Federal Trade Commission v. Lending Club Corporation

Filing 291

ORDER RE: CROSS MOTIONS FOR SUMMARY JUDGMENT; PLAINTIFF'S MOTION FOR PARTIAL JUDGMENT ON THE PLEADINGS; CROSS MOTIONS TO EXCLUDE EXPERT TESTIMONY; ADMINISTRATIVE MOTIONS TO FILE UNDER SEAL. Signed by Magistrate Judge Jacqueline Scott Corley on 6/1/2020. (ahm, COURT STAFF) (Filed on 6/1/2020)

1
2

UNITED STATES DISTRICT COURT

3

NORTHERN DISTRICT OF CALIFORNIA

4
FEDERAL TRADE COMMISSION,

5

Plaintiff,

6
v.

7

LENDINGCLUB CORPORATION,

8

Defendant.

9
10

ORDER RE: CROSS MOTIONS FOR
SUMMARY JUDGMENT;
PLAINTIFF’S MOTION FOR PARTIAL
JUDGMENT ON THE PLEADINGS;
CROSS MOTIONS TO EXCLUDE
EXPERT TESTIMONY;
ADMINISTRATIVE MOTIONS TO
FILE UNDER SEAL
Re: Dkt. Nos. 137-40, 143, 145-47, 155,
201, 211, 215, 218, 221, 234, 236, 239, 241,
248-265, 268

11
United States District Court
Northern District of California

Case No.18-cv-02454-JSC

12
13
14

The Federal Trade Commission (“the FTC”) brings causes of action against LendingClub

15

Corporation (“LendingClub” or “Defendant”) alleging deceptive and unfair business practices in

16

violation of Section 5(a) of the Federal Trade Commission Act (“FTC Act”), 15 U.S.C. § 45(a),

17

and violation of the Gramm-Leach-Bliley Act’s (“GLB Act”) Privacy of Consumer Financial

18

Information Rule (“Privacy Rule and Regulation P”), 16 C.F.R. § 313, recodified at 12 C.F.R. §

19

1016. (Dkt. No. 57 at ¶ 1.)1 Now pending before the Court are LendingClub’s motion to exclude

20

expert testimony, (Dkt. No. 138), the FTC’s motion for partial judgment on the pleadings on

21

LendingClub’s affirmative defenses, (Dkt. No. 139), LendingClub’s motion for partial summary

22

judgment, (Dkt. No. 143), the FTC’s motion for summary judgment, (Dkt. No. 147), the FTC’s

23

motion to exclude expert testimony, (Dkt. No. 155), and the parties’ administrative motions to seal

24

portions of those filings, (Dkt. Nos. 137, 140, 145, 146, 201, 211, 215, 218, 221, 234, 236, 239,

25

241, 248-265).2 After careful consideration of the parties’ briefing and having had the benefit of

26
27
28

1

Record citations are to material in the Electronic Case File (“ECF”); pinpoint citations are to the
ECF-generated page numbers at the top of the documents.
2
Both parties have consented to the jurisdiction of a magistrate judge pursuant to 28 U.S.C. §
636(c). (Dkt. Nos. 6 & 12.)

1

oral argument on April 27, 2020, the Court rules as set forth below.
BACKGROUND

2
3

I.

The Parties

4

A.

5

“The FTC is an independent agency of the United States Government created by statute.”

6

(Dkt. No. 57 at ¶ 5 (citing 15 U.S.C. §§ 41-58).) As part of its duties, and as relevant here, “[t]he

7

FTC enforces Section 5(a) of the FTC Act, 15 U.S.C. § 45(a), which prohibits unfair or deceptive

8

acts or practices in or affecting commerce.” (Id.) In addition, the FTC “enforces the Privacy

9

Rule, 16 C.F.R. Part 313, recodified at 12 C.F.R. § 1016, which requires financial institutions to

10

The FTC

protect the privacy of consumer information.” (Id.)

United States District Court
Northern District of California

11

B.

LendingClub

12

LendingClub is a Delaware corporation formed in 2006; it is headquartered in San

13

Francisco, California. (Dkt. Nos. 140-4 at 9 (filed under seal) & 140-8, Ex. 1 at 6 (30:21) (filed

14

under seal).) LendingClub has operated an “online peer-to-peer lending platform” since 2007

15

“that connects borrowers and investors.” (Dkt. No. 140-4 at 9.) “Through LendingClub’s

16

platform, consumers can apply for and obtain personal loans underwritten and issued by a third-

17

party bank, WebBank.” (Id.) LendingClub’s platform also “facilitates the purchase of interests in

18

the loans by retail and institutional investors.” (Id.) LendingClub is not a lender; instead, it

19

“services the loans, manages customer service, and acts as an intermediary between borrowers and

20

investors.” (Id.) More than three million customers “have used LendingClub’s platform to obtain

21

loans worth, in aggregate, more than $43 billion.” (Id.)

22

II.

Factual Background

23

The following facts are undisputed.

24

A.

25
26

LendingClub’s Website and Online Loan Application
1.

“No Hidden Fees” and Disclosure of Origination Fee

Lending Club charges borrowers an “origination fee of 1% to 6%” of the total requested

27

loan amount in connection with its unsecured personal loans. (Dkt. No. 81 at ¶ 10.) The

28

origination fee “is deducted up-front from the proceeds disbursed to the borrower.” (Id.) Thus,
2

1

“the amount of money disbursed to a borrower’s bank account is less than the total loan amount by

2

the amount of the origination fee.” (Id. at ¶ 24.) Between 2013 and 2018, the origination fee

3

averaged nearly 5% of the borrower’s requested loan amount. (Id. at ¶ 23.) Borrowers are

4

obligated to repay the entire loan amount they initially requested, including the origination fee.

5

(Dkt. No. 145-4 at ¶ 3 (filed under seal).)

6

From 2012 until mid-2018, LendingClub advertised its personal loans to consumers

7

through mail, email, and online advertisements as containing “no hidden fees.”3 (Dkt. Nos. 81 at ¶

8

10 & 145-4 at ¶¶ 1, 31-38.) During that time consumers could apply for personal loans directly

9

through LendingClub’s website, www.lendingclub.com. (Dkt. No. 81 at ¶ 15.) Online loan
applicants were required to complete a form providing basic information that LendingClub used to

11
United States District Court
Northern District of California

10

assess “baseline creditworthiness criteria” and conduct a “front-end” credit check. (Dkt. No. 81 at

12

15.) LendingClub rejected applicants who failed to pass the front-end check’s baseline criteria.

13

(Id.) Applicants who met the baseline criteria were directed to a loan offer webpage (“Offer

14

page”) that included a bold-faced “Loan Amount.” (Id. at 15.) The webpage represented that the

15

loan included “[n]o hidden fees” below a masked smiley face (“bandit”) icon. (Id.)

16
17
18
19
20
21
22
23
24
25
26
27
28

3

LendingClub removed the statement “no hidden fees” from its website in May 2018. (Dkt. No.
81 at 9.) Some mail advertisements continued to state “no hidden fees” until September 2018.
(See Dkt. No. 145-4 at ¶ 32.) LendingClub adopted the FTC’s preferred disclosures on the Offer
page in February 2019. (Dkt. No. 218-4 at 27.)
3

1

(Dkt. No. 145-4 at ¶ 10.)4 On desktops, the webpage allowed users to click on a “tooltip,” which

2

is “a green dot with a white question mark inside” appearing beside the word “APR.” (Dkt. No.

3
4

81 at ¶ 20.) If users clicked on or hovered over the tooltip, a pop-up window appeared with text
explaining the Annual Percentage Rate (“APR”) and notifying users that the APR includes “a onetime origination fee . . . that is collected out of your loan proceeds.” (See id.; see also Dkt. No.

5
145-4 at ¶ 13.) The pop-up window provides the exact dollar amount of the origination fee. (Id.)
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17

(Dkt. No. 145-4 at ¶ 13.) Users accessing the webpage through a mobile device would have to
click on the APR rate itself to produce the same pop-up window. (Dkt. No. 81 at ¶ 22.) An
applicant was not required to click on the tooltip to proceed with the loan application.5 (Id. at ¶
21.)

18
19
20
21
22
23
24
25
26
27
28

4

All screenshots are taken from the FTC’s motion for summary judgment and LendingClub’s
opposition, and the exhibits in support of same, or the Court’s October 2018 Order. In some
instances, the Court cropped the screenshots for formatting purposes. The FTC’s motion specifies
that the screenshots reflect the December 2016 version of LendingClub’s desktop website. (See
Dkt. No. 150-3, Ex. 3 at 2.) There is no dispute that the screenshots reflect LendingClub’s online
loan application flow as it existed “for most of the relevant time period.” (Dkt. No. 145-4 at ¶ 8.)
5
In June 2017, LendingClub added a footer to the Loan Offer Page stating: “All loans made by
WebBank, Member FDIC. Your actual rate depends upon credit score, loan amount, loan term,
and credit usage & history. The APR ranges from 5.99% to 35.89%. For example, you could
receive a loan of $6,000 with an interest rate of 7.99% and a 5.00% origination fee of $300 for an
APR of 11.51%. In this example, you will receive $5,700 and will make 36 monthly payments of
$187.99. The total amount repayable will be $6,767.64. Your APR will be determined based on
your credit at the time of application. The origination fee ranges from 1% to 6% and the average
origination fee is 5.49% as of Q1 2017. There is no down payment and there is never a
prepayment penalty. Closing of your loan is contingent upon your agreement of all the required
agreements and disclosures on the www.lendingclub.com website. All loans via LendingClub
have a minimum repayment term of 36 months or longer.” (Dkt. No. 149-6, Ex. 34 at 4.) As of
February 28, 2019, the Loan Offer page discloses the origination fee without use of the tooltip and
in addition to the footer. (Dkt. No. 145-4 at ¶ 77 (citing Dkt. No. 149-6, Ex. 34).)
4

1

The origination fee was next disclosed on the Loan Rate & Terms webpage containing the

2

federally-required Truth In Lending Disclosure Statement (“TILA disclosure”). (Id. at ¶¶ 29, 30.)

3

The TILA disclosure’s Federal Box accounted for the origination fee in the amount listed in the

4

“Amount Financed” box stating: “The amount of credit provided to you or on your behalf.”

5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

(Dkt. No. 53 at 13.)6

26

“[I]n most standard screen configurations,” the user must then scroll down webpage to see

27
28

6

The Court has altered the screenshot by presenting it as two separate images to increase the
clarity of the images and provide a better representation of the actual user view.
5

1

an itemized breakdown that includes the “Origination Fee[ ].” (Dkt. No. 81 at ¶ 30.)

2
3
4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13

(Dkt. No. 53 at 14.) Applicants must scroll to the bottom of the page and click “Next” to complete

14

the loan application. (Dkt. No. 81 at ¶ 30.) Prior to January 2017, some mobile users could click

15

“I Agree” to complete the loan application without having to scroll past the TILA disclosure and

16

itemized breakdown listing the origination fee. (Id. at ¶ 32.)

17

After clicking “Next,” applicants entered their bank account information and were taken to

18

an “Account Summary” page that until November 2017, stated: “Your $[amount requested] loan

19

is on the way. What’s next?” (Id. at ¶ 34; see also Dkt. No. 145-4 at ¶ 80.)

20
21
22
23
24
25
26
27
28
6

1

(Dkt. No. 220-5, Ex. 26 (redacted version of document filed under seal).) The loan amount listed

2

did not reflect subtraction of the origination fee. The Account Summary page is the final page of

3

the application flow. (Dkt. No. 145-4 at ¶ 80.)
2.

4
5

Disclosure of the Origination Fee on the “Rates & Fees” Webpage

LendingClub’s website contains a “Rates & Fees” page that is separate from the online

6

loan application flow. (Id. at ¶ 29.) To access the page from the website’s homepage, users

7

navigate to the “Personal Loans” page and then click on “Rates & Fees.” (Id.) The page explains

8

that the origination fee is an up-front fee that is deducted from the amount applied-for.

9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18

(Dkt. No. 223-3, Ex. 2 at 5.) Because the “Rates & Fees” page is not included in the loan

19

application flow, an applicant need not view the page to apply for a loan.

20

During the relevant time period, LendingClub received tens of thousands of inquiries and

21

hundreds of formal complaints from consumers regarding the origination fee. (Dkt. No. 145-4 at

22

¶¶ 45-52.) LendingClub’s internal consumer studies and surveys, compliance reviews, employee

23

training materials, and communications among employees also addressed issues regarding

24

disclosure of the origination fee. (Id. at ¶¶ 53, 55, 59-66, 73-75.) Between April 2013 and

25

February 2019, LendingClub issued over 3 million personal loans and collected over $2 billion in

26

origination fees. (Dkt. No. 152-1, Ex. 8 at ¶ 13 (filed under seal).)

27

B.

Communication of Loan Approval

28

Until November 2017, the Account Summary page informed applicants that their loans
7

1

were “on the way”; however, the page also included a “To-Do List” and informed applicants that

2

they must complete the items on the list and LendingClub would then “finish the final review” of

3

the application. (See Dkt. No. 220-5, Ex. 26 (“Please complete all tasks on your To-Do List. Once

4

complete, we can finish the final review of the application.”); see also Dkt. No. 145-4 at ¶ 80.)
Between April 2015 and mid-November 2015, LendingClub sent applicants who had

5
6

completed the application through the Account Summary page an email stating in large bold type:

7

“Your Loan is 100% Backed.” (Dkt. No. 177-1, Ex. 128 at 5-11 (filed under seal).) One version

8

stated:

9
10

United States District Court
Northern District of California

11
12
13
14
15
16

(Id. at 7.) Other versions stated that the applicant had to complete the “To-Do List” and

17

LendingClub had to complete its “final review” before the applicant received the loan.

18
19
20
21
22
23
24
25
26
27

(Id. at 11.)

28
8

1

Another version provided additional detail regarding the final review process.

2
3
4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13

(Id. at 9.)
Beginning in mid-November 2015, LendingClub replaced the “Your Loan is 100%

14

Backed” large-print language with “You’re one step closer to getting your loan.” (See id. at 12-

15

14.) However, the emails retained the language in the body: “Great news! Investors have backed

16

your loan 100%.” (See id.) The emails included this language until October 2018. (Dkt. No. 145-

17

4 at ¶ 195.)

18

Some applicants who completed all the items on their To-Do List were ultimately denied a

19

loan after LendingClub conducted its “back-end” review of their applications. (Dkt. No. 81 at ¶¶

20

35-40.) Similarly, prior to June 2016, LendingClub denied loans to some applicants after they had

21

received the “100% Backed” emails. (Id. at ¶ 36.) “LendingClub voluntarily changed its

22

processes for nearly all classes of loans” in June 2016, and since that time, “only consumers who

23

successfully completed the verification process would have their loan applications listed for

24

investor backing and be notified when their loans were fully backed by investors.” (Id.)

25

LendingClub received complaints from consumers that the Account Summary page’s

26

language that their loans were “on the way” was confusing and indicated that applicants were

27

already approved, and internal reports acknowledged that the page was confusing. (Dkt. No. 145-

28

4 at ¶¶ 81-86.) LendingClub also received complaints regarding the “100% Backed” language
9

1

from applicants who thought the emails indicated that their loans were approved. (Id. at ¶ 94

2

(citing Dkt. Nos. 190-21, Ex. 247 (filed under seal) & 150-7, Ex. 7 (filed under seal)).) Internal

3

LendingClub communications recognized that the language could be confusing. (Id. at ¶¶ 95, 96.)

4

C.

Withdrawal of Fees

5

LendingClub’s “default method of receiving consumers’ scheduled monthly payments is

6

automated electronic bank account withdrawal via ACH [Automated Clearing House] transfer.”

7

(Dkt. No. 57 at ¶ 42 & 81 at ¶ 42.) Consumers can choose to opt out of ACH transfers and pay by

8

check. (Dkt. No. 81 at ¶ 42.) 98% of LendingClub’s consumers pay by ACH transfer. (Dkt. No.

9

140-6 at ¶ 11.)
LendingClub’s use of ACH transfers has resulted in erroneous withdrawals, “including

10

United States District Court
Northern District of California

11

double payments and charging of customers who cancelled automatic payments.” (Id. at ¶ 43.) In

12

some cases, consumers were erroneously charged hundreds or thousands of dollars and incurred

13

overdraft fees. (Dkt. No. 145-4 at ¶¶ 105-06.) LendingClub has refunded erroneous withdrawals

14

and overdraft fees when notified by consumers of the issue. (Dkt. No. 81 at ¶ 47.)
“Between August 2016 and June 2019, consumers contacted their financial institutions to

15
16

dispute Defendant’s unauthorized ACH withdrawals 5,490 times, totaling at least $3.8 million.”

17

(Dkt. No. 145-4 at ¶ 109.)

18

D.

Privacy Policy

19

LendingClub collects nonpublic personal information from prospective borrowers in its

20

loan application process. (Dkt. No. 81 at ¶ 48.) In May 2016, the FTC sent LendingClub a civil

21

investigative demand (“CID”) inquiring about its privacy disclosures to loan applicants. (Dkt. No.

22

145-4 at ¶ 139.) Since late 2016, LendingClub has required prospective borrowers to check a box

23

acknowledging its Privacy Policy “as a necessary step to completing a personal loan application

24

through LendingClub.” (Dkt. No. 81 at ¶ 51.) Prior to December 2016, “LendingClub required

25

prospective borrowers to check a box agreeing to its Terms of Use, which included a consent to

26

the Privacy Policy, as a necessary step to completing a personal loan application through

27

LendingClub.” (Id.)

28

//
10

1

III.

