Mullen v. Wells Fargo & Company et al

Filing 137

ORDER RE 109 MOTION TO DISMISS. (whalc1, COURT STAFF) (Filed on 5/6/2022)

1
2
3
UNITED STATES DISTRICT COURT

4
5

NORTHERN DISTRICT OF CALIFORNIA

6
7
8
9
10

EMPLOYEES’ RETIREMENT SYSTEM OF
THE STATE OF HAWAII, on behalf of itself
and similarly-situated individuals,
Plaintiff,

United States District Court
Northern District of California

11
12
13
14
15
16

No. C 20-07674 WHA

v.
WELLS FARGO & COMPANY, C. ALLEN
PARKER, TIMOTHY J. SLOAN, JOHN R.
SHREWSBERRY, PERRY PELOS, MARK
MYERS, and KARA MCSHANE,

ORDER RE MOTION
TO DISMISS

Defendants.

17
INTRODUCTION

18
19
20
21
22

In this putative securities class action, defendants move to dismiss. To the extent stated,
defendants’ motion is GRANTED.
STATEMENT
Here follow the facts, as pleaded. At all material times, Wells Fargo & Company was a

23

financial services and bank holding company. Among other services, it originated commercial

24

real estate (CRE) loans to fund the purchase, remodeling, or refinancing of commercial

25

property, as well as commercial and industrial (C&I) loans to fund business operating expenses

26

(Amd. Compl. ¶ 51). In addition to originating commercial loans, Wells Fargo also bundled

27

and sold the right to collect on those loans, a process known as sponsoring a securitization (id.

28

¶¶ 52, 220). The investment products it bundled included Commercial Mortgage Backed

1

Securities (CMBS), which star in the complaint (id. ¶ 53). Employees’ Retirement System of

2

the State of Hawaii invested in one or more Wells Fargo CMBS and serves as court-appointed

3

lead plaintiff in this putative class action. The putative class consists of persons or entities

4

damaged as a result of acquiring stock in commercial loans sponsored by Wells Fargo between

5

October 13, 2017, and October 13, 2020 (id. ¶ 1). The consolidated amended complaint asserts

6

claims against Wells Fargo and certain officers (id. ¶¶ 42–46).

7

The claims concern the ways in which Wells Fargo assessed the financial strength of its

8

commercial borrowers during loan origination. It further concerns assessment of the borrowers

9

whose existing loans Wells Fargo sponsored into a CMBS. The complaint “describes a
pervasive problem of lenders and securities issuers have [sic] regularly altered financial data

11

United States District Court
Northern District of California

10

for commercial properties without justification to make the properties appear more valuable,

12

and borrowers more creditworthy, than they actually are.” The complaint incorporates articles

13

and studies that concerned the entire industry but also calls out information and trends specific

14

to Wells Fargo, which allegedly dominated the commercial lending industry (id. ¶¶ 87 (cleaned

15

up), 99, 100, 101, 111, 158, 159).

16

As stated, when it issued a loan, Wells Fargo evaluated borrowers’ ability to pay, as well

17

as the value of any property securing the loan (collateral), in order to determine the size of any

18

loan. Prior to sponsoring a loan into a CMBS, Wells Fargo similarly underwrote the existing

19

loan. As used herein, underwriting was this art of predicting a business’ future ability to pay.

20

“Wells Fargo’s process for originating and underwriting commercial mortgage loans” were

21

allegedly identical. Both included, among other things, evaluating credit, rent, operating

22

budgets, predicted future cash flow, and real property (sometimes using appraisers). Central

23

here, the “underwriting process” before CMBS sponsorship could include “adjustments” to a

24

borrower’s stated financial figures in order to accommodate the underwriter’s opinion about a

25

borrower’s long-term ability to pay (id. ¶¶ 106, 221; id. n.15; see also ¶¶ 73, 94, 106, 146, 152,

26

221–22, 225–31, 239).

27

Wells Fargo specialized in two major types of securitized commercial investment

28

products: collateralized loan obligations (CLOs), which contained commercial loans of many
2

1

types, and CMBSs, which contained only CRE mortgages. The complaint also notes that the

2

inflationary practices applied to all commercial lending, including loans to alternative asset

3

managers who in turn issued commercial loans using similar risky inflationary practices.

4

CMBS data availability makes Wells Fargo’s CMBS sponsorship the focus of the complaint,

5

however (Amd. Compl. ¶¶ 15, 19, 51–54, 152, 243).

6

Critical concepts in the complaint include:

7

•

NOI was “total rent and other revenues minus general operating
expenses like management, utilities, cleaning, repairs, and
maintenance” (id. ¶ 161). NOI was a key input that ultimately
helped to determine the size of a loan that a business could
receive (id. ¶ 105).

•

“Net Cash Flow (NCF) [was] NOI minus replacement of capital
items such as building and tenant improvements, and leasing
commissions” (id. ¶ 161).

•

Loan-to-value (LTV) ratio referred to the size of the loan relative
to the value of the business (see id. ¶ 213).

•

“[R]eserves, allowances, charge-offs, and other impairments”
referred to the capital Wells Fargo needed to hold in order to
offset any loan that borrowers would not fully repay (see id. ¶¶
21, 24, 256).

8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16

Between 2016 and 2019, Wells Fargo developed at least twenty-six CMBSs. The “deals”

17

were valued at between $192 million and $1.04 billion. Wells Fargo originated up to 48.9% of

18

the commercial loans that it bundled into CMBSs in this period. The remainder Wells Fargo

19

sponsored from other originators. A CMBS usually contained less than one hundred bundled

20

commercial loans, far fewer than the residential mortgage-backed securities made famous in

21

the financial crisis (id. ¶¶ 54, 231).

