WILLIAM ALSUP, District Judge.
This certified consumer class action challenges hundreds of millions of dollars in overdraft fees imposed on depositors of Wells Fargo Bank, N.A. through allegedly unfair and fraudulent business practices. This order is the decision of the Court following a two-week bench trial.
Overdraft fees are the second-largest source of revenue for Wells Fargo's consumer deposits group, the division of the bank dedicated to providing customers with checking accounts, savings accounts, and debit cards. The revenue generated from these fees has been massive. In California alone, Wells Fargo assessed over $1.4 billion in overdraft penalties between 2005 and 2007. Only spread income—money the bank generated using deposited funds—produced more revenue.
This action does not challenge the amount of a single overdraft fee (currently $35). That is accepted as a given. Rather, the essence of this case is that Wells Fargo has devised a bookkeeping device to turn what would ordinarily be one overdraft into as many as ten overdrafts, thereby dramatically multiplying the number of fees the bank can extract from a single mistake. The draconian impact of this bookkeeping device has then been exacerbated through closely allied practices specifically "engineered"—as the bank put it—to multiply the adverse impact of this bookkeeping device. These neat tricks generated colossal sums per year in additional overdraft fees, just as the internal bank memos had predicted. The bank went to considerable effort to hide these manipulations while constructing a facade of phony disclosure. This order holds that these manipulations were and continue to be unfair and deceptive in violation of Section 17200 of the California Business and Professions Code. For the certified class of California depositors, the bookkeeping device will be enjoined and restitution ordered.
Plaintiffs commenced this action in November 2007, alleging violations of the "unfair" and "fraudulent" restrictions of Section 17200. Two of Wells Fargo's business practices were initially targeted: (1) a high-to-low "resequencing" practice, challenged herein, and (2) an "including and deleting" practice, which plaintiffs no longer challenge.
The "resequencing" class, however, survived for trial. This class was defined as (Dkt. No. 98):
Plaintiffs were allowed to conduct a new restitution study covering Wells Fargo transaction data for the entire "resequencing" class period. Following the completion of this study, Wells Fargo again moved for summary judgment and class decertification (Dkt. No. 292). These motions were denied.
The instant order follows a two-week bench trial that commenced on Monday, April 26, 2010, and concluded on Friday, May 7. Following the close of evidence, both sides submitted lengthy proposed findings of fact and conclusions of law, followed by responses (Dkt. Nos. 452-55). The undersigned also denied without prejudice a motion for judgment on partial findings submitted by Wells Fargo during trial and allowed the bank to reargue its points in its proposed findings of fact (Dkt. Nos. 417, 446). Closing arguments were heard on the morning of July 9.
Rather than merely vet each and every finding and conclusion proposed by the parties, this order has navigated its own course through the evidence and arguments, although many of the proposals have found their way into this order. Any proposal that has been expressly agreed to by the opposing side, however, shall be deemed adopted (to the extent agreed upon) even if not expressly adopted herein. It is unnecessary for this order to cite the record for all of the findings herein. Citations will only be provided as to particulars that may assist the court of appeals. In the findings, the phrase "this order finds. . ." is occasionally used to emphasize a point. The absence of this phrase, however, does not mean (and should not be construed to mean) that a statement is not a finding. All declarative statements set forth in the findings of fact are factual findings.
1. The core of this controversy is a bookkeeping device adopted by the bank called "high-to-low resequencing" that transforms one overdraft into as many as ten overdrafts—ten being the voluntary limit the bank imposed on what could otherwise be an almost limitless prospect. The bank instituted this device for California accounts in April 2001 and then soon magnified its impact through closely allied practices. What now follows is an explanation of the bookkeeping device and how it changed overdrafting at Wells Fargo.
2. "Posting" is the procedure followed by all banks to process debit items presented for payment against accounts. During the wee hours after midnight, the posting process takes all debit items presented for payment during the preceding business day and subtracts them from the account balance. These items will typically be debit-card transactions and checks (plus a few other occasional items described below). If the account balance is sufficient to cover all such debit items, there will be no overdrafts regardless of the bookkeeping method used. If, however, the account balance is insufficient to cover all such debit items, then the account
3. Prior to April 2001, Wells Fargo used a low-to-high posting order, as did most banks (then and now). Low-to-high posting meant that the bank posted settlement items from lowest-to-highest dollar amount. Low-to-high posting paid as many items as the account balance could possibly cover and thus minimized the number of overdrafts. This was because the smallest purchases were always deducted from the customer's checking account first and the balance was used up as slowly as possible.
4. This changed in April 2001. Then, Wells Fargo did an about-face in California and began posting debit-card purchases in highest-to-lowest order. The reversal of the bank's previous low-to-high posting order had the immediate effect of maximizing the number of overdraft fees imposed on customers. This was exactly the reason that the bank made the switch.
5. To illustrate, assume that a customer has $100 in his account and uses his debit card to buy ten small items totaling $99 followed by one large item for $100, all of which are presented to the bank for payment on the same business day. Using a low-to-high posting order, there would be only be one overdraft—the one triggered by the $100 purchase. Using high-to-low resequencing, however, there would be ten overdrafts—because the largest $100 item would be posted first and thus would use up the balance as quickly as possible. Scenarios very much like this happened to plaintiffs Veronica Gutierrez and Erin Walker, as will be shown momentarily.
6. The switch in April 2001 to high-to-low posting in California was followed by two closely allied practices, both intentionally "engineered"—to use the bank's own term at the time—to amplify the overdraft-multiplying effect of high-to-low ordering: (1) a switch to commingling of debit-card purchases with checks and automated clearing house ("ACH") transactions in December 2001, and (2) the deployment of a secret "shadow line" in May 2002 to authorize debit-card purchases into overdrafts.
7. Regarding commingling, before December 2001, all debit-card purchases were posted prior to checks, and all checks were posted prior to ACH transactions. While transactions for each transaction type were already being resequenced in high-to-low order (since April 2001), the different transaction types were posted separately.
8. In December 2001, however, Wells Fargo began commingling debit-card purchases, checks, and ACH transactions together and posting the entire group from highest-to-lowest dollar amount. This amplified the overdraft-multiplying effect of high-to-low posting. Checks and ACH transactions—which tended to be the larger items—now consumed the account balance even faster than if all debit-card transactions had been deducted first (debit-card purchases typically being smaller).
9. For the commingling change, the "before and after" looked like this (high-to-low posting having already been instituted in April 2001):
10. The last step in the three-step plan was executed in May 2002. Wells Fargo implemented a practice involving a secret bank program called "the shadow line." Before, the bank declined debit-card purchases when the account's available balance was insufficient to cover the purchase amount. After, the bank authorized transactions into overdrafts, but did so with no warning that an overdraft was in progress. Specifically, this was done without any notification to the customer standing at the checkout stand that the charge would be an overdraft and result in an overdraft fee. Thus, a customer purchasing a two-dollar coffee would unwittingly incur a $30-plus overdraft fee. (This very scenario happened to plaintiff Walker.) Internally, Wells Fargo called this its "shadow line," as in shadow "line of credit." The amount of the credit ceiling per customer was and still is kept secret. Again, customers were not even alerted when shadow-line extensions were made to them—until it was too late and many overdraft fees were racked up. In this program, the bank correctly expected that it would make more money in overdraft fees than it would ever lose due to "uncollectibles" (i.e., overdrafts that were never paid back).
11. Ms. Gutierrez opened a checking account and a savings account at the Rancho Cucamonga branch of Wells Fargo in San Bernadino County on October 25, 2002. She was eighteen at the time.
12. Ms. Gutierrez had never held a bank account in her own name prior to opening these two accounts. She had also never used a debit or ATM card. This was her first real banking experience.
13. A Wells Fargo customer service representative ("CSR") assisted Ms. Gutierrez in opening her checking and savings accounts. The CSR provided Ms. Gutierrez with various Wells Fargo documents during this process.
14. One such document given to Ms. Gutierrez was a Wells Fargo brochure entitled "Checking, Savings and More" (TX 89). Certain portions of this brochure were shown to Ms. Gutierrez by the CSR, including sections highlighting the features of "Student Checking," "Wells Fargo Advantage Checking," and "Wells Fargo Cards" (Tr. 373-76). The CSR highlighted these sections after asking Ms. Gutierrez what kind of account she was interested in opening and learning that Ms. Gutierrez was going to be a college student.
15. With respect to the "Wells Fargo Cards" section of the brochure, Ms. Gutierrez reviewed the subsections entitled "overdraft protection," "ATM card," "ATM & Check Card," and "Visa Credit Cards" while at the branch (id. at 374-76).
17. The "ATM & Check Card" subsection of the brochure included the following statements regarding Wells Fargo debit cards (TX 89) (emphasis added):
18. After reviewing the "Wells Fargo Cards" section of the brochure, Ms. Gutierrez asked the CSR for guidance about how to use an ATM and debit card (Tr. 381). The CSR told Ms. Gutierrez that a debit card could be used in any store where a MasterCard logo was shown (the bank has since switched to VISA-based debit cards). The CSR then assisted Ms. Gutierrez in setting up a PIN number to use for ATM transactions.
19. The CSR also mentioned to Ms. Gutierrez that she would have overdraft protection through the savings account she was opening. Ms. Gutierrez was not informed by the CSR, however, as to what "having overdraft protection" meant (Tr. 376). She had no idea at the time.
20. Ms. Gutierrez was also provided with a Wells Fargo "Welcome Jacket" while opening these accounts (id. at 377; TX 86). The front of the Welcome Jacket stated "Welcome to Wells Fargo" and "Thank You for Opening Your Account With Us." Inside the Welcome Jacket was a pocket to hold additional materials.
21. The Welcome Jacket provided to Ms. Gutierrez contained a copy of Wells Fargo's then-current Consumer Account Agreement ("CAA"), effective April 1, 2002 (TX 13). The CAA was 69 pages in length and was densely packed with single-spaced text in ten-point font.