Procedural History
The FTC filed its original complaint in May 2018, bringing three claims against

2

LendingClub for the following alleged violations of Section 5 of the FTC Act: (1) deceptive

4

practices in charging the origination fee (“Count I”); (2) deceptive practices in communicating

5

loan approval prior to back-end review (“Count II”); and (3) unfair practices in making

6

unauthorized withdrawals from customer accounts (“Count III”). (Dkt. No. 1.) The original

7

complaint also brought a claim for violation of the GLB Act’s Privacy Rule and Regulation P

8

(“Count IV”). (Id.) LendingClub moved to dismiss the complaint pursuant to Federal Rule of

9

Civil Procedure 12(b)(6), and the Court granted that motion in part and denied it in part in October

10

2018; specifically, the Court granted dismissal with leave to amend as to Count III and denied the

11
United States District Court
Northern District of California

3

motion as to Counts I, II, and IV. (Dkt. No. 53.)
The FTC filed an amended complaint (the “FAC”) on October 22, 2018, asserting the same

12
13

four claims set forth in its original complaint arising from LendingClub’s representation of “no

14

hidden fees” to prospective consumer loan applicants, LendingClub’s practices related to its

15

consumer loan application process and consumer privacy protections, and LendingClub’s loan

16

repayment collection methods.7 (See generally Dkt. No. 57.) LendingClub filed its answer to the

17

FAC on November 13, 2018, asserting 19 affirmative defenses. (Dkt. No. 58.) Shortly thereafter

18

the FTC moved to strike certain affirmative defenses, (Dkt. No. 62), and the Court granted that

19

motion in part, (Dkt. No. 80). LendingClub filed an Amended Answer in May 2019, denying

20

liability as to all claims and asserting nine affirmative defenses. (Dkt. No. 81.)
Pursuant to the Court’s pretrial scheduling order, fact and expert discovery has closed, and

21
22

the matter is set for bench trial commencing on October 19, 2020. (Dkt. No. 93.) The parties filed

23

their respective motions for summary judgment on February 27, 2020. (Dkt. Nos. 143 & 147.)

24

The motions are fully briefed. (See Dkt. Nos. 216; 223; 240, 242.) The parties also filed their

25
26
27
28

7

Counts I and II of the FAC allege deceptive practices in violation of Section 5(a) of the FTC Act
based on a hidden loan origination fee, and LendingClub’s practice in communicating loan
approval, respectively. (Dkt. No. 57 at ¶¶ 56-61.) Count III alleges unfair practices under Section
5(a) based on unauthorized withdrawals from the accounts of LendingClub loan recipients. (Id. at
¶¶ 62-64.) Count IV alleges violations of the Privacy Rule and Regulation P based on
LendingClub’s failure to deliver an initial privacy notice to customers. (Id. at ¶¶ 65-67.)
11

1

respective motions to exclude expert testimony, (Dkt. Nos. 138 & 155), and the FTC filed its

2

motion for partial judgment on the pleadings, (Dkt. No. 139), on February 27, 2020. The motions

3

are fully briefed. (See Dkt. Nos. 202; 207; 212; 224; 235; 238.) The Court heard oral argument

4

on April 27, 2020.
DISCUSSION

5
6
7
8
9
10

I.

Motions for Summary Judgment
The FTC moves for summary judgment on all claims and LendingClub cross-moves for

summary judgment on Counts III and IV.
A.

Section 5 Deceptive Acts (Counts I and II)

Section 5 of the FTC Act prohibits “unfair or deceptive acts or practices in or affecting

United States District Court
Northern District of California

11

commerce.” 15 U.S.C. § 45(a)(1). To establish a Section 5 deceptive practices claim, the FTC

12

must prove three elements: (1) “there is a representation, omission, or practice that,” (2) “is likely

13

to mislead consumers acting reasonably under the circumstances, and” (3) “the representation,

14

omission, or practice is material.” FTC v. Stefanchik, 559 F.3d 924, 928 (9th Cir. 2009) (internal

15

quotation marks and citation omitted).

16

“Deception may be found based on the ‘net impression’ created by a representation.” Id.

17

(quoting FTC v. Cyberspace.Com LLC, 453 F.3d 1196, 1200 (9th Cir. 2006) (“A solicitation may

18

be likely to mislead by virtue of the net impression it creates even though the solicitation also

19

contains truthful disclosures.”)). Proof of actual deception is not necessary for purposes of

20

Section 5 liability; instead, the FTC need only show that the representation is likely to mislead

21

reasonable consumers. See, e.g., FTC v. AMG Capital Mgmt, LLC, 910 F.3d 417, 424 (9th Cir.

22

2018) (“This consumer-friendly standard does not require the Commission to provide proof of

23

actual deception.”) (alteration, internal quotation marks, and citation omitted); Cyberspace.Com,

24

453 F.3d at 1201 (noting that evidence of actual deception is not required but is “highly probative

25

to show that a practice is likely to mislead consumers acting reasonably under the circumstances”).

26

Accordingly, LendingClub’s insistence that the FTC must prove that LendingClub’s statements

27

“caused a significant minority of consumers to be deceived” (see, e.g. Dkt. No. 218-4 at 15-16) is

28

contrary to Ninth Circuit law and not the test the Court will apply.
12

1.

1
2

Count I

There is no dispute that LendingClub made the representation “no hidden fees” until mid-

3

2018. Thus, the first element of the FTC’s Section 5 deceptive practices claim is undisputed.

4

The question then is whether the FTC has satisfied its initial burden on summary judgment of

5

showing that “no reasonable factfinder could conclude that the solicitation was not likely to

6

deceive consumers acting reasonably under the circumstances,” and that the representation was

7

material. See Cyberspace.Com, 453 F.3d at 1201. Whether LendingClub’s representation that

8

prospective borrowers would receive a specific loan amount with “no hidden fees” was likely to

9

mislead a reasonable consumer hinges on whether the origination fee was in fact “hidden” or

10

United States District Court
Northern District of California

11

concealed in the loan application flow; thus, the starting point is the loan application flow itself.
A reasonable trier of fact could find that the net impression of “no hidden fees” was likely

12

to mislead based on a facial review of the loan application flow itself. To recap: LendingClub’s

13

advertisements and loan application flow represented to prospective borrowers that they could

14

obtain a specific loan amount with “no hidden fees.” The loan application then concealed the

15

origination fee behind a tooltip and in a fine print footer before affirmatively disclosing it at the

16

bottom of the TILA disclosure in a single line between two paragraphs containing bold text.

17

There is no dispute that prior to the May 2018 removal of the “no hidden fees” representation, the

18

loan application flow itself contained no other disclosures of the origination fee. Thus, the FTC

19

has satisfied its initial burden of demonstrating that the “no hidden fees” representation was likely

20

to mislead consumers acting reasonably under the circumstances because the loan application flow

21

concealed the origination fee; that is, the origination fee was hidden.

22

The FTC asserts that the Court’s inquiry should end there because a facial review of the

23

loan application flow is all that is necessary to resolve Section 5 deception cases on summary

24

judgment. In other words, the FTC contends that every reasonable trier of fact would have to find

25

that the net impression was misleading. While a reasonable trier of fact could find that the TILA

26

disclosure was insufficient to make the “no hidden fees” representation not misleading, drawing

27

all reasonable inferences in LendingClub’s favor, a reasonable trier of fact could find that it was

28

sufficient.
13

1

First, as LendingClub has vigorously argued throughout this case, the disclosure of the

2

origination fee was TILA compliant; in other words, federal law did not require any further

3

disclosure of the origination fee (putting aside the FTC Act). From that fact, a reasonable trier of

4

fact might find that the origination fee was not hidden. The cases upon which the FTC relies for

5

the proposition that such an inference is not reasonable are distinguishable. In FTC v. AMG

6

Capital Mgmt, LLC, for example, the Ninth Circuit affirmed the district court’s summary

7

judgment ruling in favor of the FTC based on the facially deceptive nature of the representation at

8

issue—a TILA disclosure page in a “Loan Note” that prominently presented loan repayment terms

9

in the Federal Box of the TILA disclosure but set forth different repayment terms in fine print
below the Federal Box that included additional charges, and borrowers were automatically entered

11
United States District Court
Northern District of California

10

into the fine print repayment plan and had to affirmatively opt-out. 910 F.3d at 422-23. The court

12

agreed “that the loan note was deceptive because it did not accurately disclose the loan’s terms,”

13

finding that the “Loan Note’s fine print does not reasonably clarify [the repayment terms] because

14

it is riddled with still more misleading statements” and “fails to cure the misleading ‘net

15

impression’ created by the TILA box.” Id. at 423-24. Here, in contrast, there is no argument that

16

the TILA disclosure was inaccurate as to the fees charged, only that the disclosure of the

17

origination fee was not sufficiently prominent to not be considered “hidden.” That may be the

18

trier of fact’s final determination, but that is not a finding that can be made without weighing

19

evidence.

20

Cyberspace.Com is similarly distinguishable. There the FTC brought a Section 5 claim

21

based on a solicitation in the form of a check for a few dollars, the net impression of which

22

conveyed to reasonable consumers the resolution of “some small, outstanding debt.” 453 F.3d at

23

1200-01 (internal quotation marks and citation omitted). However, a fine print notice “on the

24

reverse side of the check” informed consumers that cashing the check activated a monthly internet

25

service. Id. The court conducted a facial review of the check and affirmed the district court’s

26

ruling “that no reasonable factfinder could conclude that the solicitation was not likely to deceive

27

consumers acting reasonably under the circumstances.” Id. at 1201. The court then went on to

28

note that undisputed extrinsic “bolstered” its conclusion; specifically, the solicitation “actually
14

1

deceived nearly 225,000 individuals and small businesses,” or roughly 99 percent of all recipients.

2

Id. No such undisputed facts regarding overwhelming actual deception are present here.
Second, at least some of LendingClub’s extrinsic evidence further supports an inference

3
4

that the net impression of the “no hidden fees” representation was not likely to mislead a

5

reasonable consumer about the origination fee. For purposes of this Order the Court will discuss

6

two areas: Consumer Financial Protection Bureau (“CFPB”) Review and Conversion Rates.
a.

7

CFPB Review

LendingClub voluntarily sought to work with the CFPB over a period of years on “Project

8

Catalyst,”8 during which the CFPB conducted a “page-by-page review of -- and were thoroughly

10

familiar with every aspect of the loan application flow, . . . includ[ing] references to the phrase,

11
United States District Court
Northern District of California

9

‘no hidden fees.’” (Dkt. No. 269 at 19:22-25.) LendingClub asserts that the CFPB “did not raise

12

any issue” with the loan application flow during its review. (Id. at 19:25-20:1.)
LendingClub’s briefing cites the deposition testimony of its Chief Executive Officer

13
14

(“CEO”) Scott Sanborn, who testified that the LendingClub applied for Project Catalyst in 2013,

15

engaged in discussions with the CFPB regarding LendingClub’s TILA disclosure, and as part of

16

those discussions, the CFPB reviewed LendingClub’s “entire end-to-end process” and “went page-

17

by-page through the loan application funnel.” (See Dkt. No. 219-11 at 10-15.) Mr. Sanborn

18

further testified that LendingClub sent the CFPB “a document containing screenshots of the entire

19

application flow” at the “[e]nd of 2015.” (Id. at 15 (235:7-12).) Mr. Sanborn testified that the

20

CFPB raised an issue with LendingClub’s since-discontinued processing fee for customers who

21

repay their loans by check, (see id. at 14-15 (234:2-235:6)); however, the CFPB “expressed no

22
23
24
25
26
27
28

8

The CFPB is an independent government agency tasked with “enforc[ing] Federal consumer
financial law consistently for the purpose of ensuring that all consumers have access to markets
for consumer financial products and services and that markets for consumer financial products and
services are fair, transparent, and competitive.” 12 U.S.C. § 5511(a). The CFPB’s statutory
objectives include “ensuring that, with respect to consumer financial products and services[:] (1)
consumers are provided with timely and understandable information to make responsible decisions
about financial transactions; [and] (2) consumers are protected from unfair, deceptive, or abusive
acts and practices and from discrimination.” 12 U.S.C. 5511(b)(1)-(2). The CFPB launched
“Project Catalyst” in November 2012 in an effort to collaborate with companies and “encourage
consumer-friendly innovation and entrepreneurship in markets for consumer financial products
and services.” (Dkt. No. 223-12, Ex. 11 at 2.)
15

1

concerns regarding the origination fee disclosures in LendingClub’s application flow,” (see Dkt.

2

No. 218-4 at 21 (emphasis omitted)).

3

LendingClub also cites the deposition testimony of James Jackson, WebBank’s Senior

4

Vice President of Strategic Partner Oversight, who testified in his capacity as a corporate

5

representative for WebBank. Mr. Jackson testified as to LendingClub’s discussions with the

6

CFPB regarding Project Catalyst, and stated that he could not “recall any interactions where [the

7

CFPB] suggested that the process or the application flow or the disclosures were out of

8

compliance.” (Dkt. No. 223-14, Ex. 13 at 10 (43:1-3).) In response to a question regarding

9

whether the CFPB would have raised an issue if they “saw a problem with the disclosures of the

10

United States District Court
Northern District of California

11
12
13
14
15

origination fee and the application flow,” Mr. Jackson answered:
So obviously it’s speculation, I can’t say what they would or would
not do, but my general experience and understanding of the CFPB is
if they think you’re doing something wrong, you know, we would
have been talking about that rather than about a test for new
disclosures. Right? We were having a positive, how-can-weimprove-things conversation. That conversation, I suspect -- and
again, only my speculation -- would have been different had they
actually thought we were doing something wrong.”

16

(Id. at 10 (43:11-20).) LendingClub asserts that “[t]he fact that CFPB staff conducted extensive

17

diligence on LendingClub’s entire application flow and did not find any issues with the origination

18

fee disclosures is strong evidence that the disclosures are neither ‘hidden’ nor deceptive.” (Dkt.

19

No. 218-4 at 21.)

20

LendingClub’s argument as to the “strong” weight of such evidence is misplaced; on

21

summary judgment it is enough if the evidence creates a genuine dispute of material fact as to

22

whether the representation was likely to mislead. However, viewing the evidence in the light most

23

favorable to LendingClub, the lack of action on the part of the agency charged with ensuring that

24

consumers of financial products and services are protected from unfair or deceptive practices—

25

after that agency’s review of the challenged service—is both genuine and material because it gives

26

rise to a reasonable inference that the net impression of the representation “no hidden fees” was

27

not likely to mislead.

28

The FTC asserts that LendingClub’s argument regarding CFPB review of the loan
16

1

application flow is “both immaterial and unsupported.” (Dkt. No. 239-5 at 13.) Not so. First, if

2

the CFPB conducted a “page-by-page” review of the loan application flow and did not flag an

3

issue as to the disclosure of the origination fee, while at the same time raising a concern about the

4

processing fee, then that evidence is material to the second element of a Section 5 claim because

5

the theory underlying Count I is that the representation “no hidden fees” was likely to mislead a

6

reasonable consumer based on the allegedly inadequate disclosure of the origination fee.

7

Second, the argument is supported by the testimony of Mr. Sanborn and Mr. Jackson.

8

Although the submitted portion of Mr. Sanborn’s deposition transcript does not directly address

9

whether the CFPB expressed concerns regarding disclosure of the origination fee, Mr. Jackson
testified that he did not “recall any interactions where [the CFPB] suggested that the process or the

11
United States District Court
Northern District of California

10

application flow or the disclosures were out of compliance.” (See Dkt. No. 223-14 at 10 (43:1-3).)

12

The FTC asserts that Mr. Jackson’s testimony constitutes “speculative opinion testimony” from an

13

obviously biased source who “was not qualified as an expert and offered no report.” (Dkt. No.

14

239-5 at 13.) The FTC is correct that the portion of Mr. Jackson’s testimony discussing his

15

“general experience and understanding of the CFPB” was speculative; however, the testimony

16

regarding his recollection of the CFPB’s review was not. Mr. Jackson is of course a competent

17

witness as to his own recollection of events. And the FTC’s argument as to bias is a classic trial

18

argument not appropriate for summary judgment.

19

In sum, the cited testimony supports LendingClub’s overarching argument that the CFPB

20

reviewed the loan application flow and did not raise an issue with the disclosure of the origination

21

fee. On summary judgment the Court cannot weigh that evidence or determine the credibility of

22

the deponents. See George v. Edholm, 752 F.3d 1206, 1214 (9th Cir. 2014) (noting that

23

“credibility determinations, the weighing of the evidence, and the drawing of legitimate inferences

24

from facts” are inappropriate on summary judgment) (internal quotation marks and citation

25

omitted). It is sufficient that the deposition testimony gives rise to a reasonable inference that the

26

CFPB did not view the loan application containing the “no hidden fees” representation as being

27

misleading. Further, evidence that LendingClub voluntarily submitted its loan application flow to

28

the CFPB for review and feedback years prior to this action “underscores how this case is very
17

1

unlike the straightforward deception cases upon which the FTC purports to rely.” See FTC v.

2

DirecTV, Inc., No. 15-cv-01129-HSG, 2018 WL 3911196, at *18 (N.D. Cal. Aug. 16, 2018)

3

(concluding following a bench trial that defendant’s “investment in substantial resources in

4

analyzing its operations, candidly identifying areas for improvement, and following through on a

5

number of improvements does not support a finding that the company violated the FTC Act”).
b.

6

Conversion Rates

LendingClub also argues that there is a genuine dispute over “conversion rate evidence”

7
8

that renders summary judgment improper. (Dkt. No. 218-4 at 27.) LendingClub defines the

9

“conversion rate” as “the percentage of applicants who ultimately become borrowers.” (Id. at 18
n.9.) The FTC similarly defines it “as the rate at which consumers proceed with the loan

11
United States District Court
Northern District of California

10

application and become revenue-generating borrowers.” (Dkt. No. 145-4 at ¶ 39.)
The FTC’s Statement of Undisputed Material Facts states, in pertinent part, that

12
13

LendingClub “rejected suggested improvements to its fee disclosures because it believed that

14

clearly and conspicuously disclosing the up-front fee would drive down conversion.” (Id. at ¶ 67.)

15

LendingClub’s opposition counters that “LendingClub always understood the origination fee to be

16

clearly and conspicuously disclosed on the TILA page,” and cites data “showing that added

17

origination fee disclosures on the Offer page do not materially impact application-to-borrower

18

conversion rates.” (Dkt. No. 218-4 at 27 (citing, among others, Dkt. No. 219-5, Ex. 6 at ¶¶ 109-10

19

(Expert report of Dr. Yoram (Jerry) Wind (filed under seal)).) The FTC seeks to exclude the

20

testimony of Dr. Wind.9 (See Dkt. No. 175-1 at 9-17 (filed under seal).) Thus, the Court first

21

addresses Dr. Wind’s testimony and whether it is admissible.

22

i.