22

In 2019, the wholesale banking division provided 55% of Wells Fargo’s net income. As

23

of June 2020, C&I loans amounted to approximately $350 billion of Wells Fargo’s $513

24

billion commercial lending business. CRE loans comprised approximately $146 billion of the

25

same (id. ¶¶ 48, 51).

26
27

Defendants now move to dismiss. This order follows full briefing, oral argument
(telephonic due to COVID-19), and supplemental briefing.

28
3

ANALYSIS

1
2

When ruling on motions to dismiss brought under Section 10(b), “courts must, as with
any motion to dismiss for failure to plead a claim on which relief can be granted, accept all

4

factual allegations in the complaint as true.” Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551

5

U.S. 308, 322 (2007). “Securities fraud class actions must,” however, “meet the higher,

6

exacting pleading standards of Federal Rule of Civil Procedure 9(b) and the Private Securities

7

Litigation Reform Act (PSLRA).” Oregon Pub. Employees Ret. Fund v. Apollo Grp. Inc., 774

8

F.3d 598, 604 (9th Cir. 2014). To state a claim under Section 10(b), a complaint must plead (i)

9

a material misrepresentation or omission; (ii) scienter; (iii) connection with the purchase or

10

sale of a security; (iv) reliance; (v) economic loss; and (vi) a causal connection between the

11

United States District Court
Northern District of California

3

material misrepresentation and the loss. Dura Pharm., Inc. v. Broudo, 544 U.S. 336, 342

12

(2005). Defendants contest falsity, materiality, scienter, and loss causation (Br. i).

13

1.

14

Section 10(b) of the PLSRA prohibits fraud or deceit in connection with the sale of

FALSITY.

15

securities. See In re VeriFone Holdings, Inc. Sec. Litig., 704 F.3d 694, 703 (9th Cir. 2012).

16

SEC Rule 10b-5 makes it illegal “[t]o make any untrue statement of a material fact or to omit

17

to state a material fact necessary in order to make the statements made, in the light of the

18

circumstances under which they were made, not misleading.” 17 C.F.R. § 240.10b-5(b). An

19

actionable “omission” must do more than leave out a fact. It must “affirmatively create an

20

impression of a state of affairs that differs in a material way from the” real one. Brody v.

21

Transitional Hosps. Corp., 280 F.3d 997, 1006 (9th Cir. 2002).

22

Our complaint chiefly alleges that Wells Fargo originated, bundled, and sold loans into

23

CMBSs while exaggerating the safety of those loans by failing to disclose its true, inflationary

24

underwriting practices. This, it’s alleged, rendered false or misleading defendant officers’

25

statements about, broadly speaking, (1) credit risk and credit quality relating to underwriting

26

practices and loan-to-value ratios, and (2) necessary credit reserves and impairments. This

27

order reiterates only a representative sample. As to (1) (all emphases in the original):

28
4

1

•

On October 13, 2017, an investor presentation slide in the
2017Q3 Quarterly Supplement presentation referred to “strong
credit quality” and to “continued credit discipline,” and also
asserted that Wells Fargo “[m]aintained our risk and pricing
discipline” (defendants Timothy J. Sloan and John R.
Shrewsberry presented) (id. ¶¶ 269–71).

•

On January 12, 2018, Shrewsberry stressed Wells Fargo’s
“continued credit discipline in a very competitive market” to
investors and, in the same call, said “Our credit quality remained
exceptionally strong” (id. ¶ 284).

•

On April 13, 2018, Shrewsberry stated during the 2018Q1
Quarterly Supplement presentation: We’ve “maintained our
credit risk discipline for new originations in commercial real
estate during a period of high liquidity and increased
competition” (id. ¶ 296).

•

On May 10, 2018, Sloan presented slides at Investor Day that
discussed, “Risk Management,” and touted “[c]ontinued
disciplined focus on credit and market risk.” Another slide,
“Building from a strong foundation,” referenced “strong credit
discipline” (id. ¶ 301 (brackets in the original)).

•

On June 10, 2020, Shrewsberry told the 2020 Morgan Stanley
Virtual US Financials Conference, “On commercial real estate,
we have a big book. We’re probably $150 billion all in of [sic]
commercial real estate, including construction. . . . [O]n an LTV
basis, oh, gosh, 90% — more than 90% of that book has less
than 70% loan-to-value based on our own underwriting of
that” (id. ¶ 397).

•

In June 2020, the “Maryland Attorney General described Wells
Fargo’s actions as directly contributing to the Financial Crisis.”
Referring to the lawsuit, “Wells Fargo stated that ‘[w]hile we
don’t agree with the state’s view on these matters, we are pleased
to be able to put these legacy issues behind us’” (id. ¶ 61).

2
3
4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19

Allegedly false representations also include references to Wells Fargo having learned its lesson
20
from the financial crisis. Concerning (2), reserves and other impairments, a press release in
21
January 2018 announced quarter-end allowances for credit losses of $12 billion, or 1.25% of
22
23

total loans. This disclosure was allegedly “false and/or misleading because Wells Fargo
materially understated the reserves and impairments needed in its commercial loan portfolio.”

24
At all material times, impairments needed referred to the financial reserves necessary to
25
compensate for losses such as nonperforming assets, non-accruals, and charge-offs (id. ¶ 256).
26
The complaint alleges that defendants made additional allegedly false or misleading
27
disclosures on this topic at various points throughout the class period (id. ¶¶ 282–83; see also
28
5

1

¶¶ 289, 311, 313; 2018: ¶¶ 299–300, 323–24, 340–42; 2019: ¶¶ 349–50, 367–69; 2020: ¶¶

2

393–96).