22. The CSR did not go through any of the 69 pages of content in the CAA with Ms. Gutierrez when her accounts were opened at the branch (Tr. 380). The CSR also failed to inform Ms. Gutierrez about the bank's resequencing practices, including the order in which debit-card purchases would be posted against her checking account (ibid.).
23. Also included in the Welcome Jacket given to Ms. Gutierrez was a copy of Wells Fargo's then-current Consumer Account Fee and Information Schedule (id. at 379; TX 14). This document, which included 37 pages of densely packed text in ten-point font, detailed the amount charged by Wells Fargo for various types of account-related fees, including overdraft and returned-item (e.g., bounced check) fees.
24. Ms. Gutierrez only briefly flipped through the hundred-plus pages of text in the CAA and Consumer Account Fee and Information Schedule while at the branch (Tr. 380).
25. Finally, before leaving the branch, Ms. Gutierrez signed at least one Wells Fargo document and received a printout of the deposits she had made to open her accounts (id. at 376-78). By the end of the process, Ms. Gutierrez had acquired a Wells Fargo checking account, savings account, and an ATM and debit card (hereinafter simply referred to as a debit card). She also applied for a Wells Fargo student VISA credit card.
26. Ms. Gutierrez received her debit card in the mail sometime after opening her checking and savings accounts (id. at 381-82). She also signed up for Wells Fargo's "online banking" service, which allows customers to view their available balance and account statements over the internet (ibid.).
28. After a few years, however, Ms. Gutierrez stopped using a check register. She also stopped verifying receipts against her account statements. This was because of "inconsistencies" between what Wells Fargo was reporting to her regarding her account balance and what she was "keeping track of, personally" (id. at 385). Instead, Ms. Gutierrez began relying on Wells Fargo's online banking information and toll-free customer service phone numbers to obtain her available balance and to check which transactions had been posted to her account (ibid.).
29. On Thursday, October 5, 2006, Ms. Gutierrez walked into a Subway sandwich shop. She spent $11.27 and used her Wells Fargo debit card to pay for the meal (TX 1). That same day, she made three purchases at AutoZone, an auto-parts store where she was then employed (Tr. 395-96). The three purchases were for $47.99, $17.23, and $3.23 (TX 1).
30. For whatever reason, Ms. Gutierrez returned one of the three AutoZone purchases—specifically, the purchase in the amount of $17.23. The return was made on the same day as the purchase (ibid.).
31. The next day, Friday, October 6, Ms. Gutierrez went to an IHOP restaurant. She used her Wells Fargo debit card to pay for her meal, totaling $26.51. Ms. Gutierrez also went to a hamburger house that same day called Farmer Boys and purchased a meal with her debit card totaling $8.10. Finally, that evening, Ms. Gutierrez used her Wells Fargo debit card at a Bank of America ATM at the Hollywood Bowl to withdraw $22.00 from her Wells Fargo checking account (ibid.). Since a non-proprietary ATM was used, Ms. Gutierrez was charged a two-dollar transaction fee by Wells Fargo.
32. That weekend, Ms. Gutierrez went to an Albertsons supermarket and spent $74.39. She used her Wells Fargo debit card to complete the purchase (ibid.).
33. Finally, on Tuesday, October 10, Ms. Gutierrez used Wells Fargo's online banking service to initiate and complete an "online transfer" of $80.00 from her Wells Fargo checking account to the checking account of a family member (ibid.; Tr. 400). That same day, a check written by Ms. Gutierrez in the amount of $65.00 was presented to the bank for payment (TX 449).
These transactions are shown as follows:
$11.27 Subway $26.51 IHOP $74.39 Albertsons $80.00 online transfer of
$47.99 Autozone $ 8.10 Farmer Boys funds to family member
$17.23 Autozone $22.00 ATM cash
withdrawal
$ 3.23 Autozone $2.00 fee for non-Wells $65.00 check presented
($17.23) return of Fargo ATM to Wells Fargo
Autozone purchase for payment
Thurs. Oct. 5 Fri. Oct. 6 Weekend Oct. 7-8 Tues. Oct. 10
34. None of the debit-card purchases was declined or otherwise rejected by Wells Fargo. In fact, all were allowed by the bank. There were no indications given to Ms. Gutierrez by Wells Fargo that any of her purchases, withdrawals, or transfers had been made against insufficient funds or would result in overdraft fees being assessed.
35. The very next day, on Wednesday, October 11, Wells Fargo assessed $88 in overdraft fees against Ms. Gutierrez's checking account (TX 1, 2). These fees were triggered by four overdrafts caused by the two smallest AutoZone purchases made by Ms. Gutierrez on October 5, her Subway purchase made on October 5, and her Farmer Boys purchase made on October 6. Each overdraft carried a penalty of $22. (The overdraft fee today is $35.)
36. The four debit-card purchases that triggered overdraft fees were among twelve transactions that posted against Ms. Gutierrez's checking account after the close of business day on Tuesday, October 10 (TX 1). These transactions are listed below in the exact order that they were posted against her checking account (ibid.). Amounts that were debited from her account are represented below in parentheses. According to Ms. Gutierrez's account statement, her checking account balance prior to the posting of these transactions was $316.90 (ibid.).
Transaction Description Amount Balance
1. Return of debit-card purchase $17.23 $334.13
made on 10/5/2006
at AutoZone
2. Online transfer of funds ($80.00) $254.13
to another account
3. ATM withdrawal on ($22.00) $232.13
10/6/2006 at a non-Wells
Fargo ATM
4. Non-Wells Fargo ATM ($ 2.00) $230.13
fee
5. Debit-card purchase ($74.39) $155.74
made on 10/7/2006 at
Albertsons supermarket
6. Check # 1103 ($65.00) $ 90.74
7. Debit-card purchase ($47.99) $ 42.75
made on 10/5/2006 at
AutoZone
8. Debit-card purchase ($26.51) $ 16.24
made on 10/6/2006 at
IHOP restaurant
9. Debit-card purchase ($17.23) -$ 0.99**
made on 10/5/2006 at
AutoZone
10. Debit-card purchase ($11.27) -$ 12.26**
made on 10/5/2006 at
Subway restaurant
11. Debit-card purchase ($ 8.10) -$ 20.36**
made on 10/6/2006 at
Farmer Boys restaurant
12. Debit-card purchase ($ 3.23) -$ 23.59**
made on 10/5/2006 at
AutoZone
** $22 overdraft fee assessed.
37. The transactions were posted by Wells Fargo in the following order: First, all credits were posted to Ms. Gutierrez's checking account. This is why the returned AutoZone purchase was the first of the twelve items to post. Second, transactions representing cash withdrawals (or their equivalent) were posted from highest-to-lowest dollar amount. The online transfer of funds, ATM withdrawal, and associated two-dollar transaction fee were part of this "priority posting" group (as called by the bank). Third, and most significantly, debit-card transactions, checks, and ACH transactions were commingled together and posted from highest-to-lowest dollar amount. The parties agree that this is how Wells Fargo posted transactions during the class period (Dkt. No. 448).
38. The adverse impact of Wells Fargo's high-to-low bookkeeping switch is clearly shown by these transactions. If the last eight transactions had been posted by Wells Fargo from lowest-to-highest dollar amount, only one of the debit-card purchases—the $74.39 purchase at Albertsons supermarket—would have been an overdraft. Thus, only one overdraft fee ($22), as opposed to four overdraft fees ($88), would have been assessed.
39. The adverse impact of commingling is also well illustrated. Assuming that debit-card
40. Finally, if Wells Fargo had posted Ms. Gutierrez's debit-card transactions in chronological order (or as close to chronological order as possible), the $74.39 purchase at Albertsons supermarket—which was made on October 7—would have been posted last. Under this scenario, only one overdraft fee ($22), as opposed to four overdraft fees ($88), would have been assessed.
41. At trial, Wells Fargo noted that it had instituted a one-dollar "overdraft courtesy threshold" during the class period (Tr. 356). It is clear—when examining Ms. Gutierrez's account statement—that this minuscule threshold was not in place when she incurred these overdrafts (TX 1).
42. With a one-dollar courtesy threshold, a checking-account holder is allowed to overdraw his or her account up to the one-dollar threshold amount before the account is considered "overdrafted" and a fee is assessed. At an unknown date during the class period, this threshold increased from zero to one dollar (Tr. 356). The fact that Ms. Gutierrez was assessed a $22 overdraft fee for overdrafting her account by $0.99 shows that the one-dollar courtesy threshold had not yet been instituted. Despite being touted by the bank, it provided no benefit to her.
43. Ms. Gutierrez first learned about these four overdraft fees via an insufficient-funds notice sent by Wells Fargo using regular "snail mail" (TX 2). While it is unclear when Ms. Gutierrez actually received the notice, it is likely that she did not receive it until a few days after the fees had been assessed. The notice informed Ms. Gutierrez that the last four debit-card purchases listed above had been presented to Wells Fargo for payment on October 10 and were paid into overdraft. It also informed Ms. Gutierrez that $88 in fees had been imposed on her.
44. The notice further instructed Ms. Gutierrez to deposit $111.59 into her checking account to cover the items paid into overdraft and their corresponding fees. Thus, despite overdrafting her checking account by only $23.59, Ms. Gutierrez was required to pay $111.59 to Wells Fargo to return her account to a zero balance.
45. The high-to-low bookkeeping devices employed by Wells Fargo caused Ms. Gutierrez to be assessed four overdraft fees ($88) rather than one overdraft fee ($22) for exactly the same conduct. These manipulations mathematically ensured that the larger items would consume her balance faster, thereby allowing Wells Fargo to impose more overdraft fees.
46. When she received the insufficient-funds notice, Ms. Gutierrez had no understanding of what the $88 fee meant, or how the amount had been calculated (Tr. 388). Ms. Gutierrez also did not understand how overdraft fees could be assessed if a debit-card purchase was "approved" by Wells Fargo while making purchases at a store or restaurant. She had not been adequately apprised of the shadow-line extension.
