. Simpson had requested chargebacks the day before he submitted data to Mr.
Shinners, but had not yet received any. ECF No. 191-3 at 3. Regardless of how these
discrepancies came about, they suggest that the Shinners Spreadsheet may understate the number
of class members who received chargebacks.
Shinners also indicated in his Victim Impact Statement that he “started a drive to help
many investors recover some of their losses through the credit card charge-back process.” ECF
No. 181-1 at 3; see also ECF No. 179-2 at 126 (post by Shinners encouraging GAW investors to
request chargebacks and indicating that “[m]ost of the [credit card] customer service reps are
already aware of what is happening with GAW, since there has been a tsunami of recent activity
and effort to charge back these transactions.”). This evidence further suggests that chargebacks
were not isolated occurrences. 2
For those class members who did receive chargebacks, the available evidence indicates
that the amount of these chargebacks relative to the class member’s total investment in GAW
products varied, but that these chargebacks were not insubstantial. For example, Crivello
testified that she recovered approximately $100,000 of her roughly $300,000 investment through
chargebacks. ECF No. 179-2 at 98, 104. And several class members appear to have recovered
the entirety of their investment. ECF No. 179-1 at 140-47.

2

Fraser also cites evidence of several other instances of chargebacks, ECF No. 179 at 9 & nn. 19-21, but this
evidence is anecdotal and does not present any method by which this Court might determine the proportion of class
members who received chargebacks.

3

Chargebacks are not reflected in the ZenCloud database. 3 The proof of these transactions
includes the testimony of class members as well as credit card statements and other documents
evidencing chargebacks. See, e.g., ECF No. 191-10 at 3 (credit card statement showing a
refund); ECF No. 191-5 at 4 (letter from Citibank indicating a refund of $24,975). Further,
although Fraser was not able to obtain any discovery from several financial institutions he
subpoenaed, Fraser’s counsel indicated at oral argument that this request did not identify any
specific class members. It is difficult to imagine that financial institutions do not have records of
the chargebacks class members received, and that they could not locate these records upon
receiving targeted requests for specific class members with accounts at the institution. Finally,
no party has suggested any reasonably accurate method for calculating the aggregate value of the
chargebacks based on available data.
B.

Sales to Third Parties

1.

Reseller Sales
Some class members acted as resellers through the Companies’ Value Added Reseller

(“VAR”) program. For example, class member Ryan Grimes participated in the VAR program
and maintained a web store, “Hoosier Miner,” that sold hashlets. ECF No. 179-2 at 167, 173.
Resellers under the VAR program could buy codes from GAW that could be used to activate
hashlets at a reduced price, and then resell those codes at the same price that GAW charged the
public. ECF No. 179-2 at 173; ECF No. 191-6 at 15-16.

3

The ZenCloud database is a relational database that stores numerous data related to purchases of GAW products,
including order history and user information. The parties dispute the reliability of this database. See ECF No. 179-1
at 27-28; ECF No. 191 at 33-35.

4

The number of active resellers does not appear to have been substantial. Fraser identifies
two class members he claims were resellers: Grimes and Tuberosi. ECF No. 179-1 at 12-13. 4
While it is undisputed that Grimes was a reseller, the Plaintiffs contend that Tuberosi was not a
reseller in the sense of purchasing hashlets for his own account and then reselling these hashlets,
but that he simply referred customers to GAW and then conveyed their payment information to
GAW. ECF No. 191. This activity would not pose the same concern as reselling under the VAR
program, since Tuberosi never owned the products, and thus could not claim losses associated
with them in a claims process.
Plaintiffs indicate that they have identified six active resellers under the VAR program,
based on the confirmation emails these resellers sent to their customers. ECF No. 191 at 16.
Plaintiffs also point to the fact that, of the 111 class members who submitted written responses to
Fraser’s questions, 108 indicated they were not resellers. ECF No. 191-2 at 3 ¶ 11. Of the
remaining three, one indicated that he had an “affiliate / reseller account,” one indicated that he
“attempted” to be a reseller on the Amazon Platform, and one indicated that he “applied to be a
reseller but never resold anything.” ECF No. 191-24. Thus, it appears that at least 109 of the
111 class members who provided written responses did not engage in any reselling, and it is
unclear whether the remaining two individuals in this subset of class members actually engaged
in reselling.