3
4

A.

THE “TRUTH” EMERGED.

The “truth” allegedly first emerged in May 2020, when ProPublica reported on an

5

investigation into alleged inflationary underwriting in the commercial lending market. The

6

findings came from “an expert with decades of experience in the commercial lending

7

industry,” financial analyst John Flynn. In a letter Flynn had written to the SEC in 2019, Flynn

8

analyzed commercial lenders’ alleged inflationary underwriting. His analysis prominently

9

featured examples from Wells Fargo. The ProPublica article also featured Wells Fargo. It
named the bank seven times, showed a Wells Fargo branch in the picture lead, and explained

11

United States District Court
Northern District of California

10

the alleged scheme using several Wells Fargo loan exemplars. Flynn had analyzed the

12

exemplars (among others) in the manner next discussed (id. ¶¶ 8, n.14, 82–100; Heather

13

Vogell, Whistleblower: Wall Street Has Engaged in Widespread Manipulation of Mortgage

14

Funds, ProPublica (May 15, 2020), https://www.propublica.org/article/whistleblower-wall-

15

street-has-engaged-in-widespread-manipulation-of-mortgage-funds).

16

Flynn did not work for Wells Fargo. Rather, as a concerned citizen who suspected

17

inflationary commercial lending, Flynn surveyed “thousands of loans” issued from large banks

18

including Wells Fargo. Specifically, he examined underwritten financial figures for

19

commercial real estate loans sponsored into CMBS bundles. He observed that financial

20

“figures were consistently inflated in” the public reporting of “new loans.” He observed this

21

inflation when comparing underwritten figures to those same borrowers’ financial “figures . . .

22

for the same historical year as reported by the servicer for prior loans” (Amd. Compl. ¶¶ 9, 96).

23

Put differently, Flynn examined on one hand underwritten financial figures for a given

24

business in given years. On the other hand, he examined unadjusted financial figures for that

25

same commercial borrower for the “same building[]” in the “same years.” This comparison

26

became possible only because that business reported its raw financial figures for the same

27

property and year to a loan servicer. The raw figures became public because the earlier CMBS

28

to which the loan belonged had published them. ProPublica examined six of the loans that
6

1

Flynn studied and reported that CMBS sponsors inflated the later underwritten financial figures

2

by as much as 30%. In all, Flynn’s 2019 estimate provided that $150 billion in inflated

3

CMBSs, made up of real estate loans, had issued from major lenders since 2013 (id. ¶¶ 8, 9,

4

15, 84–86, 88, 96).

5

Flynn has also allegedly elaborated on the findings he reported to the SEC to discuss

6

Wells Fargo in particular. He did so after his letter to the SEC and for the benefit of our

7

amended complaint. Per the complaint, Flynn stated that the pattern of inflation seen across

8

the industry applied “with equal force” to loans that Wells Fargo originated. Specifically,

9

Flynn examined “multiple CMBS trusts that include[d] loans that Wells Fargo originated and
sold directly into the trusts.” Therein Flynn identified “precisely” the same practice as he

11

United States District Court
Northern District of California

10

found industry-wide: “inflated historical income and cash flow figures, and changes to

12

identifying information.” The changes to identifying information, according to Flynn,

13

reflected the banks’ attempts to make the different figures appear to come from different

14

businesses, even though they did not. Flynn concluded that no valid explanation existed for

15

such a wide gap between raw and underwritten figures from old to new loans. Underwriters

16

allegedly ought to have used conservative standards to predict borrower-business’ future

17

abilities to pay. By that logic, adjustments should rarely have changed the underwritten

18

figures. If anything, any change should have reduced expected future income. The reason is

19

that if adjustments did the opposite and inflated borrowers’ expected future income, such

20

inflation would justify issuing a bigger loan — perhaps one that the borrower could not pay.

21

This, in turn would endanger CMBS investors. In sum, Flynn contended that inflation on the

22

scale he described revealed unjustifiably risky commercial lending at Wells Fargo (id. n.15, ¶¶

23

9, 16, 84–87, 92–96, 159, 209).

24

Second, finance professor John M. Griffin, PhD and his PhD-candidate coauthor

25

“confirmed” Flynn’s findings. They released a preprint study dated November 2020 on the

26

Social Science Research Network (SSRN), an online repository of preprint academic articles.

27
28
7

1

Third, the last alleged revelation about CMBS inflation came from The Intercept. In

2

April 2021, the publication spoke with Flynn, concurred with him after reviewing his data, and

3

also reported on Griffin’s findings (id. ¶¶ 101–29, 130–41).

4

Preliminarily, this order must decide which version of Griffin’s paper to consider.

5

Plaintiff did not append a copy to the complaint. Instead, the complaint drops a footnote with

6

the URL and summarizes the study. The body of the complaint names a release date of

7

November 18, 2020. Today, however, the same URL links a version dated September 20,

8

2021. While the parties appear to agree that the latter paper reaches largely the same

9

conclusion, the latter is substantially revised and fills nearly double the pages (id. n.16, ¶¶ 101–

10

United States District Court
Northern District of California

11

29).
Plaintiff contends that the latter version was “not mentioned in the complaint” (Dkt. No.

12

132 at 2 (cleaned up)). Thus, says plaintiff, the operative complaint cannot incorporate it by

13

reference. See Khoja v. Orexigen Therapeutics, 899 F.3d 988, 1003 (9th Cir. 2018).

14

Defendants disagree, arguing that the 2021 version should be deemed incorporated (Dkt. No.

15

131). That version added a graph favoring defendants’ cause.