48. Over the next two days, October 12 and 13, two additional overdraft fees were assessed on Ms. Gutierrez by Wells Fargo (TX 1). The fee imposed on October 12 was caused by a debit-card purchase of $5.68 made by Ms. Gutierrez on October 9 at a Jack in the Box restaurant. The fee on October 13 was caused by a debit-card purchase of $20.01 made by Ms. Gutierrez on October 10 at an ExxonMobil gas station. Unlike the prior overdraft fees, however, the October 13 overdraft fee was $33 (rather than $22)—a result of Wells Fargo's two-tiered fee schedule that imposed harsher fees on account holders who had overdrafted their accounts on multiple occasions within a twelve-month period.
49. To sum up the damage, by October 13, Ms. Gutierrez had overdrafted her account by only $49.28. Due to the multiplication of overdraft penalties, however, she was required to deposit almost four times that amount—$192.28—into her checking account to return it to a zero balance (TX 4). Stated differently, three-fourths (in round terms) of her $192.28 checking-account deficit consisted of Wells Fargo overdraft penalties.
50. Had the bank not commingled Ms. Gutierrez's debit-card purchases with her checks, she would have only had to pay $126.28 (rather than $192.28) to return her checking account balance to zero. Similarly, if Ms. Gutierrez's transactions had been posted in low-to-high order or chronologically, she would have only had to pay $126.28 to return her checking account to a zero balance.
51. Ms. Gutierrez received a payroll payment via direct deposit from her employer, AutoZone, on October 13. The deposit, which went straight into her Wells Fargo checking account that evening, was in the amount of $412.90. While this returned Ms. Gutierrez's checking account to a positive balance, over a third of this payroll payment went directly into Wells Fargo's pockets to cover the overdraft penalties (TX 1).
52. While the high-to-low bookkeeping device did not directly affect the overdraft fees incurred by Ms. Gutierrez on October 12 and 13 (since her account was already overdrafted on those days), it catapulted her into a higher overdraft bracket and upped her per-occurrence fee. Moreover, as stated, the high-to-low posting of Ms. Gutierrez's debit-card purchases dramatically increased the rate at which her balance was consumed, thereby multiplying the number of overdraft penalties she incurred on October 11.
53. Given the costs of Ms. Gutierrez's college education and her limited income at the time, $143 in overdraft fees was a significant amount of money for her to have to pay in overdraft penalties (Tr. 405). Ms. Gutierrez would not have made the four purchases that resulted in overdraft fees if she had been advised or warned that each purchase—some as small as $3.23—would be covered by a secret shadow line of credit and assessed a $22 or $33 overdraft fee (id. at 403-05).
54. Ms. Gutierrez did not benefit in any way, shape, or form from Wells Fargo's high-to-low bookkeeping device.
55. After receiving these fees, Ms. Gutierrez decided not to depend on the available balance information Wells Fargo communicated to her online and over the phone (id. at 406). Instead, she had to "sort of trick [her] brain" to remind herself that she had less money in her checking
56. Even after receiving and reviewing her Wells Fargo account statement for the period covering the transactions discussed above, Ms. Gutierrez could still not decipher the process or order in which the various transactions were posted to her checking account and how her debit-card purchases could lead to multiple overdraft fees. The Court has studied her account statements and finds that it was impossible for her or anyone else to reconstruct how the bank came up with its number of overdrafts.
57. Even if Ms. Gutierrez had meticulously maintained a chronological check register of all of her transactions (as Wells Fargo insists that she and other customers should), it could not have accurately reflected or predicted how the bank would have posted transactions. Stated differently, Ms. Gutierrez had no reason to know that Wells Fargo would resequence her debit-card transactions in a way that would maximize the number of overdraft fees it could assess on her. She would not have known that she was at risk of incurring four overdraft fees on October 10. This is significant, for over and again at trial, bank counsel and bank witnesses waxed eloquent about the virtues of keeping a check register. In reality, the most exacting register would not have told Ms. Gutierrez that she would be hit with four rather than one overdraft fee.
58. Despite being new to banking, Ms. Gutierrez understood the importance of managing her checking account and not spending more than she had. She acknowledged that she made a mistake by overdrafting her account. She did not, however, expect or believe it was fair for Wells Fargo to capitalize upon her mistake by artificially resequencing her debit-card transactions to maximize fee revenue for the bank.
59. Ms. Gutierrez was harmed by this practice. She was deceived by the bank's obfuscation of its practice regarding high-to-low posting and the risk that one overdraft could be transformed into as many as ten overdrafts. She also relied upon the bank's misleading marketing materials that reinforced her natural assumption that debit-card transactions would post chronologically.
60. Ms. Gutierrez's testimony in support of these findings was and remains imminently credible.
61. Erin Walker, a member of the "resequencing class," suffered a similar piling-on encounter. Ms. Walker opened a checking account and a savings account in July 2006 at a Wells Fargo branch in Culver City. Like Ms. Gutierrez, she was around eighteen at the time she opened these accounts and this was her first real banking experience.
62. Ms. Walker chose Wells Fargo because she had seen its ATMs throughout the campus of Arizona State University, where she was about to start school.
63. The Wells Fargo CSR who assisted Ms. Walker in opening her accounts provided her with a Wells Fargo "Welcome Jacket" (Tr. 756-57; TX 18, 84). Ms. Walker did not read the Welcome Jacket while at the branch, but read it at home (Tr. 757-58). In particular, Ms. Walker read the same portion of the Welcome Jacket pertaining to debit cards that Ms. Gutierrez read, which stated that "[e]ach purchase is automatically deducted from your primary checking account" (TX 84).
64. Ms. Walker did not recall receiving a Wells Fargo CAA at the time she opened her account. She also did not recall if her Welcome Jacket contained a copy of a CAA (Tr. 758). This order finds, however, that it was Wells Fargo's customary business
65. During the account-opening process, Ms. Walker signed up for overdraft protection and to receive a Wells Fargo debit card. The debit card arrived in the mail sometime thereafter. While she was at the Wells Fargo branch, the CSR did not instruct Ms. Walker about how to use her debit card (id. at 759). Nevertheless, Ms. Walker ended up using her debit card more often than checks to make purchases (id. at 759-60).
66. Like Ms. Gutierrez, Ms. Walker also signed up for Wells Fargo's online banking services after opening her accounts. She used this service periodically to check her available balance online (id. at 760).
67. On Tuesday, May 29, 2007, Ms. Walker made a $9.66 purchase at Jackson Market in Culver City using her Wells Fargo debit card (TX 21). Two days later, on Thursday, May 31, she used the same debit card to purchase $10.92 of merchandise at a store called Essence of Nature in Los Angeles.
68. The following day, Friday, June 1, Ms. Walker purchased $20.27 worth of gasoline from a Shell gas station using her Wells Fargo debit card (ibid.). That same day, Ms. Walker purchased food items at a restaurant called Sandbags and at Jamba Juice, an establishment that serves smoothies. The purchase at Sandbags totaled $7.52 while the Jamba Juice purchase totaled $4.00. Both purchases were made using her Wells Fargo debit card.
69. Finally, on Saturday, June 2, Ms. Walker went to an ARCO gas station and spent $20.61 using her Wells Fargo debit card. That same day, she also went to two coffee shops—Starbucks and The Coffee Bean & Tea Leaf—and spent $5.55 and $4.10, respectively, using her Wells Fargo debit card.
These transactions are shown as follows:
$9.66 Jackson Market $10.92 Essence of Nature $20.27 Shell gas station $20.61 ARCO gas station
$ 7.52 Sandbags $ 5.55 Starbucks
$ 4.00 Jamba Juice $ 4.10 The Coffee Bean
Tues. May 29 Thurs. May 31 Fri. June 1 Sat. June 2
70. Ms. Walker completed each of these debit-card purchases without any indication from the bank that she lacked sufficient available funds to cover them. At no time did Wells Fargo warn her that she was overdrawn or that any one of these purchases could trigger an avalanche of overdraft penalties from earlier purchases she had already made. Instead, Wells Fargo allowed each transaction to go through with nary a warning.
71. On Tuesday, June 5, Wells Fargo assessed eight overdraft fees against Ms. Walker's checking account. The amount of each fee was $34. In total, Ms. Walker was assessed $272 in overdraft fees.
72. The eight debit-card purchases were posted in high-to-low order against her account (TX 21). Commingled with these debit-card purchases was a check Ms. Walker had written in the amount of $8.47. This check was presented for payment on June 4.
74. The balance of Ms. Walker's checking account before any of these transactions were posted was $36.41. Shown below is an illustration of how the eight debit-card purchases, the written check, and the overdraft-protection transfer posted against her checking account after the close of business day on June 4. As before, items that were debited are represented in parentheses.
Transaction Description Amount Balance
1. PIN-based debit-card ($20.61) $15.80
purchase on 6/2/2007 at
ARCO gas station
2. Debit-card purchase on ($20.27) -$ 4.47**
6/1/2007 at Shell gas
station
3. Overdraft protection $ 3.72 -$ 0.75
transfer from Ms.
Walker's savings account
4. Debit-card purchase on ($10.92) -$11.67**
5/31/2007 at Essence of
Nature
5. Debit-card purchase on ($ 9.66) -$21.33**
5/29/2007 at Jackson
Market
6. Check # 104 presented ($ 8.79) -$30.12**
for payment on 6/4/2007
7. Debit-card purchase on ($ 7.52) -$37.64**
6/1/2007 at Sandbags
restaurant
8. Debit-card purchase on ($ 5.55) -$43.19**
6/2/2007 at Starbucks
9. Debit-card purchase on ($ 4.10) -$47.29**
6/2/2007 at The Coffee
Bean & Tea Leaf
10. Debit-card purchase on ($ 4.00) -$51.29**
6/1/2007 at Jamba Juice
** $34 overdraft fee assessed.
75. As shown, by commingling the eight debit-card purchases with the written check and posting them in high-to-low order, Ms. Walker's available balance was depleted to a negative balance using the least number of transactions possible, allowing the maximum number of overdraft fees to be assessed.