4

Fraser suggests that Grimes’ losses are misrepresented in the Shinners Spreadsheet. Fraser argues that the
$140,000 listed in the Spreadsheet are not real losses, because Grimes resold the hashlets he purchased for at least as
much as he paid for them, and that Grimes in fact suffered “no losses.” ECF No. 179-1 at 13. But as a reseller,
Grimes asserts that he was subject to losses from chargebacks from the customers to whom he resold. See, e.g., ECF
No. 191-7 at 3 ($2,422.58 chargeback); ECF No. 191-8 at 2 ($747.50 chargeback). Further, at his deposition, he
indicated that to estimate his losses he had “totaled up what we had bought and totaled up what we had charged back
to us from our clients.” ECF No. 191-6 at 8. This seems to suggest that the purchases he included in this calculation
do not include purchases of hashlets that were later resold, but only those purchased for his own account.

5

Resellers under the VAR program appear to have received email receipts from GAW
memorializing their purchase and sent email receipts to their customers memorializing their
sales. ECF No. 179-2 at 169. 5 It further appears that whether an individual was engaging in
reselling under the VAR program can be determined by examining the ZenCloud database.
Fraser’s expert opined that “non-marketplace transactions,” including resale transactions, “leave
a specific trail in the ZenCloud database that [can be used] to determine whether any given
device held by a given account was either purchased from GAW Miners, purchased indirectly
from another user in the GAW Miners marketplace, or purchased indirectly from a seller outside
of the marketplace.” ECF No. 179-2 at 228. The terms of the transaction between the reseller
and the purchaser, however, do not appear to have been tracked in the ZenCloud database. Id. at
228-29. It is unclear whether the ZenCloud database can be used to estimate the aggregate total
of reseller transactions with reasonable accuracy.
2.

Paycoin Sales
Paycoins, unlike the other GAW products at issue in this case, could be withdrawn from

ZenCloud and sold on public exchanges. ECF No. 179-2 at 63. Proceeds from the sale of
Paycoin appear to be relevant to the calculation of damages for two types of transactions. The
first is purchases of Paycoin directly from GAW. Plaintiffs state that such transactions were
“relatively uncommon.” Id. The second is purchases of GAW products that produced income in
the form of Paycoins, such as hashlets, in which case Fraser argues that proceeds from the sale of
these Paycoins should be subtracted from any losses associated with the purchase of these
products.

5

Class member Grimes, who was a reseller, testified that he “[didn’t] know if [he had] a hundred percent of the
records.” ECF No. 179-2 at 169.

6

The magnitude of the potential offset from Paycoin sales depends heavily on the timing
of any sales. According to Fraser’s expert, if all class members sold their Paycoins immediately
upon withdrawing it from the ZenCloud database, this would reduce class-wide damages under
the Exchange Act by $4.3 million. ECF No. 179-2 at 234. 6 On the other hand, today, each
Paycoin is worth a fraction of a penny, and all of the Paycoins in circulation are worth
approximately $10,000. 7 Thus, if all class members held on to their Paycoins and sold them only
recently, the value of any offset would be negligible.
Fraser has produced evidence that some class members sold Paycoin before it had
become worthless. For example, the Shinners Spreadsheet indicates that class member Kevin
Calabrese sold off all of his Paycoin “when it became apparent that GAW was not going to live
up to the Honors program” and “sold off the last of [his Paycoin] . . . (at a substantial loss) by
early February [2015].” ECF No. 191-3 at 3. 8 Class member Tuberosi also testified that he sold
some of his Paycoin, including approximately two percent of his holding, on an exchange called
“Cryptsy.” ECF No. 191-14 at 3-4. Two additional class members indicated that they sold some
Paycoins but do not have complete records of these sales. See ECF No. 179-2 at 407 (Roman
Gorodnev), 662 (Christopher Alan Crane). At least one other class member, Martin Ammann,
sold Paycoin. This class member provided a transaction log from a public exchange. ECF No.
191 at 18; ECF No. 191-29.
Fraser’s expert also suggests that the high volume of Paycoin trading through April 3,
2014 (the end of the available data on Paycoin withdrawals from ZenCloud), which was six