16

The framework for this dispute is the incorporation by reference doctrine. Orders like

17

ours may consider any document “whose contents are alleged in a complaint and whose

18

authenticity no party questions, but which are not physically attached to the plaintiffs’

19

pleading.” Knievel v. ESPN, 393 F.3d 1068, 1076 (9th Cir. 2005) (quoting In re Silicon

20

Graphics Inc. Sec. Litig., 183 F.3d 970, 986 (9th Cir. 1999) (superseded by statute on other

21

grounds)). Only two district courts (and no appellate courts) appear to have reached the

22

immediate issue. Each decision simply accepted two versions of a website. The issues arose at

23

motions on the pleadings. In each case, one party cited a version of an internet source different

24

from one in the complaint. In Sabin v. Curt Manufacturing Company, Sabin moved for, inter

25

alia, declaratory judgment that his copycat website did not infringe on the company’s

26

trademark. 2009 WL 10673588, at *1–3 (D. Ariz. May 4, 2009) (Judge Susan R. Bolton). The

27

decision simply listed some minor differences between versions of the website and accepted

28

both. The second relevant decision concerned an online list of corporate directors attached as
8

1

an exhibit to a motion to dismiss. That decision accepted a later version because it contained

2

“the same information” as the earlier. That earlier document was already “incorporated by

3

reference into the Complaint. . . .” In re Yahoo! Inc. S’holder Derivative Litig., 153 F. Supp.

4

3d 1107, 1118 (N.D. Cal. 2015). In re Yahoo! took judicial notice of the site because

5

incorporation by reference of the first version incorporated the latter as well. This order finds

6

both Sabin and In re Yahoo! useful, though Griffin’s massive update far exceeded the

7

differences between the websites relevant to those decisions.

8

We must keep in mind the apparent point made in citing to Griffin in the first place in the
complaint. The article appeared after the class period. The point being made was to confirm,

10

after the fact, that inflation of financial figures had indeed occurred. Put differently, Flynn had

11

United States District Court
Northern District of California

9

earlier alerted the SEC. Griffin later, after the class period, allegedly confirmed that Flynn had

12

been correct.

13

For this purpose, we should look to the most recent version of the Griffin article. Since

14

all versions occurred after the fact and no version was part of the critical events affecting stock

15

price, we should look to Griffin’s most recent and updated article in determining whether

16

Griffin confirmed Flynn. To take a different example, suppose a linked document first stated

17

that a witness had told the author a certain fact but that a later version of the article at the same

18

link corrected it to say, no, that was a mistake, the witness did not say it after all. We would

19

surely want to erase any reliance on the original statement, now shown to be incorrect via the

20

very same link. This would be true even if the complaint somehow tried to lock in only the

21

original version at the link. But it is all the more true when the complaint refers merely to the

22

link without capturing the original version. So we should look to the later Griffin version even

23

though it now undercuts the complaint.

24
25

B.

FALSE?

The complaint fails adequately to allege that Wells Fargo unjustifiably inflated the

26

financial figures of up to one-third of loans sponsored into CMBS trusts, that it understated

27

loss reserves, or that defendants misstated practices to investors. None of the sources cited

28

reveal that Wells Fargo’s true underwriting standards contradicted descriptions of “continued
9

1

credit discipline,” or of “conservative,” “rigorous,” “solid,” or “strong” credit quality flowing

2

from its lending practices (see, e.g., Amd. Compl. ¶¶ 5, 270–77, 280, 289–296, 402). Nor does

3

the complaint adequately plead false or misleading statements with respect to LTV ratios, a

4

critical input in the underwriting process. Similarly, allegations of false disclosures about

5

credit reserves and impairments do not succeed.

6
7

(i)

Underwriting and Credit Risk.

The gist of the following long analysis is this: First, Wells Fargo’s underwriting

8

standards proved largely accurate or conservative, not inflationary, as measured by borrowers’

9

actual NOI after the first year of a loan. Second, the complaint does not adequately allege that,
in light of the inflation and deflation rates Griffin detailed, Wells Fargo issued “truly risky”

11

United States District Court
Northern District of California

10

commercial loans (Opp. Br. at 23; Amd. Compl. ¶¶ 88, 99, 100, 108, 110; Exh. 132-2 at 21,

12

Figure IA.7. Panel A).

13

A more detailed explanation now follows. Griffin used CMBSs’ publicly-reported

14

financial figures to examine whether Wells Fargo had been “inflat[ing]” those financial

15

figures. He decided yes that the data “point[ed] to originators knowingly inflating

16

underwritten income” (Dkt. No. 132-2 at 35). Griffin analyzed loan-level data in “a sample of

17

39,522 CMBS loans,” sponsored by various banks. This sample carried “a market

18

capitalization of $650 billion underwritten between January 1, 2013 and December 31, 2019”

19

(Amd. Compl. ¶¶ 104, 108, n.17). “Griffin compared NOIs for the two years prior to the loan

20

at issue as reported in the securitization materials versus the NOIs for those same prior years as

21

reported for an earlier CMBS” (id. ¶ 110). The latter category, “as reported from an earlier

22

CMBS,” referred to raw numbers that borrowers reported to their loan servicers (servicer-

23

reported NOI), i.e., not underwritten. Thus, Griffin, like Flynn, used CMBS reporting to

24

examine whether underwriting generally inflated commercial borrowers’ apparent financial

25

health.