76. Taking into account the availability of $3.72 in overdraft protection funds, if Wells Fargo posted these same transactions in low-to-high order, only three overdraft fees ($101) rather than eight overdraft fees ($274) would have been assessed on Ms. Walker on June 5—a difference of $173. This again illustrates how a high-to-low ordering dramatically increases the number of overdraft items for exactly the same conduct.
77. Similarly, if all debit-card transactions had been posted from low-to-high before the written check, as Wells Fargo used to do in California prior to April 2001, only four overdraft fees would have been assessed (or, alternatively, the check would have been returned, resulting in a $34 non-sufficient funds fee) (TX 353).
78. Finally, if the transactions had been posted in chronological order (or as close to chronological order as possible), Ms. Walker would have been assessed between five and seven overdraft fees, depending upon where the check was posted relative to the debit-card purchases and the time-stamp information for the June 1 purchases.
79. Ms. Walker did not learn that she had been assessed hundreds of dollars in overdraft fees until a few days after the eight fees had been assessed. She found out by using Wells Fargo's online banking services, although insufficient funds notices had been mailed to her mother's address—the address of record on her account—by the bank (Tr. 762-63, 801-02).
80. By that time, a number of additional debit-card purchases had triggered more overdraft fees to be assessed. For example, a $6.60 purchase at The Coffee Bean & Tea Leaf on June 3 was posted after the close of business day on June 5. That transaction triggered an additional $34 overdraft fee to be assessed against Ms. Walker's checking account on June 6.
81. Three other transactions, including an ATM withdrawal of $20, a $20.78 debit-card purchase of gasoline made on June 5 at an ARCO gas station, and a $4.69 debit-card purchase made on June 4 at Coldstone Creamery (an ice cream parlor) posted to Ms. Walker's checking account after the close of business day on June 6. This caused three more $34 overdraft fees to be assessed on June 7. Additionally, Ms. Walker was assessed five dollars for a socalled "Continuous OD Level 2 Charge" by Wells Fargo on June 7. This charge was imposed due to Ms. Walker's failure to pay off her negative balance within a certain number of days.
82. By the end of day on June 7, Ms. Walker's checking account had reached a negative balance of -$516.36. Of this deficit, $413 were Wells Fargo overdraft fees (including continuous overdraft fees). In other words, despite overdrafting her account by one hundred dollars, thanks to high-to-low resequencing and the shadow-line program, Ms. Walker was required to pay Wells Fargo nearly five times that amount to return her checking account to a zero balance.
83. That same day, Ms. Walker made a purchase at a Starbucks coffee shop in the amount of $5.60. She used her Wells Fargo debit-card to complete the purchase. Despite being overdrafted by hundreds of dollars at this point, Wells Fargo nevertheless approved the purchase at the point-of-sale under its shadow line. Ms. Walker was assessed an additional $34 overdraft fee for this purchase on June 11.
84. Finally, for a $10.34 debit-card purchase made by Ms. Walker on June 5 at Jackson Market, she was assessed an additional $34 overdraft fee on June 12, again due to the shadow line of credit.
85. By June 14, Ms. Walker had been assessed six five-dollar continuous overdraft fees and fourteen $34 overdraft fees by Wells Fargo. This amounted to $506 in penalties for overdrafting her checking account by approximately $120.
86. While only a portion of these penalties can be directly tied to the bank's high-to-low posting (since posting order will not change the number of overdrafts assessed once the account is already overdrafted), high-to-low bookkeeping made it much more difficult for Ms. Walker to return her account to a positive balance. Stated differently, by enabling Wells Fargo to assess the maximum number of overdraft
87. Ms. Walker was able to convince Wells Fargo—after calling the bank, meeting in-person with a Wells Fargo representative, and complaining about these fees—to reverse four overdraft fees and four continuous overdraft fees. She was still liable for $350 in penalties (TX 22). After repaying these fees, she closed her Wells Fargo checking and savings accounts (Tr. 784).
88. Like Ms. Gutierrez, Ms. Walker understood the importance of not spending more than she had in her checking account (id. at 788-89). Nevertheless, she acknowledged that she made a mistake and forgot to check her balance prior to making the above purchases (id. at 779). She neither understood nor expected, however, that her purchases would be deducted from her account in a way that would maximize her exposure to overdraft fees. She had no idea that all of her smallest purchases made at a Starbucks or The Coffee Bean would be posted last so as to allow Wells Fargo to charge her a $34 overdraft fee for each cup of coffee purchased.
89. Ms. Walker also did not understand how she could be assessed over $500 in overdraft-related fees for purchases that were "approved" at the point-of-sale, the majority of which were less than ten dollars in value.
90. Ms. Walker was harmed by Wells Fargo's unfair and deceptive business practices. Like Ms. Gutierrez, she was deceived by the bank's non-disclosures regarding high-to-low posting and the risk that one overdraft could be transformed into as many as ten overdrafts. She also reasonably relied upon the bank's misleading marketing materials reinforcing a customer's natural assumption that debit-card transactions would post chronologically.
91. Ms. Walker's testimony in support of these findings was and remains imminently credible.
92. All banks post transactions after the close of every business day. This occurs in the wee hours after midnight. During this process, transactions that were presented for settlement during the day are posted against the customer's checking account (Tr. 593-94). In recent years, this has become a computerized process.
93. Wells Fargo calls its automated system HOGAN. The HOGAN deposit system does not exercise individualized discretion based upon the nature of a customer's transactions. Unlike only a few years ago when individualized discretion was sometimes exercised in a manual sort, Wells Fargo does not prioritize payments, for example, to the Internal Revenue Service versus payments to a wine shop. Rather, HOGAN is designed to sequence and post transactions in a pre-programmed order selected by the bank.
94. When a credit transaction is "posted" during batch posting, the amount of the credit is added to the ledger (i.e., actual) balance of the customer's checking account. Examples of credit transactions include a deposited check, a return of a debit-card purchase, a transfer of funds into a checking account from another source (e.g., a savings account), and the direct deposit of payroll. This increases the balance, although a temporary hold is sometimes placed on deposited checks.
96. All possible types of debit transactions are posted during batch processing, including debit-card transactions, traditional written checks, ACH transactions, cash withdrawals or transfers, cashiers-check purchases, and cashed checks (Tr. 604). As stated, debit-card purchases, checks, and ACH transactions are commingled together and posted in high-to-low order by Wells Fargo. Most banks, by contrast, still use the opposite low-to-high sequencing.
97. Once the funds in an account are consumed, further debits will be overdrafts. When the HOGAN deposit system at Wells Fargo is presented with a debit transaction for posting, two steps are performed: First, the system determines whether the transaction is a "must pay" item. A "must pay" item means exactly that—the bank is obligated to pay it and cannot reject it, even if the customer lacks sufficient funds to cover it. This is because the bank has already authorized the purchase and the merchant has relied on the authorization by letting the customer leave the store with merchandise. Significantly, the vast majority of debit-card purchases are "must pay" items. This is significant because it undercuts the bank's rationale presented at trial for high-to-low posting, as will soon be seen.
98. Second, if the transaction is not a "must pay" item, the HOGAN system determines whether to pay the item or return it unpaid. To make this decision, the HOGAN system first looks at the customer's balance. If there are insufficient funds to cover the transaction, the HOGAN system then looks for overdraft-protection sources—e.g., savings or credit-card accounts that are linked to the customer's primary checking account.
99. If the customer has sufficient funds available through overdraft protection, Wells Fargo automatically transfers funds from the customer's savings account (or charges the customer's credit card) to cover the overdraft amount (Tr. 612-14). A fee, however, is charged every time an overdraft-protection transfer is made (id. at 1147; TX 6, 347, 353, 355). These overdraft-protection fees—though lower than overdraft fees—are also a profit center for the bank (Tr. 1368).
100. Finally, if there are still not enough funds to cover the transaction, the HOGAN system looks to a hidden source of funds that the bank internally calls the "shadow line." The bank computer performs a secret "risk assessment"—akin to a credit analysis—to determine whether the customer would be likely to repay the overdraft amount and corresponding fee if the item is honored (id. at 1564-65). As stated, this hidden amount that the bank is willing to "spot the customer" is the customer's "shadow line"—as in shadow "line of credit." If the bank decides, based upon the shadow line, to pay the item into overdraft, an overdraft fee is then assessed. If, however, the bank elects to return the item unpaid, a non-sufficient funds ("NSF") fee is assessed. During the class period, NSF fees were lower than overdraft fees.
101. Importantly, during the class period, the same shadow line was employed before the posting process to authorize debit-card transactions in the first place. Put differently, rather than declining debit-card purchases when the customer lacked sufficient funds (as the bank did prior to May 2002), the bank approved such transactions with the shadow line without any warning to the customer at
102. The trial record is most telling about the true reasons Wells Fargo adopted high-to-low bookkeeping and the two allied practices—commingling and the shadow line—described herein. Internal bank memos and emails leave no doubt that, overdraft revenue being a big profit center, the bank's dominant, indeed sole, motive was to maximize the number of overdrafts and squeeze as much as possible out of what it called its "ODRI customers" (overdraft/returned item) and particularly out of the four percent of ODRI customers it recognized supplied a whopping 40 percent of its total overdraft and returned-item revenue. This internal history—which is laid bare in the bank's internal memos—is so at odds with the bank's theme of "open and honest" communication and that "overdrafts must be discouraged" that the details will be spread herein.
103. Internal bank memos were presented at trial pertaining to a bank-wide strategic plan called "Balance Sheet Engineering"—or "BSE" for short (see, e.g., TX 36, 61). The documents and testimony surrounding the BSE plan provided clear evidence that the challenged practices were implemented for no other purpose than to increase overdrafts and overdraft fee revenue (Tr. 134-35, 237).
104. Both the commingling of debit-card transactions with checks and ACH transactions and the extension of the shadow line to debit-card purchases were part of Wells Fargo's BSE plan. The commingling change was deemed "BSE Phase 1" while the extension of the shadow line to point-of-sale (POS) debit-card purchases was deemed "BSE Phase 2a" (ibid.; TX 36, 61).