6

Exchange Act damages under the model suggested by Plaintiffs’ expert, which do not account for Paycoin sales,
are approximately $12 million. ECF No. 179-2 at 594.
7
PayCoin, CoinMarketCap, https://coinmarketcap.com/currencies/paycoin2 (last visited April 28, 2020).
8
In early February, one Paycoin was worth approximately one dollar. PayCoin, CoinMarketCap,
https://coinmarketcap.com/currencies/paycoin2 (last visited April 28, 2020).

7

times the cumulative volume of withdrawals from ZenCloud, indicates that class members
actively traded Paycoin. ECF No. 179-2 at 232. Plaintiffs argue that much of this activity can be
attributed to the 12,000,000 Paycoins released into the market by GAW, which, according to
Plaintiffs, was almost four times as many coins as were assigned to class members. ECF No.
191 at 18.
There is also some evidence, however, suggesting that the sale of Paycoin for more than
negligible value was not widespread. Plaintiffs note that five of the seven class members Fraser
deposed testified that they did not sell Paycoin, and the sixth witness testified that he initially
“doubled and tripled down on PayCoin” and only sold it “when it was worthless,” probably for
“cents on the dollar.” ECF No. 191-28 at 4-5.
In short, while it is possible that a substantial number of class members received
significant value in exchange for Paycoin withdrawn from ZenCloud, it is unclear at this stage
whether this is the case. Similarly, while there is some indication that documentation of Paycoin
sales is limited, it is unclear how many class members lack documentation of their Paycoin sales.
Finally, other than the estimate provided by Fraser’s expert, which assumes that all Paycoin was
sold the day it was withdrawn from ZenCloud, ECF No. 179-2 at 234, and which likely
significantly overstates the value of Paycoin sales, there does not appear to be a reasonably
accurate method for calculating the aggregate value of this offset.
3.

Exchange of Paycoin for Ion
In June of 2015, a project called xpy.io was started to try to make Paycoin a viable

cryptocurrency. ECF No. 191-1 at 2. Eventually, it was determined that Paycoin would never
recover, and the team working on xpy.io decided to create a new cryptocurrency called Ion. Id.

8

Ion launched in April 2016. Id. One of the ways to acquire Ion was to trade Paycoin for Ion. Id.
One Ion coin could be acquired for eight Paycoins. Id.
Of the 111 class members who submitted responses to written questions, twenty-two
(approximately 20%) indicated that they exchanged their Paycoin for Ion. ECF No. 179-1 at 15.
It is unclear, however, whether class members received substantial value from such conversions.
By the time Ion was launched, Paycoin was trading at roughly $0.02 per coin, less than 1% of its
all-time high. 9 In the first six months after its launch, Ion generally traded for roughly $0.20 per
coin or less. 10 A class member exchanging Paycoin for Ion during this time period would have
received approximately $0.025 in value per Paycoin—still less than 1% of Paycoin’s all-time
high. But the price of Ion began to rise in early 2017, peaking at $8.18 in January of 2018,
before falling again in late 2018. 11 As of April 28, 2020, Ion traded at approximate $0.025 per
coin. 12 To the extent that class members exchanged Paycoin for Ion during its peak period in
early 2018, they could have derived significant value from the exchange. There is no evidence,
however, indicating that a material number of class members held on to their withdrawn Paycoin
until early 2018 before exchanging it for Ion around its peak.
No party has suggested a method for estimating the aggregate value of the Ion received
by class members in exchange for Paycoin. While there is some indication that the availability
of documentation for these transactions is limited, it is unclear how many class members lack
documentation for these exchanges.