26

The chart below displays the comparison between underwritten and servicer-reported

27

NOI. The orange pillars in the chart immediately below represent the proportion of loans with

28

underwritten NOI inflated by >5%. The blue pillars represent the proportion of loans with
10

1

underwritten NOI deflated by >5%. Griffin found “[s]ubstantially more than 30% of

2

commercial loans that Wells Fargo originated had NOI inflated as compared to the amount of

3

NOI reported for the same year in a prior transaction.” Less than 5% of these loans showed

4

deflation in underwriting by >5% (id. ¶ 111, n.20; Dkt. No. 132-2 at Fig. 6. Panel A (oval

5

added)).

6

SERVICER-REPORTED NOI:
RELATIONSHIP TO UNDERWRITTEN NOI

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
26
27

This, plaintiff says, is the smoking gun. The inflation seen is unjustifiable, per the
complaint: underwriters had no valid reason to adjust the figures upward. The pictured rates
of inflation and deflation allegedly mean that Wells Fargo systematically and intentionally
exaggerated when estimating borrowers’ future NOI in about 30% of loans, making the pool of
securitized loans inordinately risky (Amd. Compl. ¶¶ 113–18). In truth, the inflation and
deflation above simply represented underwriters’ opinions about borrower-business’
creditworthiness. Underwriters predicted about 30% of commercial borrowers would have
greater future NOI than the raw figures suggested and that >60% of borrowers would realize
NOI equal to or lower than what borrowers reported in the base year. How well did the
underwriters forecast future NOI?

28
11

1

Griffin contrasted underwritten NOI with the NOI that a business saw in the first year of

2

payment on the loan (first-year realized NOI). The chart below shows the comparisons. The

3

businesses whose loans Wells Fargo sponsored into CMBS trusts saw “underwritten income

4

exceed[] actual net operating income by 5% or more in 29% of loans,” during the first year of

5

payment on the respective loans (Dkt. No. 132-2 at 2).*
Actual net operating incomes also exceeded underwritten income >5% in about 40% of

6

loans. In other words, business’ real-life NOIs showed that underwriting predictions had been

8

conservative more often than not (id. at Figure IA.7, Panel A (oval added)). Both versions of

9

the Griffin paper addressed this. The latter version, however, examined the frequency with

10

which first-year realized NOI surpassed underwritten. Here is the Griffin chart that plaintiff

11

United States District Court
Northern District of California

7

wishes to expunge, but which this order now considers (ibid. (oval added)):

12
FIRST-YEAR REALIZED NOI:
RELATIONSHIP TO UNDERWRITTEN NOI

13
14

Figure IA.7. Panel A. Proportion Overstated and Understated

15
16
17
18
19
20
21
22
23
24
25
26
27
28

*

This order estimates the orange pillar in Figure IA.7. Panel A (above) as 28%, rather than the
29% quoted by Griffin (see Dkt. No. 132-2 at 2).
12

1

In sum, Wells Fargo’s underwriting standards proved to be conservative. Defendants

2

correctly argue: “[U]nderwritten numbers should not necessarily match borrower-reported

3

financials but should reflect the context-specific assumptions and adjustments of . . . [l]enders

4

assess[ing] the income-producing capability of a property and summariz[ing] their expectations

5

of sustainable cash flows with the underwritten” NOI (Br. 21–22, quoting Amd. Compl. ¶ 106

6

(cleaned up, emphasis in the original)). With respect to first-year realized NOI, defendants

7

contend that the approximately 28% overestimation and approximately 40% underestimation

8

“negate[] Plaintiff’s assertion that Wells Fargo intentionally and improperly inflated

9

underwritten NOI” (Dkt. No. 132 at 3).

10

United States District Court
Northern District of California

11

In short, Wells Fargo’s 28%:40% ratio represented, defendants argue, a valid over-under.
Defendants say that this alone should defeat falsity. This order agrees.

12

First, nothing about the professional underwriting standards that plaintiff cites suggested

13

that the frequency of upward adjustments to servicer-reported NOI ranked as beyond the pale.

14

The complaint has not tried to explain, for instance, what the proper amount of inflation or

15

deflation would be either with respect to servicer-reported or first-year realized NOI.

16

According to plaintiff, however, the Griffin chart showing the conservative first-year

17

realized underwriting practice does not refute plaintiff’s point, for the paper “in no ways

18

suggests that a lender’s understated loans somehow compensate for its overstated ones” (Dkt.

19

No. 131 at 3). Wells Fargo systematically and unjustifiably inflated its commercial borrowers’

20

ability to pay on their loans, plaintiff says.

21

Why? Put simply, underwriting must very rarely inflate raw financial figures. In

22

support, plaintiff points to the Commercial Real Estate Finance Council’s (CREFC) standards.

23

These provided that underwritten NOI “should reflect . . . minimum expected cash flow over

24

an extended time period” (Amd. Compl. ¶ 106; Dkt. No. 132-2 at 7). CREFC standards also

25

provided that adjustments “usually result in cash flow decreases” while increasing the expected

26

cash flows only occurred when such an increase was “clearly” warranted (Dkt. No. 132-2 at 7).

27

Griffin concluded as well that “underwritten income should be a stable and conservative

28
13

1

measure” based on “conservative guidelines” and that “deal documents indicate that lenders

2

often take conservative measures” (Dkt. No. 132-2 at 2, n.8).
The complaint further relied on disclosures from the prospectus for a Wells Fargo deal,

3
4

however (Amd. Compl. ¶¶ 220–31) (see, infra, Section 2). Defendants argue that the

5

prospectus acknowledges underwriters’ ability to adjust financial figures. It states that

6

revenues and expenses are “often highly subjective values,” and “[a]ctual net cash flow for a

7

Mortgaged Property may be less than the Underwritten Net Cash Flow presented with respect

8

to that property” (Tang Decl. Exh. J at 170–173). Plaintiff disagrees. Noting that the

9

prospectus claimed to abide by Item 1125 of Regulation AB (17 C.F.R. 229.1125), which
delineated exclusions and inclusions for calculating operating expenses (adjustments), plaintiff

11

United States District Court
Northern District of California

10

contends that the regulation does not permit Wells Fargo’s apparent “open-ended exclusions”

12

(Opp. Br. 11; Lieberman Decl. Exh. 1). It also argues that the prospectus discussed properties’

13

performance at the time of the loan, not historical figures. At the very least, however, the

14

portion discussing historical figures tracks the amended complaint’s own acknowledgement

15

that underwriting allows for certain adjustments (Tang Decl. Exh. J at 163; Amd. Compl. ¶

16

106).