105. The internal BSE memos show that both the commingling change and the adoption of the shadow-line program were intended to capitalize on the overdraft-multiplying effect of high-to-low ordering, which had been recently implemented in California in April 2001, a few months before.
106. A graphical timeline of Wells Fargo's BSE initiatives, taken straight out of a Wells Fargo internal memo from February 2002, is shown below (TX 36). As the internal timeline illustrates, both BSE Phase 1 and BSE Phase 2a were part of a larger suite of OD/NSF (overdraft/non-sufficient funds) initiatives being deployed by the bank. The internal timeline also shows that these initiatives were implemented by Wells Fargo specifically to increase revenue.
Here is the internal bank graphical timeline dated February 2002:
107. Again, this internal timeline picked up just after high-to-low sequencing was deployed in California. As shown in the timeline, Wells Fargo expected the commingling change in combination with the high-to-low change—labeled as BSE Phase 1—to generate an additional $40 million annually in overdraft fees nationwide (TX 36; Tr. 73). This was over and above the millions that the high-to-low change alone was already going to add. These financial projections were echoed in a number of internal spreadsheets pertaining to these initiatives (see, e.g., TX 61).
108. The commingling change was implemented in December 2001. Prior to this date, the bank posted all debit-card transactions from highest-to-lowest dollar amount, then all checks from highest-to-lowest dollar amount, and then all ACH transactions from highest-to-lowest dollar amount. In other words, the different transaction types were posted separately. In December 2001, however, the bank began commingling debit-card transactions with checks and ACH transactions and posting the entire group from highest-to-lowest dollar amount. Since checks and ACH transactions tended to be the larger items, commingling assured that balances would be consumed more rapidly so as to increase the number of overdrafts under a high-to-low ordering. This was the meaning of the $40 million revenue boost predicted by the memo.
109. Consistent with the $40 million projection in the February 2002 document, Wells Fargo Executive Vice-President Ken Zimmerman, who personally took part in the decision-making process for the commingling change, admitted at trial that the bank was well aware that commingling was expected to produce a "significant increase in overdraft income" (Tr. 1422). According to Mr. Zimmerman, the increase in overdraft income was "one of the significant factors in the decision-making" process for the commingling change. Indeed,
110. The undeniable connection between commingling and high-to-low resequencing was also clearly evidenced by internal bank memos. In the February 2002 BSE memo, the bank explained that the December 2001 commingling change was designed "to more-closely mirror true High-to-Low sort order" (TX 36). This is significant because Wells Fargo knew— and its own expert witness, Professor Christopher James, confirmed at trial— that high-to-low posting would "mechanically. . . lead to more overdrafts" than other posting order (Tr. 613, 1863-64). It is a mathematical certainty. Indeed, Mr. Zimmerman, who was personally involved in the bank's decision, admitted that he recommended against a high-to-low posting order due to the adverse impact it would have on customers (id. at 113, 160-61, 340). His objection was overruled. Thus, the additional revenue expected from commingling was premised upon the adverse impact of high-to-low posting.
111. In this connection, the February 2002 memo went on to cheer that "[t]he two most important OD/NSF [revenue] drivers [we]re up significantly" following the December 2001 commingling change (TX 36). These "two most important OD/NSF drivers" were: (1) the percentage of Wells Fargo checking accounts incurring overdraft fees and (2) the number of overdraft items per overdrafted account. This is also clearly shown in another bank slide from the February 2002 memo:
112. Based upon Wells Fargo's own analysis of its California customers, the percentage of checking accounts incurring overdraft fees rose 7.8 percent after the commingling change. As telling, the number of overdraft items per overdrafted account rose 10.1 percent over that same time period (ibid.). The commingling of
113. A separate internal bank memo dated April 2002 provided additional analysis and commentary on the effect of commingling (TX 37). After crunching the numbers over the prior year, the April 2002 analysis stated that the number of overdraft items per occurrence was up strongly due to the fact that debit-card transactions were "now commingled with checks in the processing stream." Yet again we see that Wells Fargo knew and expected that commingling would further amplify the overdraft-multiplying effect of high-to-low posting.
114. In sum, this order finds that Wells Fargo intended to boost its overdraft revenue by $40 million via the December 2001 commingling change, relying upon the overdraft-multiplying effect of high-to-low resequencing to generate this additional revenue.
115. Turning to the shadow line, the BSE timeline in the February 2002 memo also showed the May 2002 extension of the shadow line to debit-card purchases, labeled in the document as "BSE Phase 2a." This was expected by Wells Fargo to generate yet another $40 million annually in overdraft fees nationwide (on top of the $40 million boost expected from commingling) (TX 36).
116. As with the commingling change, extracting more overdraft penalties was the motivation behind the extension of the shadow line to the authorization of debitcard purchases in May 2002. The additional $40 million in overdraft revenue that the bank expected from this change was predicated upon the overdraft-multiplying effect of high-to-low posting.
117. At all relevant times, including now, the vast majority of debit-card purchases followed a two-step process during the class period: authorization and settlement. When a Wells Fargo debit-card holder used his or her card to complete a purchase at a store or restaurant, an authorization request was almost always submitted to the bank then and there. Within a matter of seconds, the authorization request was approved or declined by the bank. An occasional exception to this rule occurred if the purchase was made from a merchant who lacked the capability to submit an electronic authorization request at the time of purchase—a rare event. For these merchants, there was no authorization step, only a settlement step.
118. Whether the authorization request was approved or declined by Wells Fargo, the merchant was notified within a matter of seconds at the point of sale. As stated, approval of an authorization request constituted a promise by Wells Fargo to pay the merchant the authorized amount—i.e., the debit-card purchase became a "must pay" item.
119. If the debit-card transaction was approved, the merchant would obtain payment from the bank via a separate process called "settlement." Once the merchant submitted the debit-card transaction to the bank for settlement, the transaction would be posted during the bank's posting process in the wee hours after midnight, as described above.
120. The time period between authorization and settlement was very short. Wells Fargo's own expert, Dr. Alan Cox, provided direct testimony at trial that for the vast majority of debit-card purchases, the merchant presented the transaction for settlement within two business days. Indeed, according to Dr. Cox, around 90 percent of debit-card purchases were presented within two business days of authorization, and over half actually cleared on the same day they were authorized (Tr. 1782). For PIN-based debit-card purchases (where the user provided her PIN
121. As stated, prior to May 2002, an authorization request for a debit-card purchase would be declined if the customer's checking account lacked sufficient available funds and the customer did not have enough overdraft-protection funds to cover the purchase amount. No fees were charged for a declination. Given the extremely short time frame between authorization and settlement, this meant that any debit-card purchase that was approved by the bank prior to May 2002 was unlikely to be assessed an overdraft fee when posted.
122. After deployment of the shadow line in May 2002, however, Wells Fargo began authorizing previously declined purchases as prospective overdrafts, assessing an overdraft fee for each overdraft. This initiative was called "overdraft via POS" within the bank, meaning the bank had now found a way to rack up overdraft fees for point-of-sale purchases that previously were protected from overdrafts.
123. From the customer's perspective—standing at the checkout counter— the shadow line was completely invisible. The bank would approve the purchase without any warning that an overdraft was in the making.
124. Given the extremely short time frame between authorization and settlement, by approving these debit-card purchases when the customer lacked sufficient funds, Wells Fargo expected a marked uptick in overdraft fees. An increase in the number of overdrafted accounts—upon which the high-to-low resequencing could work its magic to the tune of $40 million in additional revenue—was the intended result. Exactly this occurred.
125. It is true, as Wells Fargo says, that when any specific shadow-line authorization is made, Wells Fargo does not know for sure that an overdraft will occur by the time the item is presented for settlement. It is theoretically possible that an intervening deposit may be made to save the day in the short interval between authorization and settlement. While this may occasionally occur, the bank specifically expected when it deployed the shadow line for debit-card purchases that it would generate $40 million annually in additional overdraft revenue.
126. Yet another powerful item of evidence illustrating the bank's motive with respect to high-to-low posting, the commingling change, and the extension of the shadow line to debit-card purchases was an April/May 2002 email exchange between Mr. Zimmerman and his boss, Les Biller (TX 38, 57).
128. This internal email shows the following: First, it is evident that Wells Fargo decision-makers closely monitored overdraft fee revenue—unexpected declines triggered an immediate investigation into the cause.
129. Second, in sharp contrast to the bank's theme that overdrafts were to be discouraged, bank executives wished to increase them and viewed any decline in overdrafts as a "cause for concern."
130. Third, Wells Fargo's true intentions regarding the shadow-line extension to debit-card purchases, referred to as "B SE 2A" in the email, were laid bare by Mr. Zimmerman. In his own words, it was intended to generate "overdraft via POS"—meaning overdraft via a point-of-sale debit-card purchase—resulting in $24 million in additional overdraft revenue for the remainder of 2002. "Overdraft via POS" was deemed a "positive" for the bank, not a negative. (Previously, "overdraft via POS" was not allowed due to declinations.)
131. Fourth, Wells Fargo was "dependent upon a small set of OD customers"— namely, four percent of them—to generate 40 percent of its overdraft fee revenue. Given this dependence, the bank did not want to lose its "high-OD" customers and, more to the point, did not want to lose the fee income they supplied.
132. In the second email of this exchange, written two weeks after the first email, Mr. Zimmerman reported on his findings regarding the decline in overdraft revenue. Excerpts from this May 2002
133. In short, Mr. Zimmerman discovered that the decline in overdraft fee revenue in early 2002 was attributable to the direct deposit of tax refunds into the checking accounts of high-OD customers. This was "good news" for the bank, because it meant that its high-OD customers would resume "normal OD behavior" once "the excess balances [created by the tax refunds] [we]re depleted." So, Wells Fargo was relieved that its high-to-low bookkeeping could continue without driving away its high-OD depositors—i.e., there was no "cause for concern."