Dr. Wind’s Expert Testimony

23

LendingClub hired Dr. Wind “to assess the effectiveness of the disclosures of the

24

origination fee associated with loans available through LendingClub’s platform.” (Dkt. No. 219-

25

5, Ex. 6 at ¶ 1.) Dr. Wind conducted an initial experiment in August and September 2018 wherein

26
27
28

9

The FTC moves to exclude testimony from each of LendingClub’s four experts. (See generally
Dkt. No. 175-1.) The Court addresses the parties’ respective Daubert motions only to the extent
necessary to resolve their cross-motions for summary judgment.
18

prospective LendinClub borrowers were assigned to a “Test” or “Control” group. (Id. at ¶¶ 1, 24,

2

36.) The groups were exposed to different “real-world stimuli” (i.e., loan application flows) in

3

“natural and realistic settings.” (Id. at ¶¶ 1, 29-30.) Dr. Wind’s initial experiment methodology

4

compared the challenged origination fee disclosures on the Offer page (experienced by the Test

5

group) to the FTC’s preferred disclosure (experienced by the Control group).10 (Id. at ¶¶ 36, 39-

6

40.) The Test group’s loan application flow did not include the “no hidden fees” representation

7

and bandit icon, which LendingClub removed from the loan application flow in May 2018. (Id. at

8

¶¶ 39, 42.) Dr. Wind’s report explains that “[b]ecause both the Test and Control flows were

9

presented to actual prospective borrowers on LendingClub’s website, surveying based on a flow

10

that included the ‘No Hidden Fees’ claim or icon would have required reinserting the claim back

11
United States District Court
Northern District of California

1

onto the live website.” (Id. at 30 n.52.)
Test and Control group borrowers who completed their loan applications, were approved

12
13

for a loan, and received loan proceeds in their bank accounts then received an email from

14

LendingClub “with a link to a survey with a sweepstakes entry incentive.” (Id. at ¶ 44.) Over 600

15

borrowers in the Test group and nearly 600 borrowers in the Control group completed the survey,

16

constituting roughly 20% of each group. (Id. at ¶ 46.) The survey respondents were asked the

17

following questions to test their understanding of the origination fee:

18

Question: Was the amount of money deposited into your bank account
the amount that you expected after you obtained your loan approval?

19

o Yes

20
21
22
23
24
25
26
27
28

10

As explained in Dr. Wind’s report, the Test group “included borrowers who purchased a
personal loan through LendingClub’s platform after experiencing a version of the personal loan
application flow on LendingClub’s desktop and mobile websites that included the same
disclosures as the flow that is challenged in the Complaint, with the exception that the “No Hidden
Fees” claim and bandit icon were not included on the Offer page because LendingClub had
removed this claim and icon from its marketing collateral, website, and application flow in May
2018 in response to the Complaint.” (Dkt. No. 219-5, Ex. 6 at ¶ 39 (footnote omitted).) The
Control group “included borrowers who purchased a personal loan through LendingClub’s
platform after experiencing a modified version of the personal loan application flow on
LendingClub’s desktop and mobile websites that included alternative disclosures of the origination
fee and a description of the amount of cash the borrower would receive in their bank account at
loan disbursement. The presentation of this information was designed to comply with the
requirements specified by the FTC Staff in a draft consent order transmitted to LendingClub prior
to the filing of the Complaint.” (Id. (footnote omitted).)
19

o No

1

o I don’t know/I’m unsure

2

If the Respondent answered “No,” the following question was asked.

3

Question: How was the amount of money deposited in your bank
account different than what you expected?

4
5

o It was greater than what I expected

6

o It was less than what I expected

7

o I don’t know/I’m unsure

8

If the Respondent answered “It was less than what I expected,” the
following question was asked.

9

Question: Why was the amount of money deposited in your bank
account less than what you expected?

10

United States District Court
Northern District of California

11

o Please explain: ___________________

12

o I don’t know/I’m unsure

13

(Id. at ¶ 54.)11 Dr. Wind’s report notes that the initial survey results indicated “only a small

14

difference in understanding of the origination fee across Test and Control groups”; specifically,

15

88.9% of Test group respondents and 94.1% of Control group respondents answered “yes” to the

16

first question: “Was the amount of money deposited into your bank account the amount that you

17

expected after you obtained your loan approval?” (Id. at ¶¶ 58-59.)
Dr. Wind presented his findings to the FTC in November 2018, and the FTC raised two

18
19

issues “regarding the reliability of the results”: (1) “the absence of the ‘No Hidden Fees’ claim and

20

bandit icon from the Test flow [was] a material distinction between the Test flow and the

21

disclosures challenged in the Complaint,” and (2) respondents likely responded more favorably

22

because they knew that the survey was done by LendingClub and were not told that their

23

responses would not affect their relationship with LendingClub or their credit score.” (Id. at ¶¶

24

78-80.) Dr. Wind did not agree that those issues undermined the reliability of the initial survey

25

results, and noted, in pertinent part:

26
27
28

11

Survey respondents were also asked questions regarding their (1) “satisfaction with the loan
application process and willingness to recommend LendingClub,” and (2) “journey through the
application process” and prior “research into various loan options.” (Dkt. No. 219-5 at ¶¶ 55-56.)
20

1
2
3
4
5
6

There is evidence in the record that the removal of the “No Hidden
Fees” claim and bandit icon from LendingClub’s application flows in
May 2018 (an action that I understand the company took to remove
any question regarding the forward-looking compliance of its loan
application flow after the FTC filed the Complaint), did not materially
impact (i) the loan origination rates on LendingClub’s platform, (ii)
the application-to-loan conversion rates, or (iii) the estimated rates of
inquiry to LendingClub regarding the origination fee. Each of these
facts strongly suggest that the “No Hidden Fees” claim and bandit
icon are immaterial (or minimally material) to borrower
understanding of the origination fee.

7

(Id. at ¶ 81 (footnotes omitted).) As to the conversion rates before and after removal of the “No

8

Hidden Fees” representation, Dr. Wind cited evidence indicating that “LendingClub’s application-

9

to-loan conversion rate decreased from an average of 6.2% in June 2017 to April 2018 to an
average of 5.6% in June and July 2018.” (Id. at 55 n.107 (citing id. at 228).) In response to the

11
United States District Court
Northern District of California

10

FTC’s concerns with the initial survey, however, Dr. Wind “designed an expanded consumer

12

experiment structured to examine th[ose] issues, and to generally evaluate the repeatability of the

13

results of [the] initial experiment.” (Id. at ¶ 82.)

14

The expanded experiment, which was conducted in January 2019, utilized the same

15

methodology as the initial experiment; specifically: “(i) a Test/Control methodology with random

16

assignments of respondents to test and control groups, (ii) a real-world stimuli, (iii) a natural,

17

realistic setting, (iv) respondents who are actual consumers, (v) appropriate context and timing,

18

and (vi) a sufficiently large sample to permit rigorous statistical analysis.” (Id. at ¶ 83; see also id.

19

at 67 n.122.) However, to address the FTC’s concerns, Dr. Wind’s expanded experiment used

20

“three separate Test groups (each with distinct Offer pages) and a Control group identical to the

21

initial consumer experiment.” (Id. at ¶ 84.) “Test Group A” used both “an Offer page that

22

included the ‘No Hidden Fees’ claim and bandit icon,” and “origination fee disclosures that are

23

analogous to the disclosures challenged in the Complaint, which do not include a footer explaining

24

the origination fee LendingClub added to the Offer page in June 2017.” (Id.) “Test Group B”

25

used the same Offer page as Test Group A, and “origination fee disclosures that are analogous to

26

the disclosures LendingClub used from June 2017 to May 2018, including the loan information

27

footer that describes the origination fee.” (Id.) “Test Group C” used an Offer page that did not

28

include “the “No Hidden Fees” claim and bandit icon” or the footer, but otherwise included
21

1

“origination fee disclosures that are analogous to the disclosures LendingClub used from June

2

2017 to May 2018.”12 (Id.) As with the initial experiment, “[t]he only differences among the

3

three Test flows and the Control flows [were] on the Offer page.” (Id. at ¶ 86.)
The expanded experiment also “modified the survey invitation email for all flows to

4
5

emphasize that survey responses would not affect the borrower’s credit score or relationship with

6

LendingClub in any way.” (Id. at ¶ 89.) Of the borrowers who received the survey, 506

7

responded from Test A, 474 responded from Test B, 483 responded from Test C, and 523

8

responded from the Control group. (Id. at ¶ 90.) The expanded survey posed the same questions

9

as the initial survey. (Id. at ¶ 83.)
Dr. Wind’s report states that the results of the expanded survey “validate the reliability and

11
United States District Court
Northern District of California

10

results of the initial consumer experiment”; specifically, “[m]isunderstanding rates for the Control

12

groups in both experiments are virtually identical,” and the “[n]et misunderstanding rates for

13

similar Test flows across the experiments are very similar.” (Id. at ¶¶ 94-95.) Further, “[r]esults

14

from the expanded consumer experiment suggest there was no material borrower

15

misunderstanding resulting from LendingClub’s disclosures.” (Id. at ¶¶ 97-103.) The results

16

instead “suggest[ed] the ‘No Hidden Fees’ claim and bandit icon modestly increased the rate at

17

which borrowers understood the origination fee.” (Id. at ¶¶ 104-05.) As for the expanded survey

18

results regarding conversion rates, Dr. Wind’s report states:

19

A low level of misunderstanding is consistent with application-topurchase conversion rates across the various Test and Control groups
being generally similar. As [a graph demonstrating the “percentage of
overall potential applicants remaining at each stage”] shows, the
overall application-to-purchase conversion rates for the Test A
(NHF/No Footer), Test B (NHF/Footer), and Test C (No NHF/No
Footer) groups ranged from 18.9 percent to 19.7 percent; the
application-to-purchase conversion rate for the Control group was
18.8 percent. This suggests that the alternative disclosures in the
Control had very little impact on the total volume of loan originations
and corroborates my conclusion that only a very small percentage of

20
21
22
23
24
25
26
27
28

12

The description of Test Group C refers to the group as the “No [No Hidden Fees]/No Footer”
group; however, the description also states that the Test Group C Offer page “includ[ed] the loan
information footer that describes the origination fee.” (See Dkt. No. 219-5, Ex. 6 at ¶ 84.) This
appears to be erroneous. The report includes a screenshot of the “Test C (No NHF/No Footer)”
mobile loan application flow Offer page with a notation stating: “Footer does not include
information regarding the origination fee.” (Id. at 62.)
22

reasonable borrowers misunderstand the origination fee.

1
2

(Id. at ¶ 109.)
ii.

3

The FTC’s Motion to Exclude Dr. Wind’s Testimony

The Federal Rules of Evidence govern the admission of testimony by expert witnesses:

4
5

A witness who is qualified as an expert by knowledge, skill,
experience, training, or education may testify in the form of an
opinion or otherwise if: (a) the expert’s scientific, technical, or other
specialized knowledge will help the trier of fact to understand the
evidence or to determine a fact in issue; (b) the testimony is based on
sufficient facts or data; (c) the testimony is the product of reliable
principles and methods; and (d) the expert has reliably applied the
principles and methods to the facts of the case.

6
7
8
9

Fed. R. Evid. 702. In determining whether expert testimony is admissible, the trial court acts as a

11
United States District Court
Northern District of California

10

gatekeeper and “must ensure that any and all scientific testimony or evidence is not only relevant,

12

but reliable.” Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 589 (1993). However, “the

13

Daubert gatekeeping obligation is less pressing in connection with a bench trial,” because “the

14

‘gatekeeper’ and the trier of fact [are] one and the same.” Volk v. United States, 57 F. Supp. 2d

15

888, 896 n.5 (N.D. Cal. 1999); see also FTC v. BurnLounge, Inc., 753 F.3d 878, 888 (9th Cir.

16

2014) (“When we consider the admissibility of expert testimony, we are mindful that there is less

17

danger that a trial court will be unduly impressed by the expert’s testimony or opinion in a bench

18

trial.”). Ultimately, courts “are entitled to broad discretion when discharging their gatekeeping

19

function.” United States v. Hankey, 203 F.3d 1160, 1168 (9th Cir. 2000).
The FTC does not argue that Dr. Wind is unqualified. Indeed, Dr. Wind’s credentials and

20
21

experience indicate that he is well qualified to offer an expert opinion in this action.13 (See Dkt.

22

No. 219-5, Ex. 6 at 106-217.) Instead, the FTC seeks to exclude Dr. Wind’s testimony as

23

unreliable because the central question posed in both the initial and expanded surveys (“Was the

24

amount of money deposited in your bank account the amount that you expected after you obtained

25

your loan approval?”) did not directly test whether the respondent was aware of the origination fee

26
27
28

13

Dr. Wind is “the Lauder Professor Emeritus and Professor of Marketing at the Wharton School
of the University of Pennsylvania.” (Dkt. No. 219-5, Ex. 6 at ¶ 5.) He has taught at Wharton for
more than 50 years, and provided expert testimony on consumer research and marketing in over 20
cases since 2014. (Id. at ¶ 5, 219-221.)
23

1

and thus produced ambiguous results. The FTC also asserts that Dr. Wind’s survey produced

2

irrelevant results because it did not test the specific loan flow described in the Complaint nor take

3

into account the discontinued “no hidden fees” advertisements and targeted mailings that enticed

4

some consumers to apply for a loan during the time period relevant to this action. The FTC

5

further argues that Dr. Wind used a biased survey methodology. Finally, the FTC asserts that Dr.

6

Wind’s opinions “about whether consumers learned of the fee from sources other than Defendant”

7

should be excluded because they “are no more than speculation.” (Dkt. No. 175-1 at 9-17.)
LendingClub counters that the FTC’s motion to exclude Dr. Wind’s testimony is

9

essentially a critique of his survey methodologies, which goes to the weight of the evidence, not

10

its admissibility. The Court addresses the FTC’s argument in turn, and agrees that they bear on

11
United States District Court
Northern District of California

8

the probative value of Dr. Wind’s testimony and are thus not proper grounds for exclusion.

12
13

(A)

Reliability and relevance

“[I]ssues of methodology, survey design, reliability, the experience and reputation of the

14

expert, critique of conclusions, and the like go to the weight of the survey rather than its

15

admissibility.” Clicks Billiards, Inc. v. Sixshooters, Inc., 251 F.3d 1252, 1263 (9th Cir. 2001); see

16

also Fortune Dynamic, Inc. v. Victoria’s Secret Stores Brand Mgmt., Inc., 618 F.3d 1025, 1036

17

(9th Cir. 2010) (“[T]echnical inadequacies in a survey, including the format of the questions or the

18

manner in which it is taken, bear on the weight of the evidence, not its admissibility.”) (internal

19

quotation marks and citation omitted). The Ninth Circuit has “long held that survey evidence

20

should be admitted as long as [it is] conducted according to accepted principles and [is] relevant.”

21

Id. (alteration in original) (internal quotation marks and citation omitted); see also Townsend v.

22

Monster Beverage Corp., 303 F. Supp. 3d 1010, 1025 (C.D. Cal. 2018) (“The admissibility

23

threshold for survey evidence in the Ninth Circuit is notably low.”).

24

Here, Dr. Wind’s report thoroughly explains the scientific principles underlying his initial

25

and expanded experiments. (Dkt. No. 219-5, Ex. 6 at ¶¶ 23-35 (detailing the following “elements

26

of ideal design of a consumer experiment”: (1) “Use of Test/Control methodology with random

27

assignment of respondents to test and control groups”; (2) “Use of real-world stimuli”; (3) “Use of

28

natural, realistic setting”; (4) “Use of actual consumers”; (5) “Appropriate context and timing”;
24

1

and (6) “Use of a large sample size and rigorous statistical analysis”).) Thus, there is no indication

2

that Dr. Wind’s surveys were not conducted in accordance with accepted scientific principles.

3

As for the relevance of the survey results, one of the test groups in Dr. Wind’s expanded
survey saw the relevant representation (“No Hidden Fees”) as it appeared on the Offer page during

5

the time period at issue, and Dr. Wind surveyed the relevant audience—actual LendingClub

6

borrowers. Cf. Kwan Software Eng’g, Inc. v. Foray Techs., Inc., No. C 12-3762 SI, 2014 WL

7

572290, at *5 (N.D. Cal. Feb. 11, 2014) (rejecting survey because defendant did not show “that

8

any members of the survey [were] the people who would see the alleged misrepresentations” or

9

“whose decision to purchase the product could be influenced”). Further, the central question

10

posed to survey respondents (“Was the amount of money deposited in your bank account the

11
United States District Court
Northern District of California

4

amount that you expected after you obtained your loan approval?”) bears on whether the

12

representation was misleading because it provides insight into whether actual LendingClub

13

borrowers were aware of the origination fee. That Dr. Wind could not replicate the advertising

14

and exact loan application flow described in the Complaint does not render his survey results

15

irrelevant and inadmissible. Instead, such arguments concerning Dr. Wind’s methodology go to

16

the probative weight of the survey results, which the FTC can challenge through cross-

17

examination and rebuttal expert testimony. See Clicks Billiards, 251 F.3d at 1263.

18

(B)

“Blatantly biased” methodology

19

The FTC’s assertion that “[Dr.] Wind used biased, unreliable survey methodologies” is

20

similarly misguided. (See Dkt. No. 175-1 at 14.) The FTC argues that the survey “results arise

21

from a ‘blatantly biased’ methodology” because Dr. Wind: (1) asked “the worst kind of leading

22

question: a yes/no question asking for agreement with a reasonable-sounding proposition”; (2)

23

disclosed LendingClub as the survey sponsor; (3) failed to screen out respondents who learned of

24

the origination fee from sources other than LendingClub’s marketing and application, including

25

this lawsuit; and (4) “surveyed respondents days or sometimes more than a week after they

26

applied, even though the claims Wind was testing were in the application flow and Wind could

27

have conducted the survey right after they had applied.” (Id. at 14-16.) All of those issues

28

concern Dr. Wind’s methodology, not whether the initial and expanded surveys were conducted
25

1

according to accepted principles and produced relevant results. Again, the FTC can challenge Dr.

2

Wind’s methodology at trial.

3

And despite the FTC’s assertion, the purported issues of bias do not reflect “blatantly
biased” methodologies warranting exclusion. (See Dkt. No. 175-1 at 14 (citing Brighton

5

Collectibles, Inc. v. RK Texas Leather Mfg., 923 F. Supp. 2d 1245, 1257 (S.D. Cal. 2013)).) First,

6

Dr. Wind’s rebuttal report explains the rationale underlying the “yes/no/I don’t know” question

7

format used in his surveys and the benefits of that format in comparison to the “open-ended

8

question” asked by the FTC’s expert, which Dr. Wind asserts created biased results. (See Dkt. No.