9

PayCoin, CoinMarketCap, https://coinmarketcap.com/currencies/paycoin2 (last visited April 28, 2020).
Id.
11
Id.
12
Id.
10

9

4.

Account Sales
Some class members sold entire ZenCloud accounts by exchanging their username and

password for payment. The acquiror would then change the password and assume ownership of
the account. Often, escrow agents were used to facilitate these transactions.
Named Plaintiff Michael Pfeiffer, for example, testified that he used escrow agents to
purchase whole accounts. ECF No. 179-2 at 84-86. Class member Daniel Simpson testified that
he had purchased four accounts. ECF No. 179-2 at 109-10. Fraser has produced evidence of six
other such transactions. ECF No. 179-2 at 501 (Martin Ruzek to Dimitrios Anastasakis), 504
(negotiations between Martin Ruzek and Bart Kant), 507 (Guillaume Barlier), 509 (Nacer
Laradji to David Mah), 516 (“limburatorul” to David Mah), 519 (Ian MacPhee).
Private sales of whole accounts were widespread enough that there was a guide for
engaging in these transactions on the HashTalk forum. ECF No. 179-2 at 529-31 (“Private
Marketplace [Official Guide].”) Also, a GAW representative posted a “Warning About Buying
and Selling Accounts” on the forum, expressing concerns that GAW had about the practice. ECF
No. 179-2 at 521-27.
In short, while the evidence before the Court indicates that these transactions were not
isolated occurrences, it is unclear what portion of the class engaged in them.
Private sales of accounts were typically negotiated and memorialized via email, providing
some documentation of the transactions. See, e.g., ECF No. 179-2 at 509. In at least some cases,
the escrow agent provided a more formal receipt. See, e.g., ECF No. 179-2 at 501. No party has
suggested a method for calculating the aggregate value of accounts sold by class members to
third parties.

10

5.

Sales on the GAW Marketplace
Hashlets and Hashstakers could also be sold to third parties on the GAW Miners

Marketplace (“GAW Marketplace”). For example, class members Ryan Grimes and Mahendra
Phagu testified that they sold hashlets on the GAW Marketplace. ECF No. 179-2 at 165
(Grimes), 544 (Phagu). Phagu further testified that he thought the GAW Marketplace was “very
active.” Id. at 544. It is unclear how many class members actually engaged in this practice.
It is undisputed that sales on the GAW Marketplace were documented in the ZenCloud
database, although Fraser questions the reliability of this database. ECF No. 179-1 at 17; ECF
No. 191 at 19. The methodology suggested by the Plaintiffs’ expert for calculating aggregate
damages, which is based on the ZenCloud database, appears to account for these transactions.
C.

Netting Gains and Losses for Individuals with Multiple Accounts
Fraser argues that, for individuals who had multiple ZenCloud accounts, some of which

made money, these gains should be used to offset any losses. Of the 173 people who submitted
documents to Shinners, 31 indicated they had multiple accounts. See ECF No. 179-2 at 243-46
(ID Numbers 2, 13, 17, 33, 35, 36, 44, 49, 53, 54, 55, 58, 62, 64, 66, 67, 75, 81, 85, 92, 98, 12,
107, 127, 128, 129, 141, 143, 156, 161, 163). According to Fraser’s expert, 27 of the 173 class
members self-reported ownership of multiple accounts, and 11 of these 27 owned at least one
account that incurred losses and at least one account that realized gains. ECF No. 179-2 at 227.
Fraser’s expert opined that netting gains and losses for these 27 individuals would reduce their
damages by $75,163. Id.
Individuals often obtained multiple accounts by purchasing whole accounts from third
parties. See, e.g., ECF No. 179-2 at 109-10, 243 (Daniel Simpson). As discussed above, such