17

The complaint never alleges what rate of inflation/deflation these ‘conservative’ industry

18

standards would approve. Plaintiff merely argues that the inflation shown implied that Wells

19

Fargo must have used a “different methodology” than AB 1125 required, and indeed a

20

different one than any regulatory standards prescribed (Amd. Compl. ¶ 106; Dkt. No. 132-2 at

21

7, n.8). This conclusion does not follow. Griffin never states an acceptable level of inflation.

22

Nor does he point to a sample institution that showed a level of inflation that he found

23

conservative/compliant with CREFC guidelines. He simply describes certain lenders as “more

24

conscientious” than others, which characterization tells us nothing about the appropriate degree

25

of inflation (Dkt. No. 132-2 at 2). Given that the inflation of servicer-reported NOI

26

represented underwriters’ predictions, this order finds the first-year realized NOI inflation

27

provides a meaningful proxy for how well the underwriters did. Our complaint does not plead

28
14

1

with particularity that a ‘risk spread’ of approximately 40% deflation and 28% inflation unduly

2

endangered investors. See Gompper v. VISX, Inc., 298 F.3d 893, 895 (9th Cir. 2002).

3

Second, plaintiff objects that Wells Fargo’s ‘risk spread’ reflected in the first-year

4

realized NOI cannot be considered conservative because the consequences of over-estimating

5

borrowers’ financial figures plausibly posed a risk of sending borrowers into distress or

6

default, whereas underestimating them did not. That risk, however, exists with every

7

securitized loan. Underwriting simply estimated a borrower-business’ future ability to pay.

8

Wells Fargo’s first-year realized NOI (in the latter chart above) revealed that roughly 70% of

9

estimates proved either overly conservative or roughly accurate. Underwriters’ judgment
proved reasonable too often to support plaintiff’s allegation (Amd. Compl. ¶ 3; Dkt. No. 132-2

11

United States District Court
Northern District of California

10

at Fig. IA.7. Panel A).

12

Third, even assuming some of Griffin’s conclusions about inflationary underwriting cast

13

suspicion on the industry as a whole, they do not as to Wells Fargo. For instance, Griffin

14

asked: “Are originators inflating historical NOIs to justify overstating underwritten NOI?” He

15

answered “yes.” Griffin observed that lenders who tended to inflate servicer-reported NOI also

16

tended to inflate first-year realized NOI: “[O]riginators who engage in more inflation of

17

historical financials are proportionally more likely to have income fall short of underwritten”

18

(Dkt. No. 132-2 at 18). Griffin was speaking about the industry as a whole. But, again, Wells

19

Fargo’s particular inflation did not appear excessively risky. An industry-wide correlation

20

does not permit this order to infer that defendants wronged the market (ibid.). Likewise, a

21

September 2020 report by Fitch Ratings that detailed lower NOIs in 2019 than in 2018 says

22

nothing specific about Wells Fargo’s underwriting (Amd. Compl. ¶ 143).

23

Fourth, Griffin wrote, “[O]riginator overstatement is highly correlated (0.903) with the

24

proportion of loans that experience[d] distress from April 2020 to April 2021” (id. at 4). Given

25

the global pandemic, Griffin’s observation makes sense. Logically, businesses strapped with

26

too-big loans will stumble first. Griffin’s observation does not necessarily indicate that Wells

27

Fargo’s underwriting lacked discipline.

28
15

1

Fifth, Griffin explored and eliminated “other plausible reasons why actual NOI might be

2

lower in the first year of a loan than in subsequent years” (Dkt. No. 116 at 10; Amd. Compl. ¶

3

108). Some plausible reasons included differing business models (Dkt. No. 132-2 at 13–14);

4

randomness (id. at 14); “cash flow volatility” (id. at 21); and different property types or classes

5

(id. at 15). He also rejected the theory that the interest rates associated with the CMBSs

6

“priced” the risk accurately (id. at 22–23). This order does not question Griffin’s findings on

7

the industry writ large. It merely notes that his analysis fails to show that Wells Fargo

8

systematically issued reckless loans. To repeat, in about two-thirds of Wells Fargo loans, the

9

lender does not appear to have inflated first-year realized income. Plaintiff has failed to
establish why approximately 32% inflation (in servicer-reported) or 28% (in first-year realized)

11

United States District Court
Northern District of California

10

amounted to rash lending. The complaint has not established falsity on this score.

12

Sixth, the complaint pleads that Wells Fargo had certain incentives to originate and

13

sponsor risky loans (to charge higher fees, to entice borrowers who wanted larger loans, and to

14

become the biggest commercial lender) (Amd. Compl. ¶¶ 157–58). Griffin endorsed these

15

motives (Dkt. No. 132-2 at 8, 26–27). The complaint does not plead, however, that these

16

motives bore out in the riskiness of Wells Fargo’s underwriting. Furthermore, Wells Fargo

17

originated certain loans that it did not sponsor into CMBS trusts. These outstanding loans

18

remained on the bank’s books. As the complaint alleges, the standards for underwriting

19

matched those used for originating commercial loans. This order accepts defendants’ point

20

that if borrowers’ promises to pay were systematically worthless, defaults would leave Wells

21

Fargo high and dry. That Wells Fargo held onto some of these allegedly risky loans cuts

22

against motive (Br. 24; see, e.g., Amd. Compl. ¶¶ 152, 241–42).