9

221-8 at ¶¶ 22-31.) Such dueling assertions of biased methodology are properly resolved at trial.

10

The FTC’s citation to Dukes v. Wal-Mart, Inc., 222 F.R.D. 189 (N.D. Cal. 2004) and Sirko v. IBM

11
United States District Court
Northern District of California

4

Corp., No. CV 13-03192 DMG (SSx), 2014 WL 4452699 (C.D. Cal. Sept. 3, 2014) for the

12

proposition that Dr. Wind’s question format was biased is unavailing because the surveys at issue

13

in both cases were designed and conducted by the sponsoring party’s legal counsel, not a

14

designated expert. See Dukes, 222 F.R.D. at 197-98; Sirko, 2014 WL 4452699, at *4. The Dukes

15

and Sirko courts did not exclude the surveys at issue on Daubert grounds but rejected them as

16

unreliable hearsay. See Dukes, 222 F.R.D. at 198 (striking portion of expert’s report relying on

17

counsel’s survey); Sirko, 2014 WL 4452699, at *4 (finding survey inadmissible as hearsay).

18

Second, as to the disclosure of LendingClub as the survey sponsor in the survey emails,

19

Dr. Wind testified that disclosure of a survey sponsor results in bias only if “knowing the identity

20

will lead respondent to respond in a way that favors” the sponsor. (Dkt. No. 224-3 at 10 (153:8-

21

10.) Dr. Wind testified at his deposition that there was “no built-in bias” here because respondents

22

were merely responding to factual questions reflecting a “direct continuation [of] their interaction

23

with LendingClub,” with no “specific incentive to respond in a way that LendingClub may be

24

interested in.” (Id. at 10-11 (153:11-154:11).) That explanation seems reasonable; at any rate, the

25

FTC can challenge Dr. Wind’s testimony on this point at trial.

26

The FTC’s reliance on Gibson v. Cty. of Riverside, 181 F. Supp. 2d 1057 (C.D. Cal. 2002)

27

for the proposition that revealing a survey’s sponsor warrants exclusion is unpersuasive because

28

that case is easily distinguishable. In Gibson, as in Dukes and Sirko, “the survey was not
26

1

conducted by experts or independently of the attorneys involved in the litigation.” See Gibson,

2

181 F. Supp. 2d at 1068-69 (“[I]nstead of employing an expert, the County and its attorneys

3

designed and executed the survey.”). Further, in addition to revealing the defendant as the

4

survey’s sponsor, “many of the recipients of the [survey] had a vested interest in [the subject of the

5

litigation], and they were informed before filling out the survey that responding to the survey

6

could help in [that] effort.” Id. at 1069. The court accordingly found that the survey responses

7

were inadmissible hearsay. Id. None of the salient facts in Gibson are present here.

8
9

Third, as for screening respondents to ensure that they did not learn about the origination
fee from sources other than LendingClub’s marketing and loan application flow, it is unclear how
Dr. Wind could have done so given that the survey involved actual LendingClub borrowers, the

11
United States District Court
Northern District of California

10

assignment to Test and Control groups was random, and the survey was voluntary. Again, to the

12

extent the FTC asserts that failure to screen out applicants tainted the survey results, the FTC can

13

argue that point at trial.

14

Finally, as to the timeliness of the survey as it relates to borrowers’ memories of their loan

15

application experience, Dr. Wind’s report states that for the initial experiment, the average elapsed

16

time between application signature date and survey response was approximately 9.5 days, and the

17

average time between loan disbursement and survey response was approximately one week. (See

18

Dkt. No. 219-5, Ex. 6 at 255-56.) For the expanded experiment, the average elapsed time between

19

application signature date and survey response was approximately 5 days, and the average time

20

between loan disbursement and the survey response was less than 3.5 days. (Id. at 399-400.) The

21

FTC asserts that Dr. Wind “could have conducted the survey right after [borrowers] applied,” and

22

the “delay made it more likely that respondents who could not remember what they expected

23

versus what they received would answer ‘yes,’ particularly because Wind did not include a ‘do not

24

recall’ answer option.” (Dkt. No. 175-1 at 16-17.) The FTC’s argument is simply wrong as to

25

both points. First, Dr. Wind could not have asked borrowers if they received what they expected

26

“right after [they] applied” because the loan proceeds had not yet been disbursed. Second, Dr.

27

Wind’s surveys included an “I don’t know/I’m unsure” option, which is substantively the same as

28

“I do not recall.” At any rate, there is no indication that the lapse between applying for a loan,
27

1

receiving the loan proceeds, and responding to the survey constitutes a “blatantly biased”

2

methodology warranting exclusion.

3

The FTC’s citation to Brighton Collectibles, Inc. v. RK Texas Leather Mfg. for its

4

overarching argument that Dr. Wind’s methodology was “blatantly biased” fails to persuade. In

5

Brighton Collectibles the court concluded that the “survey’s design was so blatantly biased that

6

the results [were] unreliable.” 923 F. Supp. 2d at 1257 (citing Daubert, 509 U.S. at 589). The

7

survey in that trademark infringement action first showed participants four authentic handbags

8

manufactured by the plaintiff, two of which were black and red and all of which “had large heart

9

ornaments.” Id. Participants were next shown “four similar handbags made by other
manufacturers,” but “only the [d]efendants’ handbag was two-color (black and brownish-red) with

11
United States District Court
Northern District of California

10

heart decorations.” Id. “Participants were then asked which, if any, item was ‘made, sponsored,

12

or endorsed’ by the same company that made the first set of handbags.” Id. Unsurprisingly, an

13

overwhelming number of participants (89%) picked the defendants’ handbag. Id. The court

14

concluded that the survey did “not prove actual customer confusion about [plaintiff’s] brand, but

15

instead tested the ability of participants to pick the most obvious match.” Id. There can be no

16

reasonable argument that Dr. Wind’s survey is akin to the simplistic and obviously biased survey

17

rejected in Brighton Collectibles.
(C)

18
19

Speculative opinion regarding “consumer
journey”

Dr. Wind’s report includes a section that discusses “the consumer journey” a loan applicant

20

takes prior to applying for a loan, and details the importance of understanding the consumer

21

journey “to appropriately evaluate whether or how a particular communication is reviewed or

22

understood by consumers, and/or may influence their purchase decisions.” (Dkt. No. 219-5, Ex. 6

23

at ¶¶ 15-22.) The gist of the section is that consumers gain information in the course of

24

researching loans, and:

25
26
27
28

[t]hese experiences inform consumers about different loan options,
alert them to important loan terms, and help them evaluate alternative
loan options to make a choice that best fits their needs. Hence, even
before applying for a loan through LendingClub, a prospective
LendingClub customer may have formed various beliefs about online
loan products and/or loans available through LendingClub, which
may influence customers’ understanding of the origination fee in a
28

1

meaningful way.

2

(Id. at ¶ 17.) The initial and expanded surveys included the following questions relevant to Dr.

3

Wind’s “consumer journey” analysis:

4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16

Question: Prior to selecting a LendingClub loan, how much research
into various loan options did you do? Check all that apply.
o
o
o
o
o
o
o

I compared the interest rates and fees of at least one other
online lender
I used a loan comparison tool (e.g., Credit Karma)
I contacted a brick-and-mortar bank or credit union
I investigated various loan options over a period of two
weeks or more
I reviewed lender customer satisfaction ratings and
reviews
I reviewed explanations of APR, interest rates, and loan
fees on LendingClub’s website
I don’t recall / I am unsure

Question: About how much time did you spend researching before
you decided to apply for a loan through LendingClub?
o
o
o
o
o
o

Less than 1 hour
1 to 3 hours
3 to 5 hours
5 to 7 hours
More than 7 hours
I don’t recall / I am unsure

17

(Dkt. No. 219-6, Ex. 6 at ¶¶ 20, 56, 83.) Dr. Wind’s report states that in both the initial and

18

expanded surveys, “more than 87 percent of respondents had conducted some form of research

19

into various loan options prior to selecting a loan through LendingClub.” (Id. at ¶ 21.) As for the

20

two answers that most directly bear on knowledge of the origination fee, in both the initial and

21

expanded surveys over 33% of respondents “compared the interest rates and fees of at least one

22

other online lender,” and roughly 29% of respondents “reviewed explanations of APR, interest

23

rates, and loan fees on LendingClub’s website.” (Id. at ¶ 21, Tables 1-2.)

24

The FTC argues that Dr. Wind’s opinions “that consumers learned of the fee from

25

independent research” should be excluded because they are “mere speculation.” (Dkt. No. 175-1

26

at 17.) LendingClub counters that Dr. Wind’s opinion regarding the “consumer journey” an

27

applicant takes prior to applying for a loan with LendingClub is not speculative but is instead

28

based on the analysis detailed in his report and the survey questionnaire asking respondents
29

1

whether they conducted research prior to applying for a LendingClub loan. (Dkt. No. 221-4 at 16-

2

17.) The Court agrees. Dr. Wind’s “consumer journey” analysis and the survey questionnaire

3

adequately support his opinion that pre-application information may inform a consumer’s

4

knowledge about the origination fee, such that the opinion is “more than subjective belief or

5

unsupported speculation.” See Daubert, 509 U.S. at 590 (noting that “knowledge” as used in Rule

6

702 “connotes more than subjective belief or unsupported speculation”). As with Dr. Wind’s

7

other testimony, the FTC can challenge the bases of his opinions at trial.

8
9

Accordingly, the Court denies the FTC’s motion to exclude Dr. Wind’s testimony because
it meets the standard for expert testimony under Rule 702, the FTC’s challenges go primarily to
the weight of Dr. Wind’s testimony and not its admissibility, and the Court’s “gatekeeper”

11
United States District Court
Northern District of California

10

function is “less pressing” because it is also the trier-of-fact. See Volk, 57 F. Supp. 2d at 896 n.5.

12
13

iii.

Conversion rate evidence

Having determined that Dr. Wind’s testimony is admissible, the conversion rate evidence

14

gives rise to a further genuine dispute of material fact. Both parties address this argument

15

primarily in terms of the third element of a Section 5 claim—materiality; however, the dispute

16

also concerns deception. Evidence that there was no material change in the conversion rate

17

between April 2018 (before LendingClub removed the “No Hidden Fees” representation from the

18

loan application flow in May 2018) and after LendingClub implemented the FTC’s suggested

19

origination fee disclosure on the Offer page in February 2019 gives rise to a reasonable inference

20

that consumers were aware of the origination fee even when the “no hidden fees” representation

21

was made and thus supports an inference that the net impression was not misleading.

22

The FTC asserts that the evidence is “immaterial because Defendant makes no attempt to

23

control for other factors that affect [conversion] rates, including market conditions; it thus proves

24

nothing about the effect of the representation itself.” (Dkt. No. 239-5 at 13.) The FTC may

25

ultimately be correct; however, that argument goes to the weight of the conversion rate evidence.

26

At this stage and viewing the record in the light most favorable to LendingClub, it is enough that

27

the evidence gives rise to a reasonable inference that consumers were aware of the origination fee

28

and thus the net impression of the “no hidden fees” representation was not misleading.
30

c.

1

Extrinsic evidence may be considered

2

The FTC suggests in its reply that the Court should not consider LendingClub’s extrinsic

3

evidence and instead rule based solely on a facial review of the loan application flow. The cases

4

the FTC cites do not persuade the Court that it can do so. In FTC v. AMG Capital Mgmt, LLC, the

5

court addressed the defendant’s extrinsic evidence—specifically, expert testimony—and

6

determined that the evidence failed to raise a genuine dispute of material fact as to deception. Id.

7

at 425-26. The same for Cyberspace.Com where the court considered the defendant’s “consumer

8

research study,” but found that it did not raise a genuine dispute of fact. 453 F.3d at 1201. Thus,

9

AMG and Cyberspace.Com do not stand for the far-reaching proposition that extrinsic evidence is
immaterial once the FTC satisfies its initial burden on summary judgment of demonstrating

11
United States District Court
Northern District of California

10

deception.

12

***

13

In sum, drawing all reasonable inferences in LendingClub’s favor, the Court cannot

14

conclude that every reasonable trier of fact would find that the net impression of the “no hidden

15

fees” representation was likely to mislead a reasonable consumer. Accordingly, the Court

16

DENIES the FTC’s motion for summary judgment on Count I.

17
18

2.

Count II

Count II alleges that LendingClub’s past practice of having “represented, directly or

19

indirectly, expressly or by implication, that . . . consumers would receive loans” constitutes a

20

deceptive act or practice prohibited by Section 5 of the FTC Act because “in numerous instances”

21

LendingClub ultimately denied loans to consumers after making those representations. (Dkt. No.

22

57 at ¶¶ 59-61.) To establish this claim the FTC must prove the same three elements required for

23

Count I; specifically: (1) “there is a representation, omission, or practice that,” (2) “is likely to

24

mislead consumers acting reasonably under the circumstances, and” (3) “the representation,

25

omission, or practice is material.” Stefanchik, 559 F.3d at 928 (internal quotation marks and

26

citation omitted). Accordingly, the same legal standard as to those elements applies: “Deception

27

may be found based on the ‘net impression’ created by a representation.” Id. (quoting

28

Cyberspace.Com, 453 F.3d at 1200).
31

1

The gravamen of Count II is that LendingClub represented to consumers that their

2

applications had been or would be approved, even though the loan applications had not received

3

final, or “back-end” approval, and some consumers’ applications were ultimately denied. There

4

are several undisputed representations underlying this claim:

5

(1) Statements from the loan application flow appearing after
applicants consent to a credit check and provide their social security
numbers “such as ‘Congratulations! You qualify for a loan’ and ‘Your
money will be deposited in this account.’” (Dkt. No. 145-4 at ¶ 79.)

6
7

(2) The statement at the top of the pre-November 2017 Account
Summary page: “Your $[amount requested] loan is on the way.
What’s next?” (Id. at ¶ 80.)

8
9

(3) The “progress bar” below the “loan is on the way” representation,
“including a green checkmark next to the word ‘FUNDING’ and the
statements ‘Investors are backing your loan,’ ‘DONE!’ and ‘Your
funds will be deposited into your bank account.’” (Id. at ¶ 81.)

10

United States District Court
Northern District of California

11
12
13

(4) Versions of the email informing applicants that their loans were
“100% Backed” sent between April 2015 and October 2018. (Id. at
¶¶ 87-93.)

14

The Court must first determine whether the FTC has satisfied its initial burden on

15

summary judgment of showing that LendingClub made the challenged representations, that “no

16

reasonable factfinder could conclude that the solicitation was not likely to deceive consumers

17

acting reasonably under the circumstances,” and that the representations were material. See

18

Cyberspace.Com, 453 F.3d at 1201.

19

//

20

//

21

//

22

//

23

//

24

//

25

//

26

//

27

//

28

//
32

1

Prior to November 2017, the Account Summary Page appeared as follows:

2
3
4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16

(Dkt. No. 220-5, Ex. 26.) Thus, on the last page of the loan application flow applicants see

17

prominent, bold text at the top of the page informing them that their “loan is on the way.” Below

18

that, applicants see the progress bar, which indicates “FUNDING” has been secured based on the

19

statement that “Investors are backing your loan.” To the right of “FUNDING,” applicants see

20

“FINAL REVIEW” in bold text above the statement: “You finish your To-Do list and we review

21

your application.” To the right of FINAL REVIEW, applicants see “DONE!” above the

22

statement: “Your funds will be deposited into your bank account.” Below the progress bar,

23

applicants see in smaller print their “TO-DO LIST,” above which is a smaller print statement

24

directing applicants to: “Please complete all tasks on your To-Do List. Once complete, we can

25

finish the final review of your application.” The bottom of the page includes a “Loan Number.”

26

(See Dkt. No. 251-1 at 32.)

27
28

At best, a consumer acting reasonably under the circumstances would likely be confused
by the prominent statement informing applicants that their loans were “on the way” before
33

1

LendingClub had completed final review of the application. Further, a consumer acting

2

reasonably under the circumstances would likely interpret the progress bar as indicating that her

3

loan had already received funding and was backed by investors because the “FUNDING” portion

4

includes a green checkmark and is not highlighted, suggesting that applicants have moved past

5

that stage of the process.

6

LendingClub contends that the references to the To-Do List” and the “Final Review”

7

“plainly communicate that the applicant must take additional steps to complete their application.”

8

(Dkt. No. 218-4 at 31.) Not so. Given that the consumer has already been told that their money is

9

on the way, their loan has been funded, and there is even a loan number, drawing all reasonable
inferences in LendingClub’s favor, at most those references communicate that before the

11
United States District Court
Northern District of California

10

consumer will receive her already-approved loan there are additional tasks the consumer must

12

complete. And the nature of those tasks reinforces the net impression that the consumer’s loan has

13

been approved and is on its way: the record reflects that in some cases those tasks consisted of

14

confirming the consumer’s email address or bank account number. (Dkt. No. 145-4 at ¶ 82.)

15

Neither of those tasks suggest to a reasonable consumer that LendingClub meant that the

16

consumer’s money is on the way only after and if LendingClub approves the loan. Similarly, that

17

the page suggests that applicants must proceed to “FINAL REVIEW” does not negate the net

18

impression that the loan has been approved and funded. There is nothing that communicates that

19

“final review” means anything more than verifying email and bank account information so that the

20

consumer can receive the already-approved loan. Because it is undisputed that LendingClub did

21

not fund loans at the Account Summary stage and that an applicant’s loan was not “on the way,”

22

the net impression of the Account Summary page was likely to mislead reasonable consumers.

23

This misleading net impression was compounded by at least one email Lending Club sent

24

consumers at the same time the consumers were presented with the pre-November 2017 version of

25

the Account Summary page. This version of the email (“Version 1”) “was sent to approximately

26

196,000 consumers, and featured the subject line ‘Hooray! Investors Have Backed Your Loan’

27

(June 2015 through July 2015) or ‘Investors Have Backed Your Loan! Here Are the Next Steps’

28

(September 2015 through November 2015).” The body of the email appeared as follows:
34

1
2
3
4
5
6
7
8
9

(Dkt. No. 177-1, Ex. 128 at 6-7, 10.) The email telling consumers that their money “was almost in
their hands” that their loan “is 100% backed,” and welcoming them to Lending Club reinforced

11
United States District Court
Northern District of California

10

the net impression of the Account Summary page that the consumer’s loan had been approved.