11

transactions were often negotiated and memorialized via email, see, e.g., ECF No. 179-2 at 509,
and in some cases, escrow agents provided more formal receipts, see, e.g., id. at 501.
The ZenCloud database can, at least to some extent, be used to identify individuals with
multiple accounts based on identifying information stored in the database. For example, Fraser’s
expert was able to identify additional, undisclosed accounts apparently owned by Named
Plaintiff Michael Pfeiffer. ECF No. 179-1. According to Fraser’s expert, however, these data
are inadequate to link together all of the accounts owned by each class member. ECF No. 179-2
at 227.
Although the Plaintiffs contend that the law does not require netting gains and losses
across multiple accounts, the Plaintiffs’ expert opined that, if necessary, the aggregate value of
this offset can be estimated using the ZenCloud database by netting gains and losses across all
accounts as if they were owned by a single individual. ECF No. 179-2 at 622-23. Such an
approach would understate damages, however, because it is likely that at least some accounts
with gains were owned by individuals with no other accounts.
III.

DISCUSSION
Before oral argument, I notified the parties that they should be prepared to discuss the

possibility of bifurcating liability and damages and trying the liability issues first, in addition to
the motion for decertification. ECF No. 203. At oral argument, the parties indicated their
amenability to bifurcation. 13 For the reasons that follow, I now adopt this approach, bifurcate the
proceedings in this case, and deny without prejudice the motion to decertify the class.

13

Both parties indicated, however, that bifurcation was not their first preference, with the Plaintiffs preferring a
single trial where a jury could determine an aggregate damages award, and the Defendants favoring decertification
of the class as to damages.

12

Rule 42 provides that, “[f]or convenience, to avoid prejudice, or to expedite and
economize, the court may order a separate trial of one or more separate issues, claims,
crossclaims, counterclaims, or third-party claims.” Fed. R. Civ. P. 42(b). District Courts in the
Second Circuit enjoy broad discretion in determining whether to bifurcate a trial. See, e.g.,
Johnson v. Celotex Corp., 899 F.2d 1281, 1289 (2d Cir. 1990) (“The decision to bifurcate is
within the discretion of the trial judge.”). “In exercising its discretion, however, the court must
consider—as the rule indicates—whether bifurcation would (1) avoid unfair prejudice to a party,
(2) provide for convenience, and (3) expedite the proceedings and be more economical.” Carson
v. City of Syracuse, 1993 WL 260676, at *2 (N.D.N.Y. July 7, 1993); see also 4 Newberg on
Class Actions § 11:5 (5th ed.) (indicating that “courts may consider a variety of factors,” and “no
one test [has emerged] as most prevalent,” but that “[w]hat the tests share is a focus on economy
and prejudice”).
In the class action context, Courts often bifurcate the determination of liability and
damages where individual inquiries are required to determine damages. See, e.g., In re
Playmobil Antitrust Litig., 35 F. Supp. 2d 231, 248 (E.D.N.Y. 1998) (“[The damages] inquiry,
however, is likely to require individualized proof. Accordingly, the Court finds that the most
efficient way to proceed in this case is to bifurcate the trial of these actions.”); In re Master Key
Antitrust Litigation, 70 F.R.D. 23, 28-29 (D. Conn. 1975) (bifurcating liability and damages in a
class action because “the proof as to damages is likely to be much more individualized” and, “if
we must reach it, may be easily resolved only by reference to a special master or through a long
series of separate minitrials”); see also Simon v. Philip Morris Incorporated, 200 F.R.D. 21, 39
(E.D.N.Y. 2001) (“The court of appeals for the Second Circuit has supported the trial judge’s