23

This order also notes that Griffin created the second chart above for a reason that neither

24

side clearly explained. Griffin wrote that the chart was intended to explore whether cash-flow

25

volatility explained the apparently-inflationary underwriting. Griffin concluded, “Inconsistent

26

with overstatement being driven by the volatility of property income, originators with the

27

greatest proportion of overstated loans have much lower levels of understated loans” (Dkt. No.

28

132-2 at 21). In other words, lenders were not inadvertently inflating business’ NOIs simply
16

1

because business’ cash flow fluctuated wildly. Rather Griffin’s results showed lenders tended

2

to inflate underwritten loans more while deflating less, and vice versa. This point does not aid

3

plaintiff’s cause since it says nothing about Wells Fargo’s particular underwriting practices.

4

A few decisions from our court of appeals have addressed the need to establish standards
against which the district court can evaluate allegedly false or misleading statements. In re

6

GlenFed, Incorporated Securities Litigation found falsity adequately alleged in part because

7

the complaint had stated that “loan underwriting and monitoring policies were inadequate” and

8

because internal documents agreed that “internal controls, policies and procedures need[ed]

9

improvement.” 42 F.3d 1541, 1550 (9th Cir. 1994). Internal reports served as the yardstick for

10

falsity. This permitted an inference that the standards were inadequate. (In that case, plaintiffs

11

United States District Court
Northern District of California

5

also alleged abandonment of underwriting standards, but the decision appears to have credited

12

allegations of inadequate standards as well.) Following similar logic, In re Vantive

13

Corporation Securities Litigation held that the complaint had not pleaded certain statements as

14

false or misleading. 283 F.3d 1079, 1086–87 (9th Cir. 2002) (partially abrogated on other

15

grounds as recognized in S. Ferry LP, No. 2 v. Killinger, 542 F.3d 776, 784 (9th Cir. 2008)).

16

Relevant here, Vantive had stated that the growth and performance of its sales team was “on

17

plan” and “extremely strong.” Id. at 1086–87. The complaint alleged that the company had

18

not, however, been able “to adequately train its new direct sales persons.” Id. at 1086. The

19

decision held: “the complaint le[ft] unclear what it would mean for Vantive to ‘adequately

20

train’ an employee.” Ibid. The complaint, in other words, failed to set forth a benchmark

21

showing why the statements were misleading.

22

In re Vantive most resembles our facts. The complaint has not supplied any standard

23

under which this order can infer “what” the inflation rates seen “would mean” about the truth

24

of Wells Fargo’s underwriting practices. Ibid.; see also Maiman v. Talbott, 2010 WL

25

11421950, at *4 (C.D. Cal. Aug. 9, 2010) (Judge Andrew J. Guilford) (complaint failed for

26

lack of “objective benchmark”). Therefore, this order must evaluate context and apply

27

“common sense.” Ashcroft v. Iqbal, 556 U.S. 662, 679 (2009). This order accepts as true that

28

Griffin found inflationary underwriting with respect to servicer-reported NOI. The complaint,
17

1

however, has not pleaded any plausible reason to think first-year realized NOI represents a less

2

important barometer. On the contrary, first-year realized NOI tested underwriters’ predictions

3

and showed that they were more often deflationary than inflationary. These findings do not

4

constitute particularized allegations that Wells Fargo’s statements were false or misleading.

5

Finally, the complaint alleges that Wells Fargo altered names and addresses associated

6

with the lenders who received inflated loans, in order to disguise the inflation in servicer-

7

reported NOI (Amd. Compl. ¶ 95). Flynn’s interview with The Intercept also details the point.

8

When a reporter asked, banks are “changing the address and they’re changing the name of the

9

property. . . Am I right?” Flynn answered, “Yes,” but did not specifically name Wells Fargo.
(The piece, however, clearly referred back to Flynn’s original analysis, which dove deep on

11

United States District Court
Northern District of California

10

Wells Fargo’s specific loans. It also named Ladder Capital, an alternative asset manager that

12

Wells Fargo funded via a C&I loan.) Flynn also allegedly observed a 95% correlation between

13

altered addresses/names and inflated financial figures. Defendants are correct that the

14

complaint does not allege who, the loan applicant or Wells Fargo, wrote down the names and

15

addresses on the “loan documents” in question. Flynn does not describe why he knew or

16

believed that Wells Fargo itself changed the names and/or addresses. These allegations do not

17

plausibly support an inference of the inflationary scheme (id. n.27, ¶¶ 85, 141, 180).

18
19
20

The complaint fails to plead false or misleading statements regarding underwriting.
(ii)

LTV Ratios.

This order turns next to Shrewsberry’s statement on June 10, 2020: “on an LTV basis,

21

oh, gosh, 90% — more than 90% of that book has less than 70% loan-to-value based on our

22

own underwriting of that” (id. ¶ 397 (emphasis in the original)). The complaint has not

23

adequately alleged that this statement ranked as false or misleading.

24

A confidential witness (CW 1) offered the relevant insights in the complaint. Our court

25

of appeals does not require that a complaint name sources “so long as the sources are described

26

with sufficient particularity to support the probability” that they would “possess the

27

information alleged and the complaint contains adequate corroborating details.” In re Daou

28

Sys., 411 F.3d 1006, 1015 (9th Cir. 2005) (cleaned up). In re Daou found it adequate to
18

1

“number each witness and describe his or her job description and responsibilities.” Id. at 1016.