12

This impression was misleading because LendingClub ultimately denied loans to “approximately

13

43,000 consumers who received [this] email.” (Dkt. No. 145-4 at ¶ 99.)

14

Finally, while not required, the FTC submits undisputed evidence that applicants were

15

actually misled by the challenged representations. See Cyberspace.Com, 453 F.3d at 1201 (noting

16

that evidence of actual deception is not required but is “highly probative to show that a practice is

17

likely to mislead consumers acting reasonably under the circumstances”). Consumers who were

18

subsequently denied loans complained that they thought they were already approved based on the

19

Account Summary page’s “on the way” statement. (Dkt. No. 145-4 at ¶¶ 80, 101 (citing Dkt. Nos.

20

150-7, Ex. 7 at 37, 49 & 190-21, Ex. 247 at 1 (filed under seal)).) Consumers also complained

21

about the misleading impression of the progress bar. (Id. at ¶ 81 (citing Dkt. Nos. 150-7, Ex. 7 &

22

190-21, Ex. 247).) In some cases, consumers did not pursue loans with other lenders, withdrew

23

their loan applications with other lenders, or made specific purchases based on the representations.

24

(Id. at ¶ 104 (citing Dkt. Nos. 150-7, Ex. 7 & 190-21, Ex. 247).)

25

LendingClub internal communications also recognized that the “on the way” statement was

26

misleading. An internal LendingClub 2017 presentation characterized the “on the way” statement

27

as “confusing and misleading.” (Id. at ¶ 83 (quoting Dkt. No. 176-2, Ex. 125 at 11 (emphasis in

28

original)).) In a September 2017 email with the subject line “‘Your loan is on the way!’
35

1

messaging,” LendingClub’s President described the message to LendingClub’s head of marketing,

2

Bill Walsh, as creating a “recurring issue” for “front line team members,” and stated that the

3

message “feels confusing to customers who assume the loan application process is complete and

4

the funds are already in transit.” (Id. at ¶ 84 (citing Dkt. No. 156-7, Ex. 29 at 2 (filed under

5

seal)).) Another September 2017 internal email characterized the “on the way” statement as

6

“fundamentally wrong (i.e. Your $10,000 loan is on the way. What’s next? . . . err not, really).”

7

(Dkt. No. 176-1, Ex. 124 at 4 (emphasis in original).)
Further, it is undisputed that LendingClub conducted a focus group with customer service

9

agents in April 2017 wherein “[p]articipants stated the progress bar and messaging within the To-

10

Do List set unrealistic expectations for customers, leading many to believe they had completed all

11
United States District Court
Northern District of California

8

necessary steps in their loan applications after signing the [TILA disclosure].” Participants noted:

12

In particular, the heading “Your [$X,XXX] loan is on the way. What’s
next? gave many customers the impression they did not need to take
further action.

13
14
15
16

Many were also confused by the language and visual treatment of the
progress bar accompanying the To-Do List, namely that all three
stages—Funding, Final Review, and Done—are displayed, with
“Funding: Investors are backing your loan” shown with the green
check mark.

17

(Dkt. No. 159-4, Ex. 35 at 3 (filed under seal).) The participants recommended, in pertinent part:

18

(1) “Do not state ‘Your loan is on the way’ before customers have completed the To-Do List”; and

19

(2) “Change the visual treatment of the progress bar so it does not mislead applicants by

20

displaying ‘Funding’ and ‘Done.’” (Id. at 4.) In 2015 an outside auditor also characterized the

21

“on the way” statement as “‘misleading at best.’” (Dkt. No. 145-4 at ¶ 86.)

22

Evidence that consumers, LendingClub employees, and third-party auditors found the pre-

23

November 2017 Account Summary page misleading, coupled with the at best confusing and at

24

worst misleading net impression of the Account Summary page itself, is sufficient to satisfy the

25

FTC’s initial burden on summary judgment. LendingClub made the challenged representations,

26

the representations were likely to mislead, and they were material because they concerned whether

27

a consumer’s loan was actually approved. See Cyberspace.Com, 453 F.3d at 1201 (“A misleading

28

impression created by a solicitation is material if it involves information that is important to
36

1

consumers and, hence, likely to affect their choice of, or conduct regarding, a product.”) (internal

2

quotation marks and citation omitted). The burden thus shifts to LendingClub to identify evidence

3

that creates a genuine issue of material fact.
LendingClub asserts that material questions of fact exist regarding the net impression of

5

the challenged representations because, “with the exception of [Version 1],” the representations

6

contain “clear and prominent statements that loan approval was not certain and that the borrower

7

must complete the remaining tasks on their ‘To-Do List.’” (Dkt. No. 251-1 at 35-36.) The Court

8

disagrees as to the pre-November 2017 Account Summary page for the reasons stated above. The

9

page’s truthful disclosures regarding the “To-Do List” did not cure the overall misleading net

10

impression that the consumers’ loans had already secured funding and were “on the way.” See

11
United States District Court
Northern District of California

4

Cyberspace.Com, 453 F.3d at 1200 (“A solicitation may be likely to mislead by virtue of the net

12

impression it creates even though the solicitation also contains truthful disclosures.”)). That the

13

pre-November 2017 Account Summary page was likely to mislead is bolstered by the undisputed

14

extrinsic evidence demonstrating that (1) consumers were actually misled and complained about

15

the specific representations contained on the page, and (2) LendingClub’s internal

16

communications recognized that, as to the “on the way” representation, the statement was

17

“confusing,” “misleading,” and “fundamentally wrong.” Thus, no reasonable factfinder could find

18

that the net impression of the representations were not at least likely to mislead.

19

LendingClub’s citation to its extrinsic evidence demonstrating “low complaint and inquiry

20

rates” also fails to raise a genuine dispute of material fact. First, the FTC does not have to show

21

any actual consumer confusion to support a finding that the net impression of its representations

22

was likely to deceive. That there is actual evidence that only a small percentage complained thus

23

does not support an inference that consumers were not likely to be deceived; especially given that

24

complaints do not reflect the “total universe of injury.” See FTC v. Amazon.com, Inc., No. C14-

25

1038-JCC, 2016 WL 10654030, at *8 (W.D. Wash. July 22, 2016) (rejecting similar argument as a

26

defense to a Section 5 unfair practices claim because it “conflates complaints with the total

27

universe of injury”). Second, it is undisputed that consumers complained about the

28

representations in large enough numbers that LendingClub recognized the problem and addressed
37

1
2

it internally through training materials, customer service focus groups, and other communications.
LendingClub also argues that “likelihood of recurrence” is another issue of material fact

3

precluding summary judgment because it ceased the challenged conduct and the FTC seeks only

4

injunctive relief under Count II. To obtain injunctive relief under Section 5 for past conduct, the

5

FTC must demonstrate that such conduct is “likely to recur.” See FTC v. Evans Prods. Co., 775

6

F.2d 1084, 1088-89 (9th Cir. 1985); see also Gill, 71 F. Supp. 2d at 1047 (noting that a permanent

7

injunction under the FTC Act requires a showing of “some cognizable danger of recurring

8

violation”) (citing United States v. W.T. Grant, 345 U.S. 629, 633 (1953)). In determining

9

whether the challenged conduct is likely to recur courts necessarily look to past conduct and a

10

United States District Court
Northern District of California

11
12
13
14

variety of factors, including:
the degree of scienter involved, whether the violative act was isolated
or recurrent, whether the defendant’s current occupation positions
him to commit future violations, the degree of harm consumers
suffered from the unlawful conduct, and the defendant’s recognition
of his own culpability and sincerity of his assurances, if any, against
future violations.

15

FTC v. Commerce Planet, Inc., 878 F. Supp. 2d 1048, 1086 (C.D. Cal. 2012), aff’d in part and

16

vacated in part on other grounds by 815 F.3d 953 (9th Cir. 2016). While those factors could

17

support a finding that there is at least a likelihood of recurrence based on the summary judgment

18

record, the FTC has not shown that is the only finding that is supported by the record. To make

19

that finding the Court would have to weigh the evidence and draw inferences in FTC’s favor.

20

Accordingly, whether an injunction is warranted has to be decided at trial when the Court can

21

weigh the evidence and make findings.

22

***

23

The FTC has satisfied its burden on summary judgment of demonstrating the absence of a

24

genuine issue of material fact as to whether the pre-November 2017 Account Summary page and

25

Version 1 of the “100% Backed” email were likely to mislead a reasonable consumer, and

26

LendingClub fails to raise a genuine dispute of material fact on that score. Accordingly, the Court

27

GRANTS IN PART the FTC’s motion for summary judgment on Count II as to those

28

representations; the Court leaves to trial whether an injunction is warranted because of a likelihood
38

1

of recurrence. See Fed. R. Civ. P. 56(a),(g).

2

B.

Section 5 Unfair Practice (Count III)

3

Count III of the FAC alleges that LendingClub has made numerous unauthorized

4

withdrawals from borrowers’ bank accounts through its default payment method of ACH transfer,

5

causing substantial and unavoidable injury to consumers. (Dkt. No. 57 at ¶¶ 62-64.) The FTC Act

6

defines an actionable unfair practice or act as one that “‘causes or is likely to cause substantial

7

injury to consumers which is not reasonably avoidable by consumers themselves and not

8

outweighed by countervailing benefits to consumers or to competition.’” FTC v. Neovi, Inc., 604

9

F.3d 1150, 1155 (9th Cir. 2010) (quoting 15 U.S.C. § 45(n)).

10

The FTC has produced evidence sufficient to meet its initial summary judgment burden on

United States District Court
Northern District of California

11

each required element. LendingClub, however, has also produced affirmative evidence of its

12

unauthorized withdrawal monitoring program for its ACH payment system, its general practice of

13

providing refunds to consumers who complain of erroneous withdrawals, and an error rate that is

14

within the guidelines set forth by the National Automated Clearing House Association

15

(“NACHA”), “the organization that sets standards for the ACH Network and regulates automated

16

withdrawals made using the Network.” (See Dkt. No. 140-4 at 11-14 (filed under seal).) That

17

evidence is material to determining whether the unauthorized withdrawals at issue constitute an

18

“unfair practice,” or if they were instead the result of inadvertent human and automated error

19

inherent in any ACH payment system that services millions of consumers.

20

A genuine dispute of material fact exists as to the efficacy of LendingClub’s ACH

21

payment system as it relates to unauthorized withdrawals, precluding summary judgment on this

22

count. Indeed, the FTC’s opposition to LendingClub’s motion for summary judgment recognizes

23

this dispute. (See Dkt. No. 215-3 at 8 (“Defendant states that the conclusion that is has ‘prudently

24

managed and monitored’ its use of the ACH Network is not in dispute. LC MSJ 6:10-7:20. The

25

parties do dispute this conclusion, which Defendant supports principally by citing to its expert and

26

its own testimony.”). “By definition, summary judgment may be granted only where there are no

27

disputed issues of material fact, and thus no factfinding by the district court,” even if the case is to

28

be tried to the court and not a jury. Animal Legal Defense Fund v. U.S. Food & Drug Admin., 836
39

1

F.3d 987, 989 (9th Cir. 2016) (en banc). Each party’s motion asks the Court to draw inferences

2

and weigh evidence in the moving party’s favor. As the Court cannot do so on summary

3

judgment, the Court denies the parties’ cross-motions on summary judgment as to Count III.

4

C.

Count IV

5

Count IV of the FAC alleges that LendingClub violated the GLB Act, 15 U.S.C. §§ 6801-

6

03, and its implementing regulations: the Privacy Rule, 16 C.F.R. § 313, and Regulation P, 12

7

C.F.R. § 1016. (Dkt. No. 57 at ¶¶ 65-67.) The FTC has statutory authority to enforce the GLB

8

Act under the FTC Act, pursuant to 15 U.S.C. § 6805(a)(7). The Privacy Rule and Regulation P

9

require financial institutions to disclose their privacy policies to consumers. 16 C.F.R. § 313.6; 12
C.F.R. § 1016.6. A Financial institution must provide “clear and conspicuous” notice of its

11
United States District Court
Northern District of California

10

privacy policy to consumers when the customer relationship is first established. 15 U.S.C. §

12

6803(a); 16 C.F.R. § 313.4; 12 C.F.R. § 1016.4(a). The initial notice must be provided in such a

13

manner “that each consumer can reasonably be expected to receive actual notice.” 16 C.F.R. §

14

313.9(a); 12 C.F.R. § 1016.9(a). Financial institutions that conduct transactions electronically

15

“may reasonably expect that a consumer will receive actual notice” if they “clearly and

16

conspicuously post the notice on the electronic site and require the consumer to acknowledge

17

receipt of the notice as a necessary step to obtaining a particular financial product or service.” 16

18

C.F.R. § 313.9(b)(1)(iii); 12 C.F.R. § 1016.9(b)(1)(iii).
There is no dispute that LendingClub is a financial institution subject to the GLB Act and

19
20

the requirements of the Privacy Rule and Regulation P because it “collects nonpublic personal

21

information,” as defined by [the Privacy Rule and Regulation P], such as Social Security numbers

22

and bank routing information.” (Dkt. No. 57 at ¶ 48 & Dkt. No. 81 at ¶ 48.) There is also no

23

dispute that prior to December 2016 LendingClub did not require loan applicants to separately

24

acknowledge receipt of its Privacy Policy prior to obtaining a loan. (Dkt. Nos. 57 at ¶ 51 & 81 at

25

¶ 11.)

26

The FTC asserts that summary judgment on this count is warranted because LendingClub

27

failed to satisfy the notice requirements under the GLB Act prior to December 2016.

28

LendingClub moves for summary judgment on this count because the FTC seeks only injunctive
40

1

relief, and LendingClub ceased the challenged conduct in late 2016 and there is no evidence that

2

LendingClub “is likely to revert to its prior practices.” (Dkt. No. 140-4 at 28.) The FTC counters

3

that there is no law “suggesting that arguments against a plaintiff’s requested relief, if successful,

4

entitle the defendant to summary judgment on liability, which is what Defendant’s motion is

5

seeking.” (Dkt. No. 215-3 at 28 (filed under seal).) The Court does not read LendingClub’s

6

motion as expansively. LendingClub is not seeking summary judgment on the merits of Count IV;

7

instead, it argues that Count IV fails because it seeks only injunctive relief for past conduct and the

8

FTC fails to show a cognizable danger that LendingClub will revert to its pre-December 2016

9

practices. Accordingly, the Court considers LendingClub’s motion for summary judgment on

10

United States District Court
Northern District of California

11

Count IV only as to the relief sought.
As previously discussed, the FTC enforces the GLB Act through the FTC Act. See 15

12

U.S.C. § 6805(a)(7). Thus, the same standard for permanent injunctive relief arising out of past

13

conduct applies; specifically, the FTC must show a likelihood or “cognizable danger” of recurring

14

violation. See, e.g., Evans, 775 F.2d at 1088-89; Gill, 71 F. Supp. 2d at 1047 (citing United States

15

v. W.T. Grant, 345 U.S. 629, 633 (1953)). The Court must determine whether such danger exists

16

based on “the totality of the circumstances” and “appropriate findings supported by the record.”

17

Amazon.com, Inc., 2016 WL 10654030, at *5 (internal quotation marks and citation omitted).

18

In considering the relevant factors regarding LendingClub’s past conduct, see Commerce

19

Planet, 878 F. Supp. 2d at 1086, the FTC has not identified facts sufficient to support a finding of

20

a reasonable likelihood or cognizable danger of recurrence. The FTC asserts that LendingClub

21

ignored warnings from its internal compliance group and only changed its Privacy Policy after the

22

FTC issued the CID in May 2016. It is undisputed that LendingClub’s compliance group issued

23

compliance reviews in October and December 2015 and June 2016 that recommended changing

24

the Privacy Policy disclosures to comply with the GLB Act. (See Dkt. Nos. 190-14, Ex. 240; 190-

25

15, Ex. 241; 190-17, Ex. 243 (filed under seal).) In pertinent part, the June 2016 review stated that

26

“[c]ompliance identified several best practices, or ‘safe harbor’ provisions of Regulation P, that

27

could be utilized to enhance LendingClub’s Privacy Policy for customer accessibility and clarity.”

28

(Dkt. No. 190-17, Ex. 243 at 5.) The review recommended “that LendingClub undergo a Privacy
41

1

Policy rewrite led by Compliance or Legal (whoever is determined to have accountable

2

ownership) with the assistance of the Technical Writing team and the advice of relevant

3

stakeholders.” (Id.)

4

In October 2016, members of the compliance group sent two internal emails regarding the

5

Privacy Policy. The first discusses a “compliance Christmas Wish List” and notes that the then-

6

Privacy Policy “is not ideal for Reg P purposes, and the preferred [(i.e., recommended)] method

7

provides us safe harbor.” (See Dkt. No. 160-5, Ex. 47 at 2-3 (filed under seal).) The second

8

October 2016 internal email states, in pertinent part: “Compliance did a Privacy review and they

9

want us to include the Privacy Policy as its own agreement . . . . Everyone else does this so I don’t
think we can avoid it much longer.” (Dkt. No. 216-9, Ex. 257 at 2.) In December 2016,

11
United States District Court
Northern District of California

10

LendingClub implemented the recommended GLB Act-compliant Privacy Policy and has

12

maintained that policy ever since. The FTC filed suit over 15 months later.

13

The internal compliance reviews and emails show that LendingClub knew about the

14

requirements under the GLB Act and that its then-current Privacy Policy did not satisfy those

15

requirements. However, the evidence does not suggest a level of scienter sufficient to support a

16

finding of a likelihood of recurrence; specifically, that LendingClub sought to maintain its non-

17

compliant policy for some improper purpose. Further, the post-CID compliance review and

18

October 2016 emails do not demonstrate that LendingClub implemented the current Privacy

19

Policy only in response to the CID; indeed, neither email even mentions the FTC investigation.