13

power to sever issues for trial before separate juries in class action lawsuits under Rule 42(b) and
23(c)(4)(A) so long as there are proper safeguards.”).
Here, too, bifurcation of liability and damages is warranted. First, a combined trial on
liability and damages, resulting in a jury verdict awarding aggregate damages to the Plaintiffs, is
likely to prejudice Fraser. The evidence before the Court suggests that the proper measure of
damages likely requires accounting for multiple, substantial offsets. The Plaintiffs have
proposed a reasonably accurate method for calculating the aggregate value of some of these
offsets. At least the chargebacks, however, appear to defy reasonable estimation on an aggregate
basis. A material number of class members have indicated that they received chargebacks, or
credit card refunds, related to their purchases of GAW products. Some class members indicated
that they recovered the entirety of their investment. ECF No. 179-1 at 140-47. Another class
member testified that she recovered around $100,000 of her roughly $300,000 investment. ECF
No. 179-2 at 98. Data collected by Named Plaintiff Dean Allen Shinners indicates that
approximately 8% of class members received chargebacks, ECF No. 191 at 12 (citing ECF No.
191-3), but there is some evidence suggesting that this figure understates the number of class
members who received chargebacks. Despite this evidence suggesting a material number of
chargebacks, the Plaintiffs have not articulated any reasonably accurate method by which a jury
might calculate the aggregate value of these chargebacks. 14
While “[a]ggregate computation of class monetary relief is lawful and proper” and
“[c]ourts have not required absolute precision as to damages,” Hart v. Rick’s Cabaret Intern.,
Inc., 73 F. Supp. 3d 382, 391 (S.D.N.Y. 2014) (quoting 3 Herbert B. Newberg & Alba Conte,

14

Plaintiffs also have not proposed any method for calculating aggregate damages for purchases of Paycoin and
Hashpoints from GAW.

14

Newberg on Class Actions § 10.5, at 483-86 (4th ed. 2002)), a District Court must nonetheless
“ensure that the damages awards roughly reflect the aggregate amount owed to class members.”
Hickory Securities Ltd. v. Republic of Argentina, 493 Fed. Appx. 156, 159 (2d Cir. 2012)
(quoting Seijas v. Republic of Argentina, 606 F.3d 53, 58-59 (2d Cir. 2010)); see also id. at 160
(If “an aggregate approach cannot produce a reasonable approximation of the actual loss, the
district court must adopt an individualized approach.”). Moreover, the Second Circuit has
rejected “[r]oughly estimating the gross damages to the class as a whole and only subsequently
allowing for the processing of individual claims” because such an approach “would inevitably
alter defendants’ substantive right to pay damages reflective of their actual liability.”
McLaughlin v. American Tobacco Co., 522 F.3d 215, 231 (2d Cir. 2008).
Bifurcating also serves the interests of judicial economy. After a trial on liability, the
Court will have before it a more fulsome factual record and will be better positioned to determine
how best to structure the subsequent proceedings. Moreover, depending on which of their claims
the Plaintiffs prevail on (if any), the relevant measure of damages may differ in material respects,
which may also impact the best approach to determining damages. And in the event that Fraser
prevails at the liability stage, the need to decide on a method for determining damages will of
course be obviated entirely.
Bifurcation subject to a later decision regarding how to proceed with the damages phase
is not uncommon in the class action context. See, e.g., In re Playmobil Antitrust Litig., 35 F.
Supp. 2d 231, 248 (E.D.N.Y. 1998) (“[T]he most efficient way to proceed in this case is to
bifurcate . . . . In the event that the jury finds the Defendant liable, the Court will then reconsider
the issue of whether class treatment of damages is feasible. At that point, the Court has a number
of options, including utilizing a formula to calculate damages, referring the damage issues to a