134. Also evident from both emails is that the bank maintained detailed information regarding the overdraft behavior of its customers. In addition to knowing that its high-OD customer segment "generated 40% of total OD/NSF revenue," Wells Fargo also knew that "new accounts generate[d] the bulk of OD revenue" (TX 38). This latter fact was also echoed in the videotaped deposition of Les Biller that was played at trial.
135. The PowerPoint presentation appended to the second email offered additional insights into the bank's views regarding overdrafts and overdraft revenue. The presentation noted that "attrition among ODRI accounts is troubling" and that "increased tax refunds" had produced a "negative ORDI revenue impact" and "negative ODRI effect"—namely a "decline in the frequency of overdrafts" (TX 57) (emphasis added). This shows that Wells Fargo did not want its customers to stop overdrafting (contrary to its litigation
136. These internal documents along with the larger trial record demonstrate that Wells Fargo implemented high-to-low posting, the commingling change, and the extension of the shadow line to debit-card purchases to turn what would ordinarily be one overdraft—which Wells Fargo Senior Vice-President Karen Moore admitted at trial was already "expensive" (Tr. 1149)— into as many as ten, and did so to extract as much ODRI revenue as possible from its ODRI customer base.
137. During the class period, overdraft revenue was the second-largest source of revenue for the Wells Fargo "Consumer Deposits" group (now called the "Consumer and Small Business Deposits" group) (id. at 623, 1515).
138. Although Mr. Zimmerman testified at trial that it was not a "core strategy" of Wells Fargo to grow this massive source of revenue, the bank's own internal memos and emails he authored directly contradicted this assertion. This testimony was not credible.
139. After resistance by the bank, a court order allowed plaintiffs' expert Arthur Olsen to access Wells Fargo's computerized customer data and crunch the numbers. Expert Olsen's study shows that the high-to-low switch boosted overdraft fees by hundreds of millions of dollars for the California class during the class period. The high-to-low switch had a colossal impact on depositors.
140. Weighing all of the evidence presented at trial, this order finds that gouging and profiteering were Wells Fargo's true motivations behind the high-to-low switch and the allied practices that soon followed. High-to-low posting was adopted exclusively to generate more overdraft fees and fee revenue at the expense of depositors. The two closely allied practices that followed were similarly motivated by a singular desire to boost overdraft fee revenue using the bank's high-to-low bookkeeping device. While these changes occurred prior to the class period, they set the stage for the profiteering that ran rampant during the class period and continues even now.
141. Two of the bank's key defenses in this action presuppose that its decision-makers based its practices on factors other than multiplying overdrafts and overdraft revenue. First, Wells Fargo argued at trial that customers wanted and benefitted from a high-to-low posting order. Second, to support its federal preemption defense, Wells Fargo argued at trial that its decision-makers anchored the decision in the four factors set forth in 12 C.F.R. 7.4002(b)(2). This order finds that these are post-hoc rationalizations merely invented for trial.
142. Starting with the argument that Wells Fargo customers "wanted" and "benefitted from" the challenged practices, no credible evidence was presented at trial to support the bank's argument that high-to-low resequencing was deployed for this purpose or that customers wanted or benefitted from it.
143. That the vast majority of debit-card purchases are "must pay" items constitutes
144. Thus, posting debit-card transactions in high-to-low order confers no benefit upon Wells Fargo customers. Unlike checks or ACH transactions, which the bank can return unpaid, there is no risk that Wells Fargo will reject a large debit-card purchase if it posts last rather than first. It is a "must pay" item. Again, this distinction is important because this lawsuit is concerned only with debit-card purchases.
145. Since there is no real benefit to Wells Fargo customers when debit-card transactions are in high-to-low order, the only impact of the bank's posting practices was and remains to multiply the number of overdraft fees assessed on its customers. That is clearly not a "benefit" to any consumer. Indeed, this adverse impact is exactly why Mr. Zimmerman recommended against switching to a high-to-low posting order (but he was overruled).
146. Apart from the must-pay aspect, there is no correlation between the dollar amount of an item and its priority of payment for a depositor. From a depositor's viewpoint, a small item may well be as or more important than a medium or large item. A smaller item to a local government, for example, may be of more importance than a larger item for magazine subscriptions. No study has ever been conducted, as far as the record shows, to establish any correlation between dollar amount and customer priority of payment. On the other hand, it is mathematically certain that posting in high-to-low order will result in more overdrafts than any other posting sequence.
147. It is extremely implausible that depositors would prefer a system guaranteed to turn what would ordinarily be one overdraft into as many as ten to merely improve the odds that a large item would be honored at the expense of others. Yes, it is possible to construct a hypothetical where a large rent check would be paid under a high-to-low scheme but not otherwise. But it is also possible to construct a hypothetical where a smaller but more important check (such as the payment of a traffic ticket) might go unpaid under a high-to-low regime. The supposed net benefit of high-to-low resequencing is utterly speculative. Its bone-crushing multiplication of additional overdraft penalties, however, is categorically assured. Thus, wholly apart from the fact that the rationalization has no applicability to debit-card transactions (due to their "must pay" aspect), the net harm to depositors from the practice is so plain and so vast that this order rejects completely that high-to-low resequencing is a net benefit to depositors or preferred by them. Customer complaints filed by Wells Fargo depositors, quoted elsewhere in this order, underscore this disapproval.
148. Turning to the actual considerations behind the bank's decision, it is true that Wells Fargo witnesses vaguely testified that the high-to-low posting order was adopted to benefit customers so that their largest obligations would be paid first (Tr. 89, 653, 1459, 1473, 1556-58). Almost always, however, this testimony was specific to checks or ACH transactions. According to the argument, a customer's largest expenditures
149. The bank produced no documentary evidence that this supposed "benefit" to customers was actually discussed or considered by Wells Fargo management when the bank made the decision to post debit-card transactions in California in high-to-low order.
150. The bank also produced no study or documentary evidence that any Wells Fargo customers preferred a high-to-low posting order for debit-card transactions (or any transactions for that matter). On this point, two Wells Fargo witnesses— Mr. Zimmerman and Mr. Biller—referred to a consumer survey supposedly performed by Norwest Bank, which Wells Fargo merged with in 1998, that allegedly evidenced a customer preference for high-to-low posting (Tr. 104). This survey proved to be an utter phantom. Despite calls for its production, it was never produced at trial and neither Mr. Zimmerman nor Mr. Biller (nor any other bank witness) knew where a copy of it could be found.
151. Given this lack of evidence and the compelling evidence of gouging in the bank's internal memos, the Court is convinced and so finds that this customer "benefit" and "preference" rationale was invented merely for public consumption and was not an actual motivating factor at the time any high-to-low decision was made, much less the high-to-low decision for must-pay items.
152. It is true that the rationale later appeared in Wells Fargo argument pieces distributed after the fact as scripts to bank employees to help justify the high-to-low posting order to customers who complained (see, e.g., TX 55, 56). These later scripts, however, were for public consumption and did not reflect the bank's true motives.
153. In the totality of the pre-decision memos, the Court has found only one that arguably points to this rationale. In January 2001, Wells Fargo Senior Vice-President Karen Moore sent an email stating that check posting would be changed from serial to high-to-low order (after the decision to make the change had already been made). She stated "[t]his may assure that the customer's most important check (perhaps their rent or mortgage) is paid first" (TX 48). The words "may" and "perhaps" exemplify the speculative nature of this rationalization. "May" necessarily implies "may not." "Perhaps" necessarily implies "perhaps not." This comment, moreover, pertained only to a change regarding checks, not debit-card transactions—the subject of the present challenge. Rent checks and mortgage payments, of course, were not and are not paid using debit cards. There are no emails or memoranda giving any reason for using high-to-low posting for debit-card transactions other than revenue enhancement. This isolated sentence cannot overcome the overwhelming evidence that high-to-low resequencing was applied to debit-card transactions in order to multiply the number of overdrafts.
154. A 2008 FDIC report introduced at trial shows that only about a quarter of surveyed FDIC-supervised banks posted consumer deposit transactions in high-to-low order (TX 184). If it were true that banking customers actually preferred a high-to-low posting order, one would expect that more banks would be doing it and indeed promoting it as a plus. Despite pretending that high-to-low posting
155. A 2009 comment in the Federal Register proffered by Wells Fargo during trial lends the bank only superficial support. In deciding not to address high-to-low posting at that time, the Board of Governors of the Federal Reserve System, the Office of Thrift Supervision, and the National Credit Union Administration stated:
74 Fed. Reg. 5498, 5548 (Jan. 29, 2009).
156. As set forth in the Federal Register, the agencies began the administrative proceeding in May 2008 by soliciting comment on whether institutions should be required to process in low-to-high order to minimize overdraft fees. Industry commentators made several arguments. First, they invoked Section 4-303 of the Uniform Commercial Code as conferring discretion on institutions with respect to posting order. (The legislative comment to Section 4303(b) of the California Commercial Code was not referenced.) Second, some banks argued that most customers prefer high-to-low posting because it results in their largest bills being paid first. Third, some banks argued that there were various alternatives such as in-sequence processing and that they should retain flexibility on a bank-by-bank basis.
Rather than decide anything, the agencies chose not to make any decisions on this particular issue "at this time." They said it would be difficult to set forth "a bright-line rule that would clearly result in the best outcome for all or most customers." They noted that causing smaller items to be paid first "may cause" an institution to return unpaid a large dollar nondiscretionary item, such as a mortgage payment. This statement falls far short of validating high-to-low posting. At most, it merely said that low-to-high posting—the proposal on the table—was not invariably the best sequence, a truism, and that the agencies were not yet ready to decide one way or the other whether to impose it on all banks in all circumstances.