20

The cited evidence instead reflects a continuing internal discussion of how to revise the policy to

21

comport with the GLB Act. It is of course reasonable to infer that the CID influenced

22

LendingClub’s decision to revise the Privacy Policy (and in deciding LendingClub’s motion the

23

Court must draw that inference), but that evidence does not support a finding that it was the

24

determining factor in light of the totality of the record. Finally, and more importantly,

25

LendingClub implemented the current Privacy Policy nearly 16 months before the FTC filed the

26

instant action. Thus, LendingClub’s conduct in changing its Privacy Policy does not evince an

27

attempt to evade impending liability under the GLB Act.

28

The FTC also does not cite any evidence showing complaints from consumers regarding
42

1

the non-compliant policy during the October 2015 to December 2016 timeframe, or that

2

LendingClub benefited from that policy. And although LendingClub is of course in a position to

3

commit future violations, the record does not support an inference that they are likely to do so

4

given that the current Privacy Policy has been in place for years, reverting to a non-compliant

5

policy will subject them to liability, and there is otherwise no reasonable business incentive for

6

them to change their current policy. Simply put, there is no showing that LendingClub is likely to

7

revert to its former non-compliant Privacy Policy after three and a half years of its current policy.
The FTC cites the Court’s October 2018 Order on LendingClub’s motion to dismiss for the

8
9

proposition that the Court “previously ruled that it . . . would only consider likelihood of
recurrence—with regard to the GLB Act—in connection with a mootness defense, and held that

11
United States District Court
Northern District of California

10

LendingClub “bears the burden of showing mootness.” (Dkt. No. 215-3 at 27-28 (citing Dkt. No.

12

53 at 24 n.10).) The Court declines the invitation to apply the same analysis at this stage because

13

its October 2018 Order was specifically addressing LendingClub’s argument that Count IV was

14

“moot” and that the FTC otherwise lacked statutory authority to pursue injunctive relief. Here, in

15

considering LendingClub’s motion for partial summary judgment as to injunctive relief, the

16

Court’s inquiry “hinges on whether the FTC has established, with evidence, a cognizable danger of

17

a recurring violation: not whether [LendingClub] has met a mootness burden.” See Amazon.com,

18

2016 WL 10654030, at *5 (emphasis added). The FTC fails to carry its burden for the reasons

19

discussed.
Accordingly, the Court GRANTS LendingClub’s motion for summary judgment on Count

20
21

IV as to the relief sought. Further, because it is undisputed that injunctive relief is the only

22

remedy available under Count IV, the Court DISMISSES this claim as moot because the Court

23

cannot provide “any effective relief.” See Nw. Envtl. Defense Ctr. v. Gordon, 849 F.2d 1241,

24

1244-45 (9th Cir. 1988) (noting that a claim is moot in the absence of “any effective relief”)

25

(emphasis omitted).

26

II.

27
28

The FTC’s Motion for Partial Judgment on the Pleadings
Under Federal Rule of Civil Procedure 12(c), “[a]fter the pleadings are closed—but early

enough not to delay trial—a party may move for judgment on the pleadings.” A court considering
43

1

a Rule 12(c) motion must accept the nonmovant’s allegations as true and construe the pleadings

2

“in the light most favorable” to the nonmovant. Gen’l Conference Corp. of Seventh-Day

3

Adventists v. Seventh-Day Adventist Congregational Church, 887 F.2d 228, 230 (9th Cir. 1989).

4

“Judgment on the pleadings is properly granted when there is no issue of material fact in dispute,

5

and the moving party is entitled to judgment as a matter of law.” Fleming v. Pickard, 581 F.3d

6

922, 925 (9th Cir. 2009). “If the motion for judgment on the pleadings is based on a failure to

7

state a claim upon which relief can be granted, the standard is the same as for a motion to dismiss

8

under Rule 12(b)(6).” Pantastico v. Dep’t of Educ., 406 F. Supp. 3d 865, 877 (D. Haw. 2019)

9

(citing McGlinchy v. Shell Chem. Co., 845 F.2d 802, 810 (9th Cir. 1988); Cafasso, U.S. ex rel. v.
General Dynamics C4 Systems, Inc., 637 F.3d 1047, 1054 n.4 (9th Cir. 2011)). In considering

11
United States District Court
Northern District of California

10

a Rule 12(c) motion, a court must limit its review to the complaint and attachments to the

12

complaint, documents incorporated by reference, and “facts that are contained in materials of

13

which the court may take judicial notice.” See Heliotrope Gen., Inc. v. Ford Motor Co., 189 F.3d

14

971, 981 n.18 (9th Cir. 1999).

15

The FTC moves for partial judgment on the pleadings as to LendingClub’s affirmative

16

defenses asserted in its Amended Answer to the FAC; specifically, the “first, third, fourth, fifth,

17

sixth, seventh, and ninth affirmative defenses.” (Dkt. No. 139 at 2.) LendingClub’s opposition

18

withdraws its fifth, sixth, seventh, and ninth affirmative defenses. (Dkt. No. 201-7, Ex. 7 at 7 n.2

19

(filed under seal).) Thus, the Court addresses the FTC’s motion only as to the first, third, and

20

fourth affirmative defenses.

21

A.

First Affirmative Defense (“Fair Notice”)

22

“A fundamental principle in our legal system is that laws which regulate persons or entities

23

must give fair notice of conduct that is forbidden or required.” FCC v. Fox Television Stations,

24

Inc., 567 U.S. 239, 253 (2012). A statute regulating commercial conduct, like the FTC Act, fails

25

to provide constitutionally adequate fair notice “only if it is so indefinite in its terms that it fails to

26

articulate comprehensible standards to which a person’s conduct must conform.” See Robles v.

27

Domino’s Pizza, LLC, 913 F.3d 898, 906 (9th Cir. 2019). Courts must examine such challenged

28

statutes “in the light of the facts of the case at hand.” See, e.g., United States v. Mazurie, 419 U.S.
44

1

544, 550 (1975); Donovan v. Royal Logging Co., 645 F.2d 822, 831 (9th Cir. 1981) (noting that

2

whether “a statute is unconstitutionally vague must be assessed in the context of the particular

3

conduct to which it is being applied”) (internal quotation marks and citation omitted).

4

LendingClub’s first affirmative defense asserts that “[t]he FTC is barred from obtaining a

5

judgment of FTC Act liability or relief on the conduct alleged in Counts I and III because

6

LendingClub lacked constitutionally adequate notice that Section 5(a)’s prohibition on ‘unfair or

7

deceptive acts or practices’ could reach the conduct challenged in those counts.” (Dkt. No. 81 at

8

40.) The Court addresses this defense as it applies to each Count, and concludes that judgment on

9

the pleadings is warranted.

10

1.

Count I (Deceptive Act)

United States District Court
Northern District of California

11

As previously discussed, an act or practice is deceptive for purposes of Section 5(a)

12

liability “if: (1) there is a representation, omission, or practice that, (2) “is likely to mislead

13

consumers acting reasonably under the circumstances, and (3) the representation, omission, or

14

practice is material.” CFPB v. Gordon, 819 F.3d 1179, 1192 (9th Cir. 2016). The Ninth Circuit

15

has recognized that “[t]he term ‘deceptive act or practice’ has an established meaning in the

16

context of the [FTC Act].” Id. at 1193 n.7; see also FTC v. Johnson, 96 F. Supp. 3d 1110, 1142

17

(D. Nev. 2015) (rejecting “fair notice” argument as to Section 5(a) liability and noting that “[t]here

18

is extensive case law and guidance on what constitutes a deceptive act or practice under Section

19

5(a).”). Count I alleges that LendingClub charges a “hidden up-front fee” constituting a deceptive

20

practice in violation of Section 5(a) of the FTC Act, based on the origination fee and

21

LendingClub’s representation in its ads and consumer loan applications that its loans contain “no

22

hidden fees.” (Dkt. No. 57 at ¶¶ 56-58.)

23

The gist of LendingClub’s “fair notice” defense as to Count I is that LendingClub lacked

24

constitutionally adequate notice that it could be liable under Section 5 for deceptive acts because it

25

disclosed the origination fee in compliance with TILA and CFPB safe harbor regulations, and

26

thus, could not have known that the representation “no hidden fees” was deceptive under the

27

circumstances. (Dkt. No. 81 at 41.) The FTC’s Rule 12(c) motion asserts that this defense fails as

28

to Count I because it is both procedurally improper and deficient on the merits. (Dkt. No. 139 at
45

1

4-6.) Because the Court agrees that the defense fails on the merits, it addresses only the latter

2

argument.

3

LendingClub’s affirmative defense does not allege that the language of Section 5(a) is

4

itself impermissibly vague. Indeed, LendingClub’s opposition expressly disclaims the argument

5

that “Section 5 has ‘no established meaning,’” and recognizes that “all industry participants are

6

unquestionably on notice that charging hidden fees may violate Section 5 of the FTC Act.” (See

7

Dkt. No. 201-7 at 12 (emphasis added)). Instead, LendingClub’s opposition clarifies that its

8

defense:

9
10

United States District Court
Northern District of California

11
12
13

invokes fair notice “as applied” to the specific facts of this case. Fox
Television, 567 U.S. at 258. While all industry participants are
unquestionably on notice that charging hidden fees may violate
Section 5 of the FTC Act, LendingClub had no notice that a fee
disclosed in compliance with TILA—which was designed by
Congress to achieve “clear and conspicuous” disclosure of loan terms,
15 U.S.C. § 1632(a)—could be challenged by the FTC as a “hidden”
fee violative of Section 5.

14

(Id.) LendingClub’s “as applied” challenge fails as a matter of law. LendingClub’s reliance on its

15

alleged compliance with TILA is effectively a defense on the merits to Section 5(a) liability; it is

16

not an affirmative defense that the FTC Act itself or the FTC’s enforcement of Section 5 fails to

17

give fair notice of the conduct that is proscribed. It is well established that deception for purposes

18

of a Section 5(a) claim “may be found based on the ‘net impression’ created by a representation.”

19

See Stefanchik, 559 F.3d at 928 (quoting Cyberspace.Com, 453 F.3d at 1200). It is also well

20

established that “[a] solicitation may be likely to mislead by virtue of the net impression it creates

21

even though the solicitation contains truthful disclosures.” See Cyberspace.Com, 453 F.3d at

22

1200 (emphasis added). Thus, unlike the unconstitutionally vague and inconsistent agency

23

enforcement policy at issue in Fox Television, the enforcement standard under Section 5(a) “as

24

applied” to the allegedly deceptive conduct here is clear—a representation may be found deceptive

25

based on the net impression it creates and the presence of “truthful disclosures” does not preclude

26

liability. LendingClub cites no authority for the broad proposition that TILA-compliance alone

27

protects a lender from liability for deceptive practices under the FTC Act; indeed, and as the Court

28

has previously noted, there is no such authority. (See Dkt. Nos. 53 at 17-19 (concluding that
46

1

compliance with TILA does not preclude liability under Section 5(a) of the FTC Act as a matter of

2

law) & 80 at 9 (same).)
Because LendingClub acknowledges that it was “unquestionably on notice that charging

3
4

hidden fees”—the same conduct alleged in Count I—“may violate Section 5 of the FTC Act,” and

5

there is no dispute regarding the standard for determining whether an act is “deceptive” within the

6

meaning of Section 5(a), LendingClub’s first affirmative defense as to Count I fails as a matter of

7

law.

8
9

2.

Count III (Unfair Practice)

The FTC asserts that LendingClub’s first affirmative defense as to Count III is similarly
deficient. The Court disagrees. As previously discussed, the FTC Act defines an actionable unfair

11
United States District Court
Northern District of California

10

practice or act as one that “‘causes or is likely to cause substantial injury to consumers which is

12

not reasonably avoidable by consumers themselves and not outweighed by countervailing benefits

13

to consumers or to competition.’” Neovi, 604 F.3d at 1155 (quoting 15 U.S.C. § 45(n)). Count III

14

alleges that “[i]n numerous instances, Defendant has withdrawn money from borrowers’ bank

15

accounts without borrowers’ authorization, or in amounts in excess of borrowers’ authorization[,]”

16

causing “substantial injury to consumers that consumers cannot reasonably avoid themselves and

17

that is not outweighed by countervailing benefits to consumers or competition.” (Dkt. No. 57 at ¶¶

18

62-63.) Count III further alleges that the unauthorized withdrawals “constitute unfair acts or

19

practices in violation of Section 5 of the FTC Act, 15 U.S.C. §§ 45(a) and 45(n).” (Id. at ¶ 64.)

20

LendingClub’s “fair notice” affirmative defense as to Count III alleges that:

21

23

the FTC failed to provide LendingClub with constitutionally adequate
notice that LendingClub’s use of its ACH payment processing
procedures would subject it to Section 5(a) liability for the
purportedly “unauthorized” account withdrawals alleged in that
count.

24

(Dkt. No. 81 at 42.) The gist of the defense is that LendingClub’s error rate for unauthorized ACH

25

withdrawals is so low that LendingClub was not on notice that “the mere use of an ACH payment

26

processing system with such error rates could subject it to Section 5(a) liability for an ‘unfair’

27

payment practice ‘under the circumstances’ alleged in Count III.” (See id. (“Even accepting the

28

allegations regarding ‘numerous’ and ‘unauthorized’ withdrawals, LendingClub will prove that its

22

47

1

ACH payment processing system has an error rate of less than 1%, and that this rate is within the

2

range of ACH errors inherent in an electronic payment system offering the features and flexibility

3

of LendingClub’s system.”).)

4

Thus, LendingClub’s affirmative defense for Count III does not argue that Section 5 is

5

itself impermissibly vague as to the conduct proscribed; instead, its affirmative defense is an “as

6

applied” challenge to the FTC’s enforcement of Section 5 against LendingClub on “the specific

7

facts of this case.” (Id. at 12.) Construing the pleadings as true and in the light most favorable to

8

LendingClub, the FTC fails to carry its burden of showing that “there is no issue of material fact in

9

dispute, and the moving party is entitled to judgment as a matter of law.” See Fleming, 581 F.3d
at 925. As pleaded, LendingClub’s as-applied challenge gives rise to a plausible inference that

11
United States District Court
Northern District of California

10

LendingClub was not on notice that its specific ACH withdrawal system and the associated

12

payment-processing errors “of less than 1%” could subject it to liability under Section 5. Cf. FTC

13

v. Wyndham Worldwide Corp., 799 F.3d 236, 256 (3d Cir. 2015) (“Fair notice is satisfied here as

14

long as the company can reasonably foresee that a court could construe its conduct as falling

15

within the meaning of the statute”). This is especially true given that liability for unfair practices

16

under Section 5 extends only to conduct that is “‘not outweighed by countervailing benefits to

17

consumers or to competition.’” Neovi, 604 F.3d at 1155 (quoting 15 U.S.C. § 45(n)).

18

LendingClub’s fair notice allegations as to Count III give rise to a plausible inference that the

19

benefits of its specific “ACH processing procedures” outweigh the low instances of “error[ ]

20

inherent in an electronic payment system.” (See Dkt. No. 81 at 42.) Thus, the Court cannot say as

21

a matter of law that LendingClub’s conduct would not “survive a reasonable interpretation of the

22

cost-benefit analysis required [under Section 5].” See Wyndham, 799 F.3d at 256 (concluding that

23

defendant’s “as-applied” challenge to Section 5 failed where defendant failed to demonstrate that

24

its allegedly unfair conduct “survive[d] a reasonable interpretation of the cost-benefit analysis

25

required [under the statute]”).

26

Accordingly, the Court grants the FTC’s motion for judgment on the pleadings on

27

LendingClub’s first affirmative defense as to Count I only. The Court denies the FTC’s motion as

28

to Count III.
48

1

B.

Third Affirmative Defense (Equitable Estoppel)

2

A party asserting equitable estoppel against a private litigant must prove the traditional
elements of estoppel: “(1) the party to be estopped must know the facts; (2) he must intend that his

4

conduct shall be acted on or must so act that the party asserting the estoppel has a right to believe

5

it is so intended; (3) the latter must be ignorant of the true facts; and (4) he must rely on the

6

former’s conduct to his injury.” Baccei v. United States, 632 F.3d 1140, 1147 (9th Cir. 2011). In

7

addition to the traditional elements, “a party asserting equitable estoppel against the government

8

must also establish that (1) the government engaged in affirmative misconduct going beyond mere

9

negligence; (2) the government’s wrongful acts will cause a serious injustice; and (3) the public’s

10

interest will not suffer undue damage by imposition of estoppel.” Id. “Affirmative misconduct on

11
United States District Court
Northern District of California

3

the part of the government requires an affirmative misrepresentation or affirmative concealment of

12

a material fact, such as a deliberate lie or a pattern of false promises.” Id. (citation omitted).

13

“[M]ere inaction . . . cannot support a claim of equitable estoppel” against the government. Id.
LendingClub’s third affirmative defense asserts that “[t]he FTC is equitably estopped from

14
15

bringing Count I and/or seeking remedies premised on borrowers not understanding the total cost

16

of their loans because LendingClub reasonably relied on and complied with the FTC’s and

17

CFPB’s regulations, guidance, and statements relating to TILA and loan price and term

18

disclosures.” (Dkt. No. 81 at 45.) The FTC moves for judgment on the pleadings on this defense

19

on the grounds that the defense fails as a matter of law because LendingClub “alleged only the

20

four traditional elements of estoppel, and not the additional three that apply when the defense is

21

asserted against the government[,]” and “its lengthy narrative does not set forth allegations that

22

could support any of [those additional elements].”14 (Dkt. No. 139 at 7.) In other words, the FTC

23

asserts that the third affirmative defense fails to state claim upon which relief can be granted.

24
25
26
27
28

14

The Court previously denied the FTC’s motion to strike this defense. (See Dkt. No. 80 at 8-9.)
The FTC’s reply briefing for that motion raised the same argument set forth here; specifically: that
the defense fails to “meet the strict ‘affirmative misconduct’ standard required to assert equitable
estoppel against the government.” (See Dkt. No. 71 at 14.) The Court declined to consider that
argument, however, because the FTC did not initially raise it in its motion to strike. (Dkt. No. 80
at 9 (citing Zamani v. Carnes, 491 F.3d 990, 997 (9th Cir. 2007) (noting that courts “need not
consider arguments raised for the first time in a reply brief”)).)
49

1

There is no dispute that LendingClub fails to specifically identify all of the requisite

2

elements for estoppel against the government; however, an affirmative defense need only “contain

3

sufficient matter to state a defense that is ‘plausible on its face.’” See Ramirez v. Ghilotti Bros.