15

special master or trying these issues, perhaps after certifying appropriate subclasses, if
necessary.” (quotation marks and internal citations omitted)); Houser v. Pritzker, 28 F. Supp. 3d
222, 254 (S.D.N.Y. 2014) (“If and when the litigation reaches [the damages] stage, the Court
will have a number of management tools at its disposal to help resolve these issues. For example,
the Court could appoint a special master to preside over individual damages proceedings, or
could decertify the class after the liability phase and provide notice to plaintiffs as to how to
proceed to prove damages. . . . There is no need to decide at this time which avenue to pursue.”
(citation omitted)); see also In re Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124, 141
(2d Cir. 2001) (“There are a number of management tools available to a district court to address
any individualized damages issues that might arise in a class action, including: (1) bifurcating
liability and damage trials with the same or different juries; (2) appointing a magistrate judge or
special master to preside over individual damages proceedings; (3) decertifying the class after
the liability trial and providing notice to class members concerning how they may proceed to
prove damages; (4) creating subclasses; or (5) altering or amending the class.” (footnotes
omitted)).
In short, taking into account the likelihood of prejudice to Fraser as well as
considerations of judicial economy, I find that this case warrants bifurcation. In light of this,
Fraser’s motion to decertify is DENIED without prejudice. The Court will determine how to
structure the damages phase of this case—including whether individual damages issues
predominate over all common issues, such that decertification as to damages is warranted—if
and when the question of liability is resolved in the Plaintiffs’ favor.

16

IV.

CONCLUSION
For the foregoing reasons, Fraser’s motion to decertify the class as to damages is

DENIED without prejudice. This case will proceed on a bifurcated basis, and the trial currently
scheduled for August will be focused on liability.
Significant questions remain regarding the scope of the liability trial and the extent to
which it should include some issues related to damages that might be resolvable on a class-wide
basis. On these questions the Court would benefit from further briefing. By May 25, 2020, the
parties are directed to submit supplemental briefs of no more than twenty-five pages addressing
the following issues:
•

Should any discrete issues pertaining to damages be determined by the jury in the
class-wide liability trial? For example, should the reliability of the ZenCloud
database be determined by the jury? Which other damages-related issues, if any,
should be determined in the class-wide liability trial? The parties should also address
whether trying any such issues during the liability phase raises concerns under the
Reexamination Clause of the Seventh Amendment.

•

What questions should the Court decide as a matter of law during the liability phase
of this case? Of these, which questions should the Court decide prior to trial, and
which questions should it decide after hearing the evidence but prior to approving the
final jury instructions? Each party should address the merits of any questions it
would have the Court decide prior to trial.

Reply briefs of no more than ten pages will be due by June 8, 2020.
Finally, as the Court discussed with the parties at the conclusion of oral argument,
although the trial remains scheduled for August 5, 2020, and the Court has set other trial-related
deadlines, there is some uncertainty about these dates at this time because of the current
pandemic. In addition, the Court recognizes that the briefing schedule set forth above will
require the parties to incur additional time and expense in litigating this case. Therefore, if the
parties would prefer to use the next month or two to engage in any settlement discussions, the
Court is prepared to accommodate such a request, and would even consider moving the trial date
17

if that is necessary. Should the parties wish to make such a request, they must file, within 14
days of this ruling, a joint statement certifying that (1) counsel have conferred with their clients and
each other, (2) the parties wish to proceed to mediation, (3) the parties are willing to participate in
settlement efforts at such mediation in good faith, and (4) counsel believe that a mediation stands at
least a reasonable chance of resolving the case without trial. Any such statement should also indicate
whether the parties seek the Court’s assistance in arranging for a mediator (i.e., a U.S. Magistrate
Judge or Court-appointed Parajudicial Officer) or whether they wish to pursue private mediation on
their own.

IT IS SO ORDERED.

/s/
Michael P. Shea, U.S.D.J.
Dated:

Hartford, Connecticut
May 4, 2020

18