More to the point, this statement cannot overturn the extensive trial record herein. Wells Fargo, for example, has been given a full and fair opportunity to prove that depositors prefer high-to-low posting, yet the bank has utterly failed to do so. It has also been given a full and fair opportunity to find and procure the illusory "Norwest Study" or any other study but it has utterly failed to do so. This is not an administrative proceeding where any blurb and argument piece can be tossed into the bin. We have had a trial with evidence and safeguards for getting at the truth. It is now evident that the bank's supposed studies and rationales have withered and vanished under effective cross-examination. The trial record here demonstrates that depositors do not prefer high-to-low posting, that there is no net benefit, and that the bank's actual motive in imposing a high-to-low regime was to multiply the
157. Turning to the December 2001 commingling change, Wells Fargo witnesses testified that "one of the key motivations" behind commingling was the supposed "migration" of consumer and merchant behavior from one transaction type to another, e.g., away from checks and towards debit-card and ACH transactions (Tr. 1164-65, 1415-16).
158. Due to these supposed trends, the bank allegedly decided to commingle checks, debit-card transactions, and ACH transactions because posting them separately had become "increasingly confusing" to customers and service personnel (id. at 1415-16).
159. Mr. Zimmerman also testified that by commingling debit-card transactions with checks and ACH transactions, the bank could honor more of its customers' high-dollar items in the form of checks and ACH transactions by "truly sorting transactions in a high to low manner" (id. at 1460). This testimony echoed a similar statement, referenced earlier, in the bank's internal BSE memos (TX 36).
160. Weighing the evidence, this order rejects the testimony provided by bank witnesses. First, no written evidence showed that Wells Fargo actually considered "migration trends" when it decided to commingle debit-card transactions with checks and ACH transactions.
161. Second, no written evidence showed that any Wells Fargo customers or service personnel were "confused" by the separate posting of checks, debit-card transactions, and ACH transactions. By contrast, the evidence presented at trial showed that most customers were not even aware of how transactions were posted by the bank. The bank did not notify customers about the old system, the new system, or the change in systems.
162. Third, the notion that customers preferred a high-to-low posting order has already been rejected by this order. Rather, by "truly sorting transactions in a high to low manner" via commingling, Wells Fargo actually exacerbated the harm that customers would suffer in the form of overdraft fees by using up their balances at a faster rate, thereby maximizing the number of items paid into overdraft.
163. This adverse impact is exactly why Mr. Zimmerman—who, as stated, had recommended against the high-to-low switch before being overruled—also recommended against the December 2001 commingling change when the decision was made. He knew that it would amplify the overdraft-multiplying effect of high-to-low posting on customers (Tr. 160, 1551).
164. For these reasons, this order finds that customers did not prefer or benefit from the December 2001 commingling change, and that Wells Fargo did not consider any of the above factors when it decided to commingle debit-card transactions with checks and ACH transactions. Any testimony to the contrary was not believable.
165. As for the shadow-line extension, according to Mr. Zimmerman, one of the "key motivations" behind this change was to increase the approval rate of debit-card purchases. This begs an important question. The increase in approval rate was achieved only by authorizing debit-card purchases into overdrafts. The whole point of the shadow line was, as Mr. Zimmerman put it, to promote "overdraft via
166. No written evidence showed that customers preferred or benefitted from this change. Rather, the documentary evidence presented at trial showed that the bank implemented this change for the main purpose of boosting overdraft revenue by $40 million (TX 36).
167. When unwittingly using the secret shadow line of credit to approve transactions, customers standing at the checkout counter were not warned that an overdraft was in the making. It is very hard to believe that customers would have preferred such a system of "overdraft via POS." This is why the Federal Reserve has recently outlawed the practice absent a customer's affirmative opt-in consent, as will soon be discussed.
168. Turning to the bank's preemption defense and its supposed considerations under 12 C.F.R. 7.4002, the regulation then stated (and continues to state) the following:
169. As set forth in detail below, Wells Fargo did not actually consider any of these factors when implementing the challenged posting practices.
170. While Mr. Zimmerman testified that the April 2001 high-to-low posting order change in California was justified in part by cost considerations, his testimony was not credible (Tr. 1429-30).
171. No documentary evidence supported his assertion that Wells Fargo considered operating costs when deciding to post debit-card transactions from highest-to-lowest dollar amount. Moreover, much of Mr. Zimmerman's testimony on this matter was not anchored in any genuine recollection of what the bank actually considered when it made this decision (see, e.g., id. at 1427-40).
172. It is true that when Wells Fargo deployed high-to-low posting for debit-card transactions in California in April 2001, it had already done so for debit-card transactions outside of California (id. at 1412). This is because Wells Fargo, following its merger with Norwest Bank in 1998, was at the tail end of an eighteen-month process of moving the two banks' accounting and deposit systems onto a single system (id. at 68, 1409-10). California was the last Wells Fargo region to move onto the new system (id. at 280-81, 1411).
173. Referencing this history, Mr. Zimmerman testified that the change in posting order in California in April 2001 "brought California into alignment with the rest of the bank, which was already processing debit cards from high to low," and that this change saved Wells Fargo costs by ensuring bank-wide uniformity (id. at 1429-30).
174. This testimony is irrelevant, however, to whether the initial bank-wide decision by Wells Fargo to deploy high-to-low posting for debit-card transactions was motivated by cost considerations. Indeed, as explained at trial, Wells Fargo's decision
175. One impetus for the decision in 1999 was Wells Fargo's merger with Norwest Bank. At the time of the merger in 1998, Wells Fargo and Norwest were posting debit-card transactions in different sequences—Wells Fargo was posting in low-to-high order while Norwest was posting in high-to-low order. Thus, Wells Fargo had to decide, among other things, on a uniform posting order for the new merged entity. In this connection, Wells Fargo considered three different posting orders: (1) high-to-low, (2) low-to-high, and (3) sequential/chronological order. After considering these three options, Wells Fargo executives—overruling the objections of Mr. Zimmerman noted earlier—made the bank-wide decision to post debit-card transactions in high-to-low order (id. at 75-76).
176. The bank produced no documentary evidence, however, that Wells Fargo management considered the impact that these three sequencing orders would have had on the bank's operating costs. Rather, the trial record shows that the bank only considered increased revenue.
177. In sum, this order finds that Mr. Zimmerman's testimony that Wells Fargo considered operating costs when deciding to post debit-card transaction in high-to-low order was not credible. The bank did not consider operating costs when it made this decision.
178. Wells Fargo similarly produced no evidence at trial showing that the bank instituted high-to-low posting of debit-card transactions to deter customers from misusing banking services. By contrast, the trial evidence showed that Wells Fargo deployed its high-to-low posting practices—including the commingling change and the extension of the shadow line to debit-card purchases-to significantly increase the number of overdrafts incurred by customers and further boost an already major source of fee revenue (TX 36, 38, 57, 61; Tr. 134-35, 237). In short, the bank's intentions behind high-to-low resequencing were to promote—not deter—customer overdrafts.
179. As showcased in the April/May 2002 emails between Mr. Zimmerman and Mr. Biller, Wells Fargo became "concerned" when overdraft frequency unexpectedly declined (TX 38). Part of this concern was that the deployment of high-to-low posting had caused increased attrition among the bank's coveted "high-OD customer segment"—the "4% [that] generate[d] 40% of total OD/NSF revenue" (TX 38). Attrition among these accounts was "troubling" to the bank (TX 57). Once the bank learned that tax refunds were to blame, however, the bank treated this as "good news"—it simply had to wait until the tax refunds were "depleted" before "normal OD behavior" would again resume (TX 38).
180. BSE documents similarly showed that Wells Fargo had no intention of deterring overdraft behavior—rather, the bank intended to promote overdrafts. The BSE initiatives were dependent upon an increase in customer overdrafts (see, e.g., Tr. 134-35, 237; TX 36, 61). On this point, Wells Fargo's own documents stated that "two most important OD/NSF drivers" for increasing OD/NSF revenue were: (1) the percentage of accounts incurring overdrafts and (2) the number of overdraft items per overdrafted account (TX 36).
181. This order rejects as untrue the testimony provided by bank witnesses that high-to-low posting was implemented to deter overdraft behavior and prevent the bank from attracting customers who would misuse bank services via overdrafting.
182. With respect to considering the "competitive position" of the bank, as already
183. When finally asked directly, Mr. Zimmerman confirmed that he had no specific recollection of whether the posting practices of competitors were specifically considered when Wells Fargo decided to post debit-card transactions in high-to-low order (id. at 1448). Mr. Zimmerman further testified that he did not recall any specific discussion about the posting practices of Wells Fargo's competitors when the decision to commingle transactions in April 2001 was made (id. at 1449).
184. No documentary evidence showed that Wells Fargo considered its competitive position or even knew how its competitors were posting debit-card transactions when it made the decision to post these transactions from high-to-low.
185. This order finds that Wells Fargo did not consider the enhancement of its competitive position when it decided to post debit-card transactions in high-to-low order.
186. Finally, no documentary evidence supported Wells Fargo's argument that it changed its posting practices to maintain the safety and soundness of the bank. Wells Fargo changed its posting order to generate millions of dollars in additional revenue at the expense of its customers. True, this would enhance the balance sheet and thus arguably enhance the bank's financial strength. But are safety and soundness served by the promotion of overdrafts? The Court doubts that the OCC would agree with this view. Certainly, no OCC authorities have been supplied so stating.
187. The consistent lack of internal records evidencing the bank's supposed consideration of these factors is even more suspect given the nine- to twelve-month planning process required to deploy the challenged practices (id. at 1421, 1558). Given this lengthy time frame, there should have been some documents or emails showing that the bank considered these factors, especially if the bank was subject to federal oversight under 12 C.F.R. 7.4002 and other federal regulations. No such evidence was presented. While Wells Fargo asserted that it did not have document retention policies at the time for these types of documents, its retention policies were never offered at trial despite an express invitation to present them at trial.
188. In sum, this order finds that the bank did not actually consider any of the factors set forth in 12 C.F.R. 7.4002 when it decided to switch to high-to-low posting for debit-card transactions, commingle transactions, and extend a "shadow line" of credit to debit-card purchases.
189. If high-to-low resequencing, the commingling change, and the shadow-line extension were subject to 12 C.F.R. 7.4002, the bank would have kept records to prove compliance with the regulation and for national bank examiners to review (id. at 1529).