4

Inc., 941 F. Supp. 2d 1197, 1204 (N.D. Cal. 2013) (quoting Ashcroft v. Iqbal, 556 U.S. 662, 678

5

(2009)). Thus, the FTC must show that the defense—on the face of the pleadings and viewing the

6

allegations in the light most favorable to LendingClub—fails to state a plausible claim. The FTC

7

has met that burden.

8

As for “affirmative misconduct,” the pleadings allege, in pertinent part: “LendingClub . . .
relied on the CFPB’s and/or the FTC’s endorsement of APR and finance charge disclosures, and

10

the CFPB’s comments and interactions with the company during LendingClub’s participation in

11
United States District Court
Northern District of California

9

the CFPB ‘Trial Disclosure Program,’ where it had discussions with the CFPB staff in an effort to

12

improve the ease of use of its disclosure of consumer credit terms.” (Dkt. No. 81 at 45.) Further,

13

LendingClub alleges: “At no point during the extensive discussions with CFPB staff were any

14

questions raised regarding the adequacy or presentation of the origination fee disclosures included

15

in the personal loan application flow on LendingClub’s platform.” (Id. at 22.)

16

The FTC asserts that the alleged government conduct was not “affirmative misconduct,”

17

but instead consists of “general business guidance on disclosures and TILA and the CFPB’s

18

silence on Defendant’s loan application flow.” (Dkt. No. 139 at 7.) The FTC notes that

19

LendingClub does not allege that “the FTC or CFPB stated that its loan flow was not deceptive[,]”

20

and “argues only that the the FTC and CFPB never previously challenged it.” (Id.) Further, the

21

FTC asserts that LendingClub “does not allege any deliberate lie or pattern of false promises, and

22

silence on the part of the government does not amount to affirmative misconduct.” (Id.)

23

LendingClub’s opposition counters that the allegations “create[ ] a question of fact

24

regarding affirmative misconduct” because “the FTC now, with full knowledge of the CFPB’s

25

extensive past engagement with LendingClub, seeks to challenge as non-compliant that exact

26

same content.” (Dkt. No. 201-7 at 19.) LendingClub further asserts that it has alleged “much

27

more than mere silence by federal officials[,]” and has instead “specifically allege[d] that

28

government officials had ‘discussions’ with LendingClub about its disclosures, and that
50

1
2

LendingClub specifically relied on those statements.” (Id. at 20.)
The question then is whether LendingClub’s allegations regarding “discussions” with the

3

CFPB about LendingClub’s disclosures are insufficient to plausibly support affirmative

4

misconduct. They are. Construing the allegations in the light most favorable to LendingClub, the

5

mere reference to “discussions” without allegations regarding an “affirmative misrepresentation or

6

affirmative concealment of a material fact, such as a deliberate lie or a pattern of false promises”

7

expressly regarding the origination fee disclosures does not give rise to a reasonable inference that

8

the CFPB engaged in “affirmative misconduct” for purposes of governmental estoppel. See

9

Baccei, 632 F.3d at 1147 (noting that “[a]ffirmative misconduct on the part of the government
requires an affirmative misrepresentation or affirmative concealment of a material fact, such as a

11
United States District Court
Northern District of California

10

deliberate lie or a pattern of false promises”). The Amended Answer does not allege that the

12

government affirmatively told LendingClub that its “no hidden fees” representations did not

13

violate the FTC Act or that the government knew that the disclosures did violate the Act but

14

deliberately lied or falsely promised that the disclosures did not do so. Thus, the defense fails.

15

LendingClub’s reliance on Clinger v. Farm Servs. Agency does not counsel a different

16

result. There local government agents “counseled [the plaintiffs] on which parcels [of land] they

17

should identify for crop destruction” in order to qualify for a USDA program. No. CV 04 424 E

18

BLW, 2006 WL 581192, at *6 (D. Idaho Mar. 8, 2006). The government then “made official

19

determinations that the [plaintiffs] qualified for the program” “notified them accordingly,” and the

20

plaintiffs “relied on the[ ] official determinations” and destroyed the designated crops. Id. at *6-7.

21

The USDA subsequently found the plaintiffs ineligible for the program because the designated

22

field was not individually-owned. Id. at *6. The court determined that the government “did much

23

more than stand by and offer negligent advice,” or “vague pronouncements or qualified

24

endorsements.” Instead, the court concluded that the government engaged in affirmative

25

misconduct by counseling the plaintiffs on which fields they should destroy to qualify for the

26

program, making an official determination that they did qualify, and notifying the plaintiffs

27

accordingly. Id. at *7. No such affirmative statements expressly regarding the origination fee

28

disclosures and “no hidden fees” representation is alleged here.
51

1

LendingClub’s reliance on FTC v. DirecTV, Inc. is also unavailing. There the court did not

2

engage in a detailed analysis of whether the governmental conduct alleged—failure to participate

3

in negotiations regarding a settlement agreement, despite being invited to do so, covering the same

4

conduct at issue in the case—constituted “affirmative misconduct.” No. 15-cv-01129-HSG, 2015

5

WL 9268119, at *2-3. The court instead denied the FTC’s motion to strike the governmental

6

estoppel defense because the defense gave the FTC notice of the defendant’s legal theory and

7

“adequately plead[ed] facts which plausibly could support a finding in [the defendant’s] favor.”

8

Id. at *3. Here, LendingClub fails to plead facts that could plausibly support a finding of

9

affirmative misconduct for the reasons previously stated.
As to the “serious injustice” element of governmental estoppel, the FTC argues that “an

11
United States District Court
Northern District of California

10

injunction prohibiting Defendant from further misrepresentations and ordering it to pay back to

12

consumers the hidden origination fees it took is not a ‘serious injustice.’” (See Dkt. No. 139 at 8.)

13

The FTC’s reply further asserts that LendingClub’s Amended Answer pleads no facts to support

14

that LendingClub would suffer “serious injustice” absent estoppel. (Dkt. No. 207 at 14.) The

15

Court agrees.

16

The Amended Answer includes no allegations regarding the “serious injustice” element.

17

However, LendingClub’s opposition asserts that if the FTC prevails on Count I “LendingClub

18

could be required to pay a significant monetary remedy premised on the inadequacy of disclosures

19

specifically reviewed, analyzed, and commented on by CFPB.” (Dkt. No. 201-7 at 20-21.) The

20

problem with that argument is that it is premised on affirmative misconduct by the government,

21

which LendingClub fails to plead.

22

Similarly, LendingClub’s failure to plead facts in support of the “public interest” element

23

renders the third affirmative defense insufficient as a matter of law. LendingClub’s opposition

24

argues that the Amended Answer’s allegations regarding LendingClub’s voluntary interactions

25

with the CFPB through the Trial Disclosure Program and incorporation of CFPB feedback into its

26

loan disclosures satisfies the public interest factor in part because “‘the public has an interest in

27

seeing its government deal carefully, honestly, and fairly with its citizens.’” (Dkt. No. 201-7 at 21

28

(quoting United States v. Wharton, 514 F.2d 406, 413 (9th Cir. 1975)).) Again, however, that
52

1

argument is premised on affirmative misconduct by the government, which LendingClub fails to

2

plead.
Accordingly, the Court grants judgment on the pleadings as to LendingClub’s third

3
4

affirmative defense because the defense as pleaded is insufficient as a matter of law.
LendingClub asserts in a footnote that it should be granted leave to amend as to this

5

defense because “the Court rejected the FTC’s previous attempt to strike this defense, . . . and

7

LendingClub therefore had no prior reason to replead its equitable estoppel defense to address the

8

FTC’s new argument.” (Dtk. No. 201-7 at 18 n.8.) “The standard for granting leave to amend

9

under Rule 12(c) is also . . . identical to Rule 12(b)(6). Pantastico, 406 F. Supp. 3d at 877 (citing

10

Pac. W. Grp., Inc. v. Real Time Sols., Inc., 321 F. App’x 566, 569 (9th Cir. 2008)). It is not clear

11
United States District Court
Northern District of California

6

that the governmental estoppel defense “could not be saved by amendment”; thus, ordinarily,

12

leave to amend would be warranted. See Harris v. Cty. of Orange, 682 F.3d 1126, 1131 (9th Cir.

13

2012).

14

Given that the trial has been continued to October, the Court will grant LendingClub 14

15

days leave to amend this affirmative defense, provided it can in good faith allege affirmative

16

government misconduct sufficient to invoke estoppel against the government. The issue of

17

whether the amendments are sufficient to allow the defense to proceed to trial can be addressed at

18

the pretrial conference rather than in another round of motion briefing, especially since some

19

evidence as to the CFPB is likely to be admitted in connection with Count I in any event.

20

C.

Fourth Affirmative Defense (Judicial Estoppel)

21

The equitable doctrine of judicial estoppel “precludes a party from gaining an advantage

22

by asserting one position, and then later seeking an advantage by taking a clearly inconsistent

23

position.” Hamilton v. State Farm Fire & Cas., 270 F.3d 778, 782 (9th Cir. 2001). Judicial

24

estoppel typically has three elements: (1) “a party’s later position must be ‘clearly inconsistent’

25

with its earlier position”; (2) the party must have “succeeded in persuading a court to accept that

26

party’s earlier position”; and (3) “the party seeking to assert an inconsistent position [must] derive

27

an unfair advantage or impose an unfair detriment on the opposing party if not estopped.” New

28

Hampshire v. Maine, 532 U.S. 742, 750-51 (2001) (noting that the three elements “do not
53

1

establish inflexible prerequisites or an exhaustive formula for determining the applicability of

2

judicial estoppel”). “The application of judicial estoppel is not limited to bar the assertion of

3

inconsistent positions in the same litigation, but is also appropriate to bar litigants from making

4

incompatible statements in two different cases.” Hamilton, 270 F.3d at 783.

5

LendingClub’s fourth affirmative defense asserts that “[t]he FTC is judicially estopped

6

from bringing Count I and/or seeking remedies premised on borrowers not understanding the total

7

cost of the personal loans available through LendingClub’s platform.” (Dkt. No. 81 at 45.) The

8

Amended Answer alleges, in pertinent part:

9
10

United States District Court
Northern District of California

11
12
13
14
15
16

Count I and/or certain of the remedies the FTC seeks are premised on
reasonable borrowers not understanding the disclosures of APR and
the total finance charge in the Federal Box of the TILA disclosure in
the personal loan application flow. The FTC previously took the
position in FTC v. AMG Services, No. 2:12-cv-00536- GMN-VCF (D.
Nev.), and potentially other cases involving theories of liability or
monetary remedies that relate to disclosures of APR or the total
finance charge, that reasonable consumers understood disclosures of
APR and finance charge, including when those disclosures were
presented in the Federal Box of a TILA disclosure. The FTC is
therefore judicially estopped from asserting in this case that
reasonable consumers did not understand truthful and accurate
disclosures of APR and the finance charge presented, among other
places, in the Federal Box of a TILA disclosure.

17

(Id.) The gravamen of the defense is that “the FTC cannot contend that reasonable borrowers

18

understand and rely on Federal Box disclosures [in TILA disclosure statements] in AMG Services

19

but then in this case contend that borrowers do not understand and rely on Federal Box

20

disclosures.” (Dkt. No. 201-7 at 26.)

21

The FTC moves for judgment on the pleadings on this defense on the grounds that

22

LendingClub “has not identified any inconsistent positions taken by the FTC”; specifically:

23

LendingClub “do[es] not identify these ‘other cases’ and do[es] not specify where in the 1,000+

24

entries of the AMG docket the FTC took [the allegedly inconsistent] position.” (Dkt. No. 139 at 8-

25

9.) Further, the FTC argues that its position in AMG was the same as its position in this case and

26

the difference between the two cases is that the defendants “incorporated deception into different

27

parts of their marketing and application process.” (Id. at 9 (“In both cases, the FTC took the

28

position that defendants engaged in deceptive practices regarding the fees associated with their
54

1

loans.”).) The Court agrees that LendingClub’s judicial estoppel defense fails as a matter of law

2

because the FTC’s positions here and in AMG are not “clearly inconsistent.”15
The Amended Answer misstates the FTC’s position on Section 5 liability in both cases.

3
4

LendingClub alleges that here: “Count I and/or certain of the remedies the FTC seeks are premised

5

on reasonable borrowers not understanding the disclosures of APR and the total finance charge in

6

the Federal Box of the TILA disclosure in the personal loan application flow.” (Dkt. No. 81 at

7

46.) LendingClub is wrong. As alleged in the FAC, Count I is premised on reasonable consumers

8

being misled by the representation “no hidden fees” based on the net impression that statement

9

creates in light of the up-front origination fee and the method by which that fee is disclosed in the
loan application flow, and the actual loan amount applicants receive. That the origination fee is

11
United States District Court
Northern District of California

10

disclosed in the TILA disclosure statement on the Loan Rates & Terms does not mean that the

12

FTC’s “position” on Section 5 liability in this case is that reasonable consumers “do not

13

understand and rely on Federal Box disclosures,” as LendingClub argues. (See Dkt. No. 201-7 at

14

26 (emphasis omitted).) It simply means that the FTC believes that the TILA disclosure in this

15

case did not fix the net impression created by the representation elsewhere in the loan application

16

flow (and in advertisements) that LendingClub charged “no hidden fees.”
The Amended Answer also misstates the FTC’s position as to Section 5 liability in AMG.16

17
18

As relevant here, the FTC brought a Section 5 claim in AMG based on an allegedly deceptive

19

TILA disclosure and moved for summary judgment. The FTC asserted that the TILA disclosure

20

was misleading because it prominently and clearly presented certain loan repayment terms in the

21
22
23
24
25
26
27
28

15

The FTC argues for the first time in its reply that LendingClub’s judicial estoppel defense also
fails as a matter of law because judicial estoppel is inappropriate in “a case where estoppel would
compromise a governmental interest in enforcing the law,’” and here, “the FTC is pursuing the
government’s interest in enforcement of the law.” (Dkt. No. 207 at 16 (quoting New Hampshire,
532 U.S. at 755).) Because the FTC did not raise this argument in their motion and provide
LendingClub with an opportunity to respond, the Court will not consider it. See Zamani, 491 F.3d
at 997 (noting that courts “need not consider arguments raised for the first time in a reply brief”).
Further, the argument is not necessary to conclude that the defense is insufficient as a matter of
law.
16
Both LendingClub’s opposition and the FTC’s reply include as exhibits several filings from
AMG. (See Dkt. Nos. 202-5 – 202-8, Exs. 4-7 & 207-2 – 207-3, 207-5 – 207-7, Exs. 1-2, 4-6.)
Judicial notice is appropriate for “undisputed matters of public record, including documents on file
in federal or state courts.” Harris v. Cty. of Orange, 682 F.3d 1126, 1132 (9th Cir. 2012).
Accordingly, the Court takes judicial notice of the proffered exhibits.
55

1

Federal Box of the TILA disclosure but borrowers were automatically entered into a different

2

repayment plan with terms set forth in fine print below the Federal Box, and the automatic

3

repayment plan included additional charges not listed in the Federal Box. (Dkt. No. 202-5, Ex. 4

4

at 4-6.) Thus, the Section 5 claim in AMG was based on the net impression of the TILA disclosure

5

itself: the Federal Box contained loan terms that conflicted with the fine print below it.

6

The FTC’s position in AMG was not, as LendingClub alleges, simply that “reasonable
consumers understood disclosures of APR and finance charge, including when those disclosures

8

were presented in the Federal Box of a TILA disclosure.” (See Dkt. No. 81 at 46.) Instead, the

9

district court’s order adopting the magistrate judge’s report and recommendation noted that as to

10

“the misleading net impression” of the TILA disclosure, “the FTC repeatedly argued in its motion

11
United States District Court
Northern District of California

7

that summary judgment was appropriate because of the ‘inconspicuous, contradictory, confusing,

12

and vague language’ in the document.” (See Dkt. No. 202-5, Ex. 4 at 18 (quoting the FTC’s

13

argument in its motion for summary judgment that “‘the loan documents were confusing,

14

particularly on the issue of the repayment terms’”).) In other words, the FTC argued that the net

15

impression of the TILA disclosure, including the Federal Box, was misleading based on the

16

discrepancy between the terms in the Federal Box and the conflicting terms in the fine print below

17

it. The AMG court agreed and granted summary judgment, finding: “[T]he net impression of the

18

[TILA disclosure] is likely to mislead borrowers acting reasonably under the circumstances

19

because the large prominent print in the TILA Box implies that borrowers will incur one finance

20

charge while the fine print creates a process under which multiple finance charges will be

21

automatically incurred unless borrowers take affirmative action.” (Id. at 16.)

22

The FTC’s positions on Section 5 liability in AMG and this case are not “clearly

23

inconsistent.” In AMG, the universe of the net impression analysis was the TILA disclosure

24

because that was the document specifically challenged; thus, the focus was on the Federal Box and

25

the fine print below it. The FTC argued in AMG that the conflicting repayment terms in the TILA

26

disclosure was likely to mislead consumer acting reasonably under the circumstances. Here, the

27

universe is the entire loan application flow as it relates to the representation “no hidden fees.” The

28

FTC asserts that the TILA disclosure statement and Federal Box does not cure the net impression
56

1

of that representation and that it is likely to mislead a consumer acting reasonably under the

2

circumstances. Indeed, even assuming without argument that LendingClub’s TILA disclosure was

3

TILA-compliant, “[a] solicitation may be likely to mislead by virtue of the net impression it

4

creates even though the solicitation also contains truthful disclosures.” See Cyberspace.Com, 453

5

F.3d at 1200. Accordingly, LendingClub’s judicial estoppel defense fails as a matter law to the

6

extent it asserts that the FTC is pursuing a “clearly inconsistent” theory of Section 5 liability in

7

this case.

8
9

The defense is similarly deficient to the extent it asserts that the FTC’s position regarding
monetary remedies for Count I is “clearly inconsistent” with its position in AMG. As explained in
LendingClub’s opposition, its “monetary remedies” argument is that in AMG: “the FTC

11
United States District Court
Northern District of California

10

calculated, and the court accepted, a monetary remedy premised on the assumption that

12

[defendant’s] borrowers understood the finance costs disclosed in the Federal Box, and that all

13

finance costs paid beyond those amounts constitute consumer harm.” (Dkt. No. 201-7 at 23.)