2

During the class period, CW 1 worked out of Phoenix, Arizona, as Executive Vice President

3

(EVP) responsible for commercial lending and Mountain Region Division Manager (2017 to

4

early 2019) as well as EVP and Commercial Banking Market Executive (early 2019 to

5

September 2020). The complaint also alleges details of CW 1’s duties and responsibilities (id.

6

¶ 210).

7

In 2018 and 2019, CW 1 allegedly began receiving requests “a couple times a month”

8

from supervisee loan officers “to make commercial loans with 80% LTV ratios” (id. ¶ 211).

9

The bank’s “commercial mortgage policy required a 75 percent LTV ratio,” however (id. ¶
213). When CW 1 refused the requests, supervisee loan officers allegedly responded, “Wait a

11

United States District Court
Northern District of California

10

minute: My colleagues in California are getting that done. Why can’t we do that?” (id. ¶ 211).

12

CW 1, however, learned this information third-hand. See In re Daou, 411 F.3d at 1015

13

(cleaned up). Nothing in the complaint corroborated those specific remarks. CWs 2, 4, and 5,

14

for instance, worked out of different Wells Fargo locations from CW 1; CW 3 worked out of

15

an undisclosed location. All other CWs simply attested to the pressure they felt to close loans

16

and make deals (id. ¶¶ 75–76, 80–81, 462). These general statements did not corroborate CW

17

1’s statement about California lending practices. Nor did other allegations in the complaint.

18

A complaint need not allege first-hand knowledge, but due to double hearsay and lack of

19

corroboration, CW 1’s allegations were not “sufficiently reliable [or] plausible” to show that

20

Shrewsberry’s statement on LTV ratios were false or misleading. Lloyd v. CVB Fin. Corp.,

21

811 F.3d 1200, 1208 (9th Cir. 2016). (Another way plaintiff frames CW 1’s allegations about

22

LTV ratios is as a deviation from Wells Fargo’s commercial underwriting standards (Opp. Br.

23

at 9). For the reasons stated, this also does not convince.)

24
25

(iii)

Credit Reserves and Impairments.

Since the complaint fails to plead reckless or inflationary underwriting standards, it has

26

not pleaded facts adequate to suggest that loan-reserve shortfall was “foreseen or foreseeable.”

27

In re Wells Fargo Sec. Litig., 12 F.3d 922, 927 (9th Cir. 1993) (partially abrogated on other

28

grounds by statute on other grounds) (cleaned up). True, plaintiff’s cited case, Maiman v.
19

1

Talbott, found misstatements about loan reserves adequately pleaded even while dismissing

2

alleged misstatements about lending practices. 2010 WL 11421950, at *4. There, however,

3

the complaint specifically alleged data showing that relevant loan liability had ballooned and

4

that the company had cut its percentage of loan reserves. Finally, the company allegedly had

5

failed to heed internal warnings about increasing risk profile for relevant loans. See ibid. No

6

data or admissions specific to loan reserves appears herein.

7
8
9
10

Since this order finds lacking the allegations of false or misleading statements, it does not
reach materiality, scienter, or loss causation.
2.

INCORPORATION BY REFERENCE.

Plaintiff objects to this order considering defendants’ cited portions of Wells Fargo’s

United States District Court
Northern District of California

11

prospectus (Opp. Br. 11; see Br. 22). The prospectus relates to Wells Fargo’s registration

12

statement for Wells Fargo Commercial Mortgage Trust 2017-C38 (Amd. Compl. ¶¶ 156, 220).

13

The complaint cites this prospectus more than ten times. It does so, at one point, to show that

14

underwriters should not have been “re-underwriting” historical financial figures but merely

15

complying with origination standards (id. ¶ 223). Plaintiff also introduced a further portion of

16

the prospectus dealing with what operating expenses must be included during underwriting

17

(Opp. Br. at 11). Plaintiff’s and defendants’ reasons for citing the prospectus — to show that

18

Wells Fargo’s underwriting was either risky or conservative — substantially matched (Amd.

19

Compl. ¶¶ 224–31; Br. at 2–3, 22).

20

More than ten references ranks as extensive. Neither side, moreover, appears to contest

21

the prospectus’ provenance. This order finds those portions of the prospectus cited by both

22

sides incorporated by reference and “consider[s]” them as “documents whose contents are

23

alleged in a complaint and whose authenticity no party questions.” Parrino v. FHP, Inc., 146

24

F.3d 699, 705 (9th Cir. 1998) (as amended (July 28, 1998)) (cleaned up). This order therefore

25

does not reach whether the remaining over-900 pages are incorporated by reference. See

26

Khoja, 899 F.3d at 1002.

27
28

Defendants did not object to articles that appeared in The Intercept or ProPublica, both
of which this order deems incorporated by reference.
20

CONCLUSION

1
2

To the extent stated, the motion to dismiss is GRANTED. By JUNE 6, 2022, AT NOON,

3

plaintiff may seek leave to amend the dismissed claims with a motion noticed on the normal

4

35-day calendar. Plaintiff must plead its best case. Any motion should affirmatively

5

demonstrate how the proposed complaint corrects the deficiencies identified in this order, as

6

well as all other deficiencies raised in defendants’ motion but not addressed herein. The

7

motion should be accompanied by a redlined copy of any proposed amendment.

8
9

IT IS SO ORDERED.
Dated: May 6, 2022.

10
WILLIAM ALSUP
UNITED STATES DISTRICT JUDGE

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
21