14

LendingClub asserts that “the FTC argued that the monetary remedy should be measured by the

15

amount ‘consumers [ ] repaid more than the amount borrowed plus [the] one finance charge’

16

disclosed in the Federal Box.” (Id. quoting Dkt. No. 202-7, Ex. 6 at 74 (alterations in original).)

17

LendingClub’s opposition highlights the following quotes from the FTC’s motion for summary

18

judgment in AMG on issues that remained after the court had granted the FTC’s motion for

19

summary judgment on its Section 5 claim:

20
21
22
23
24
25

[T]he Court determined that Defendants’ loan documents were likely
to mislead consumers by prominently disclosing the fixed cost to
repay their loans (the principal borrowed plus one finance charge),
but then automatically enrolling customers into a repayment schedule
resulting in much higher costs, often several multiples of the disclosed
amount.
...
Thus, the most reasonable approximation of the consumers’ monetary
loss is found by totaling, for consumers who actually repaid more than
the disclosed cost (the principal amount and one finance charge), the
aggregate amount those consumers paid above the disclosed cost.

26

(Dkt. No. 202-7, Ex. 6 at 74.) So? The TILA disclosure itself was the challenged document in

27

AMG and the Federal Box set forth one set of loan repayment terms while the fine print set forth

28

conflicting terms that resulted in additional finance charges. Thus, the FTC used the set of loan
57

1

repayment terms “prominently disclos[ed]” in the Federal Box to determine the “reasonable

2

approximation of the consumers’ monetary loss” caused by the additional finance charges in the

3

fine print. That approximation was entirely reasonable because the loan repayment terms and

4

additional finance charges were directly at issue in that case.
The FTC’s reasonable approximation of consumer harm in AMG does not mean, as

5
6

LendingClub argues, that the remedy was “premised on the assumption that [defendant’s]

7

borrowers understood the finance costs disclosed in the Federal Box, and that all finance costs

8

paid beyond those amounts constitute consumer harm.” (See Dkt. No. 201-7 at 23.) It was instead

9

premised on the amount of monetary harm suffered as a result of the misleading representation at
issue—the TILA disclosure. That the FTC calculates harm in this case based on the amount of

11
United States District Court
Northern District of California

10

origination fees collected by LendingClub and seeks restitution of those fees does not mean that

12

the FTC’s position on monetary remedies here is “clearly inconsistent” with its position in AMG

13

for purposes of judicial estoppel.
Because LendingClub cannot satisfy the first element of judicial estoppel, its affirmative

14
15

defense fails as a matter of law. Accordingly, the Court grants the FTC’s motion for judgment on

16

the pleadings as to LendingClub’s fourth affirmative defense.

17

V.

Administrative Motions to File Under Seal

18

A party seeking to seal a document filed in conjunction with a motion related to the merits

19

of a case must overcome the “strong presumption in favor of public access” to judicial records by

20

meeting the “compelling reasons” standard. Kamakana v. City and Cty. of Honolulu, 447 F.3d

21

1171, 1178-79 (9th Cir. 2006) (noting that the “strong presumption of access to judicial records

22

applies fully to dispositive pleadings, including motions for summary judgment and related

23

attachments”); see also Ctr. for Auto Safety v. Chrysler Grp., LLC, 809 F.3d 1092, 1098, 1101

24

(9th Cir. 2016) (noting that the “compelling reasons” test applies “to most judicial records,”

25

including documents attached to nondispositive motions that are “more than tangentially related to

26

the merits of a case”) (internal quotation marks and citation omitted); In re Midland Nat. Life Ins.

27

Co. Annuity Sales Practices Litig., 686 F.3d 1115, 1120-21 (9th Cir. 2012) (applying “compelling

28

reasons” standard to Daubert motions submitted “in connection with” motions for summary
58

1

judgment). The requesting party “must articulate[ ] compelling reasons supported by specific

2

factual findings . . . that outweigh the general history of access and the public policies favoring

3

disclosure.” Kamakana, 447 F.3d at 1178-79 (alteration in original) (internal quotation marks and

4

citations omitted). “Compelling reasons” exist when court documents “might have become a

5

vehicle for improper purposes,” such as the release of trade secrets and other competitively

6

sensitive business information. Nixon, 435 U.S. at 598. It is not sufficient that public disclosure

7

of court records may lead to a litigant’s embarrassment, incrimination, or exposure to further

8

litigation. Foltz v. State Farm Mut. Auto. Ins. Co., 331 F.3d 1122, 1136 (9th Cir. 2003).

9

A party seeking to seal documents must also comply with the Civil Local Rules, which
provide that sealing is appropriate only where the requesting party “establishes that the document,

11
United States District Court
Northern District of California

10

or portions thereof, are privileged, protectable as a trade secret or otherwise entitled to protection

12

under the law.” Civ. L.R. 79-5(b). Further, parties must “narrowly tailor” their requests only to

13

the sealable material and redact documents accordingly. Civ. L.R. 79-5(d). The Local Rules also

14

provide that where “the Submitting Party is seeking to file under seal a document designated as

15

confidential by the opposing party or a non-party pursuant to a protective order,” the Designating

16

Party must file within four days of the Submitting Party’s motion “a declaration as required by

17

subsection 79-5(d)(1)(A) establishing that all of the designated material is sealable.” Civ. L.R. 79-

18

5(e)(1).

19

On April 14, 2020, the Court issued an order addressing the parties’ motions to seal and

20

clarifying that the “compelling reasons” standard applied to the motions. (Dkt. No. 233.) The

21

Order directed LendingClub to submit supplemental declarations in light of the “compelling

22

reasons” standard and new unredacted and redacted versions of the material it seeks to seal if

23

application of the standard changed its previously proposed redactions. LendingClub complied

24

with that Order. (See Dkt. Nos. 245-245-1; 248-265.)

25

A.

LendingClub’s Motions to Seal

26

LendingClub moves to file under seal: (1) portions of certain exhibits submitted in support

27

of its motion to exclude the expert testimony of Mr. Loewenstein, (Dkt. No. 137 & 249); (2)

28

portions of its motion for partial summary judgment, and portions of the declaration and exhibits
59

1

submitted in support of same, (Dkt. No. 140 & 248); (3) portions of its opposition to the FTC’s

2

motion for partial judgment on the pleadings and exhibits submitted in support of same, (Dkt. No.

3

201 & 255); (4) portions of its opposition to the FTC’s motion for summary judgment and exhibits

4

submitted in support of same, (Dkt. No. 218 & 251); (5) portions of its opposition to the FTC’s

5

motion to exclude expert testimony and exhibits in support of same, (Dkt. No. 221 & 253); and (6)

6

portions of its reply in support of its motion for partial summary judgment and exhibits in support

7

same, (Dkt. No. 241). LendingClub also filed amendments to its motions to seal at Dkt. Nos. 140

8

& 201, (see Dkt. No. 214).

9

LendingClub subsequently withdrew its confidentiality designations as to some of the
documents filed under seal in connection with its motion for partial summary judgment, (Dkt. No.

11
United States District Court
Northern District of California

10

143); some of the documents filed under seal in connection with its opposition to the FTC’s

12

motion for summary judgment, (Dkt. No. 223): some of the documents filed under seal in

13

connection with its opposition to the FTC’s motion for judgment on the pleadings, (Dkt. No. 202),

14

and some of the documents submitted in connection with its opposition to the FTC’s motion to

15

exclude expert testimony, (Dkt. No. 224). (See Dkt. Nos. 214 & 245.) Accordingly, those

16

documents shall be filed without redaction on the public docket.

17

The Court has reviewed LendingClub’s supplemental proposed redactions and declaration

18

in support of sealing material it designated as confidential and is satisfied that LendingClub’s

19

requests meet the “compelling reasons” standard. (See Dkt. No. 245 & 245-1.) The material

20

constitutes non-public, competitively sensitive business information regarding LendingClub’s

21

business operations and strategy, consumer research, financial data, or personally identifiable

22

consumer information. Further, the proposed redactions are narrowly tailored to only sealable

23

information. Accordingly, the Court GRANTS LendingClub’s requests to seal as described in the

24

declaration of Richard H. Cunningham, (Dkt. No. 245-1).

25

In response to LendingClub’s motion to seal portions of its opposition to the FTC’s motion

26

for summary judgment and exhibits submitted in support of same, (Dkt. No. 218), and pursuant to

27

Civil Local Rule 79-5(e)(1), the FTC filed a declaration in support of sealing four exhibits, (Dkt.

28

Nos. 219-12, Ex. 14; 219-13, Ex. 15; 219-14, Ex. 16; 219-15, Ex. 17), designated as confidential
60

1

by the FDIC pursuant to the Protective Order in this case, (see Dkt. No. 44). The FTC’s

2

declaration, (Dkt. No. 230), includes an attachment from Daniel H. Kurtenbach of the FDIC in

3

support of sealing, (Dkt. No. 230-1). Mr. Kurtenbach, counsel in the Legal Division of the FDIC,

4

attests that the documents:

5

consist of bank examination reports and summaries containing (1)
confidential business information about WebBank’s operations and
practices; (2) confidential business information about LendingClub’s
operations and practices; and (3) FDIC bank examiner opinions,
conclusions, and recommendations concerning WebBank’s
operations and LendingClub’s operations.

6
7
8

(Id. at ¶ 8.) Mr. Kurtenback further attests that the FDIC provided the material to the FTC on the

10

condition that the FTC “mark and ensure that the documents are treated as Confidential under the

11
United States District Court
Northern District of California

9

Protective Order in this case” because of the potential for harm to WebBank and LendingClub if

12

the documents were exposed to the public. (Id. at ¶¶ 9-10.) The Court has reviewed the

13

documents and agrees that “compelling reasons” standard is met and sealing is warranted.

14

B.

The FTC’s Motions to Seal

15

The FTC moves to filed under seal: (1) portions of its motion for summary judgment and

16

certain exhibits submitted in support of same, (see Dkt. No. 145); (2) portions of its motion to

17

exclude expert testimony and exhibits submitted in support of same, (see Dkt. No. 146); (3)

18

portions of its opposition to LendingClub’s motion to exclude expert testimony and one exhibit

19

submitted in support of same, (see Dkt. No. 211); (4) portions of its opposition to LendingClub’s

20

motion for partial summary judgment and exhibits submitted in support of same, (see Dkt. No.

21

215); (5) portions of its reply in support of its motion to exclude expert testimony, (see Dkt. No.

22

234); (6) errata exhibits, (see Dkt. No. 236); and (7) portions of its reply in support of its motion

23

for summary judgment, (see Dkt. No. 239).

24

With the exception of the FTC’s motion to seal portions of exhibits in support of its motion

25

for summary judgment that contain personally identifiable consumer information, (see Dkt. Nos.

26

148-6, Ex. 10 at 5-6; 149-4, Ex. 14 at 6-13; 158-3, Ex. 214; 164-1, Ex. 7; 164-9, Ex. 246; 164-10,

27

Ex. 247), the FTC’s motions concern material produced by LendingClub during discovery and

28

designated as confidential by LendingClub under the stipulated Protective Order in this case, (see
61

1

Dkt. Nos. 44 & 123). The Court grants the FTC’s motion to seal the personally identifiable

2

consumer information because such material satisfies the “compelling reasons” standard. See

3

Nursing Home Pension Fund v. Oracle Corp., No. C01-00988 MJJ, 2007 WL 3232267 (N.D. Cal.

4

Nov. 1, 2007) (noting that “compelling reasons exist to keep personal information confidential to

5

protect an individual’s privacy interest and to prevent exposure to harm or identity theft”).

6

Pursuant to Civil Local Rule 79-5(e)(1), LendingClub filed responses and declarations in

7

support of sealing certain portions of the FTC’s briefing and exhibits that LendingClub had

8

designated as confidential. (See Dkt. Nos. 198 & 214 (addressing Dkt. Nos. 145 & 146); 231

9

(addressing Dkt. Nos. 211 & 215).) Following the Court’s Orders on April 14 & 21, 2020
regarding the parties’ motions to seal, LendingClub filed supplemental declarations and briefing in

11
United States District Court
Northern District of California

10

support of sealing. (See Dkt. Nos. 245-1 at ¶¶ 10-74, 87, 91-95, 102 (declaration of Mr.

12

Cunningham); 250 (addressing Dkt. No. 146)); 252 (addressing Dkt. No. 215); 254 (addressing

13

Dkt. No. 211); 256-65 (addressing Dkt. No. 145); 266 (declaration of James C. Jackson for non-

14

party WebBank in support of sealing Dkt. Nos. 156-9, Ex. 31; 187-9, Ex. 196; 201-4, Ex. 2

15

(designated portions)); 268 (addressing Dkt. Nos. 234; 236; 239).)

16

LendingClub’s responses withdraw its confidentiality designations for certain documents.

17

First, LendingClub withdraws its designations for certain exhibits the FTC filed under seal in

18

connection with its motion for summary judgment, (Dkt. No. 147). (See Dkt. Nos. 198 at 6-16 &

19

245 at 6-17.) LendingClub also withdraws its designations for certain documents the FTC filed

20

under seal in connection with its opposition to LendingClub’s motion for partial summary

21

judgment, (Dkt. No. 216). (See Dkt. Nos. 231 at 3-4 & 245 at 20.) Further, LendingClub

22

withdraws its designations for certain documents the FTC filed under seal in connection with its

23

motion to exclude expert testimony, (Dkt. No. 155). (Dkt. No. 245 at 21.) Accordingly, those

24

exhibits shall be filed without redaction on the public docket.

25

LendingClub also withdraws its confidentiality designations for the following exhibits

26

filed in connection with the FTC’s administrative motion to seal errata exhibits designated by

27

LendingClub as confidential, (Dkt. No. 236): Dkt. Nos. 236-13, Ex. 261 (filed as “PX 2” in

28

connection with Dkt. No. 236); 236-14, Ex. 227 (filed as “PX 3” in connection with Dkt. No.
62

1

236); 236-15, Ex. 129 (filed as “PX 4” in connection with Dkt. No. 236); 236-19, Ex. 195 (filed as

2

“PX 8” in connection with Dkt. No. 236); 236-20, Ex. 200 (filed as “PX 9” in connection with

3

Dkt. No. 236); 236-21, Ex. 201 (filed as “PX 10” in connection with Dkt. No. 236).17 (See Dkt.

4

No. 268 at 3.) However, as to Dkt. No. 236-14, Ex. 227, LendingClub’s supplemental submission

5

at Dkt. No. 265 seeks to seal the same information. (See Dkt. No. 265 at 8 (citing Dkt. No. 265-5

6

at 4).) Further, Mr. Cunningham’s supplemental declaration in support of sealing asserts that

7

compelling reasons exist to seal the designated portion of Exhibit 227 because it contains

8

competitively sensitive business information that is not relevant to the FTC’s claims. (See Dkt.

9

No. 245-1 at ¶ 72.) Thus, it is unclear whether LendingClub’s subsequent withdrawal of its
confidentiality designation as to Exhibit 227 is in error. Because the document in question is an

11
United States District Court
Northern District of California

10

internal “Compliance Regulatory Risk Assessment Report” issued in June 2018 and

12

contemporaneously marked “Highly Confidential,” and the redacted material consists of an

13

executive summary detailing the results of the risk assessment, LendingClub shall clarify its

14

position as to Exhibit 227 within 10 days of this Order. The other exhibits for which LendingClub

15

withdraws its confidentiality designations shall be filed without redaction on the public docket.
LendingClub’s supplemental responses in support of sealing also narrow or maintain the

16
17

designations for certain documents. The Court has reviewed LendingClub’s supplemental

18

proposed redactions and declaration in support of sealing material it designated as confidential and

19

is satisfied that LendingClub’s requests meet the “compelling reasons” standard and Mr.

20

Cunningham’s declaration complies with Civil Local Rule 79-5(e)(1). (See Dkt. No. 245 & 245-

21

1.) The material constitutes non-public, competitively sensitive business information regarding

22

LendingClub’s business operations and strategy, consumer research, financial data, or personally

23

identifiable consumer information, and the proposed redactions are narrowly tailored to only

24

sealable information. Accordingly, the Court GRANTS LendingClub’s requests to seal as

25
26
27
28

17

The FTC designated the exhibits submitted in connection with Dkt. No. 236 with “PX” numbers
that do not coincide with the original “PX” numbers used in connection with the original,
underlying motion (i.e., “PX 3” for purposes of Dkt. No. 236 is designated as “PX 227” for
purposes of the FTC’s motion for summary judgment). The Court maintains the original “PX”
number for consistency and to avoid further confusion.
63

1

described in the declaration of Richard H. Cunningham, (Dkt. No. 245-1).

2

The Court warns that it may revisit the sealing of any of these documents if a party seeks

3

to introduce them at trial. A party will not meet its burden of showing that the exhibit should be

4

shielded from the public by merely pointing out that the Court allowed the document to be sealed

5

in connection with summary judgment or other dispositive motions.

6

CONCLUSION

7

For the reasons set forth above, the Court grants in part and denies in part the FTC’s

8

motion for summary judgment. The Court GRANTS summary judgment on Count II, and

9

DENIES summary judgment on Counts I, III, and IV.
The Court grants in part and denies in part LendingClub’s cross-motion for partial

11
United States District Court
Northern District of California

10

summary judgment on Counts III and IV. The Court GRANTS summary judgment on Count IV’s

12

request for injunctive relief and DISMISSES that claim as moot because there is no available

13

remedy. The Court DENIES summary judgment on Count III.

14

The Court DENIES the FTC’s Daubert motion to the extent it seeks to exclude the

15

testimony of Dr. Wind. The Court otherwise DENIES the parties’ respective Daubert motions as

16

moot because the challenged testimony was not necessary to resolve the parties’ cross-motions for

17

summary judgment.

18

The Court grants in part and denies in part the FTC’s motion for judgment on the

19

pleadings. The Court GRANTS judgment on the pleadings on LendingClub’s First Affirmative

20

Defense as applied to Count I, and LendingClub’s Third and Fourth Affirmative Defenses. The

21

Court DENIES judgment on the pleadings on LendingClub’s First Affirmative Defense as applied

22

to Count III.

23

The Court GRANTS the parties’ respective motions to seal as set forth above.

24

IT IS SO ORDERED.

25

Dated: June 1, 2020

26
27
JACQUELINE SCOTT CORLEY
United States Magistrate Judge

28
64