190. The fact that no documents show that Wells Fargo was attempting to comply with 12 C.F.R. 7.4002 shows that the federal regulation governing bank charges and fees was a non-factor in these decisions.
191. The only motives behind the challenged practices were gouging and profiteering.
192. The bank's argument that its one-dollar "courtesy threshold" implemented during the tail end of the class period evidences its "good faith" towards its customers is rejected (id. at 355-56). Similarly, the bank's argument that after-the-fact fee reversals or posting credits before debits evidences the bank's "good faith" is also rejected. These were non-factors when the bank decided to deploy the practices challenged herein.
193. In December 2002, a class action lawsuit—Sean M. Smith v. Wells Fargo Bank, N.A. (Case No. GIC802664)—was brought against Wells Fargo in San Diego Superior Court over an alleged failure by the bank to adequately disclose the extension of the shadow line to debit-card purchases in a 2002 "Policy Change Notice." Pillsbury Winthrop Shaw Pittman LLP represented the bank, while Finkelstein & Krinsk LLP served as class counsel.
194. A settlement agreement was reached in May 2007, and the class action settlement was approved in November 2007 by Judge Ronald S. Prager. Judgment was thereafter entered. Under the settlement agreement, class members were allowed to submit a claim form to Wells Fargo (postmarked by January 8, 2008) stating that they had incurred an overdraft fee during the twelve-month period following May 24, 2002. Wells Fargo would then refund up to $20 in overdraft fees to that class member, subject to an aggregate cap on total claims of three million dollars.
195. What Judge Prager did not know—since it only came to light in this proceeding—was that Wells Fargo paid merely $2080 to 166 claimants. This is not a typo. Barely two thousand dollars were paid to class members.
196. By contrast, Wells Fargo agreed to pay Finkelstein and Krinsk $2.2 million in attorney's fees and costs.
197. On February 1, 2008, Judge Prager ruled on a motion concerning whether and how Smith encompassed the instant case. Judge Prager held that while our named plaintiff was a member of the settlement class in Smith, the claims in our case were not released by the Smith settlement. While Judge Prager noted that the release language in the Smith settlement agreement appeared quite broad, he found that the release was actually limited to the claims in the Smith complaint. (Judge Prager still, it should be noted, was not made aware of the minuscule amount paid to dispose of Smith.)
198. After examining the Smith complaint, Judge Prager concluded that the claims in the instant action—targeting Wells Fargo's high-to-low posting practices—were not released because the claim in Smith was limited to disclosures in a 2002 "Policy Change Notice" and overdraft fees incurred during the twelve-month period between May 2002 and May 2003. In reaching this conclusion, Judge Prager noted that he did not agree with Wells Fargo's argument that the Smith release barred the claims asserted herein.
199. Given the harsh impact of the bank's high-to-low practices, the bank was
200. The Wells Fargo Consumer Account Agreement ("CAA") was and remains a lengthy document given to customers when opening a new account (Tr. 275). The trial record contains numerous versions of Wells Fargo's CAA, including version from November 1999 (TX 49), August 2000 (TX 75), April 2001 (TX 76); April 2002 (TX 13), October 2004 (TX 79), October 2005 (TX 81), October 2006 (TX 19), and November 2007 (TX 339). All of these CAAs were distributed nationwide, and each is approximately 60 pages in length (Tr. 1134-35).
201. While plaintiff Walker did not recall receiving a copy of the CAA when she opened her Wells Fargo bank accounts in July 2006, this order nevertheless finds that the CAA was customarily provided to all Wells Fargo customers when they opened a new account (Tr. 796). The CAA claims to "govern[] the relationship between the bank deposit account holder[] and the bank" and "outlines the terms and conditions that apply to [Wells Fargo] checking accounts, and outlines the contractual obligations both of the bank, as well as with [Wells Fargo] consumers" (Tr. 586, 1134; see also TX 13 at 4).
202. None of the CAAs introduced at trial adequately disclosed Wells Fargo's posting practices. For example, the April 2002 CAA contained a section entitled "Debiting Your Account; Order of Posting" (TX 13 at 24-25). This section stated, in relevant part:
203. While the April 2002 CAA used the phrase "[w]e may choose to pay Items in the order of highest dollar amount to lowest dollar amount[,]" it is undisputed that Wells Fargo was then actually posting cash withdrawals, debit-card, check, and ACH transactions from highest-to-lowest dollar amount (see Dkt. No. 448; Tr. 2770-78).
204. The phrasing "[w]e may choose" suggested to customers that the bank would either exercise discretion or that it had not yet chosen to go to a high-to-low scheme. In fact, the bank knew good and well that it was already imposing and would continue to impose high-to-low bookkeeping—the worst possible system from the customer's perspective.
205. This phrasing was almost identical, in fact, to the phrasing used before the switch. No notice was ever given of the high-to-low change.
206. In the October 2004 CAA, Wells Fargo altered the prior language and added new language to the "Order of Posting" section of the CAA. The new language is highlighted below in italics (TX 79 at 23):
207. A similar explanation was included in Wells Fargo CAAs dated October 2005, October 2006, and November 2007 (TX 81, 19, 339). (The parties did not put a CAA from 2003 into evidence.)
208. This revised language compounded the deception. It did not adequately disclose that the bank had already adopted the high-to-low scheme. Rather, a reasonable consumer would have been deceived— just as with earlier CAAs—by the bank's representation that it "may, if it chooses, post Items in the order of the highest dollar amount to the lowest dollar amount." This disclosure reinforced the misleading impression that Wells Fargo had not yet chosen to post in the order of the highest dollar amount to the lowest dollar amount or that it would exercise discretion on a case-by-case basis.
209. Apart from its content, the "disclosure" on posting order was buried within a sea of single-spaced text stretching over 60 pages in tiny ten-point font. For example, in the 2002 CAA, the posting-order "disclosure" was 24 pages into the document. This order finds that no reasonable depositor could be expected to read the entire document or locate the "disclosure" within the bank's CAA. Moreover, even if the customer read the CAA, no reasonable depositor could be expected to understand the disclosure regarding posting order and overdraft fees, especially given the deceptive use of "may" throughout the disclosure.
210. Tellingly, Wells Fargo's own expert, Itamar Simonson, who spoke to the adequacy of the bank's disclosures regarding posting order, testified that the length and complexity of the CAA made it "completely unrealistic to assume that . . . many consumers would actually read those lengthy documents" (Tr. 977-78). Additionally, Mr. Simonson confirmed that the length and complexity of the CAA made it "so difficult for consumers to understand" (id. at 982). This remarkably candid opinion was echoed throughout Expert Simonson's direct testimony (id. at 977, 979):
211. This order also agrees with Expert Simonson on these points as well as his opinion that those Wells Fargo customers who read the CAA "are probably not the same people who just have $20 in their
212. The bank, however, knew—when it wanted—how to speak plainly. After the fact, when a customer complained about getting hammered with overdraft fees, Wells Fargo provided a clear explanation of its posting process. This explanation, however, would only be provided once the customer submitted a written complaint expressing anger or confusion regarding overdraft fees. Wells Fargo would send these customers a response letter containing standard "form" language describing in very straightforward terms the actual posting order used by the bank (Tr. 1319-20; TX 127).
213. An example of this "form" language, found in a Wells Fargo letter dated February 26, 2008, in response to a customer complaint, is reproduced below (TX 127, BATES 4387) (emphases in original):
214. In addition to these form-letter responses, Wells Fargo provided a similarly clear disclosure regarding posting order to customers if they went to the trouble of complaining about excessive overdraft fees over the phone (Tr. 1313-14).
215. The very existence of these clear after-the-fact explanations further highlights the bank's before-the-fact obfuscation. Before the fact, the bank piled ambiguity onto vagueness and hit it all at page 27 of a 60-page document. Once the customer was hammered by the bank's resequencing practice and went to the trouble of complaining, then and only then did the bank roll out the clear-cut script.
216. In sum, the bank intentionally made its disclosure in a way that was calculated to go unnoticed by class members and used language that, even if read, obfuscated the practices challenged herein.
217. Additionally, given that Wells Fargo's own expert admitted that most bank customers did not read the CAA and, even if they did, the length and complexity of the document made it "so difficult for consumers to understand," this order also finds that it was not within the reasonable expectations of class members that the bank had in fact already put in place a posting practice that could transform a single mistake into as many as ten overdrafts.
218. No other Wells Fargo documents adequately disclosed the challenged resequencing practices to the bank's customers in California. With respect to Wells Fargo's Consumer Account Fee and Information Schedule, which—like the CAA—was provided to Wells Fargo customers when they opened a new account, this document also failed to disclose how transactions were resequenced prior to batch posting or
219. Nor did the bank's account statements alert depositors. Throughout the class period, Wells Fargo separated credit transactions, debit-card transactions, and checks on its printed account statements. The separation of these transactions made it nearly impossible for a customer to determine the actual posting order being employed by the bank.
220. Further obfuscating the bank's high-to-low posting order was the fact that certain "priority debit" transactions—like cash withdrawals and balance transfers— were always posted before debit-card transactions, checks, and ACH transactions. The account statement of Ms. Gutierrez discussed earlier in this order provided a clear example of this obfuscation, as her $22 ATM withdrawal posted before her $74.39 debit-card purchase at Albertsons supermarket (TX 1). Thus, even if a reasonable consumer viewed the actual order that his transactions were posted by the bank, a high-to-low ordering would not be self-evident.
221. Finally, none of Wells Fargo's marketing documents admitted into evidence disclosed the bank's posting order or its potential impact on overdraft fees.
222. The bank's murky disclosures regarding posting order were exacerbated by misleading information disseminated by Wells Fargo reinforcing the perception that transactions would be deducted from their accounts in chronological order.
223. A Wells Fargo marketing theme was that debit-card purchases were "immediately" or "automatically" deducted from an account. This likely led the class to believe: (1) that the funds would be deducted from their checking accounts in the order transacted, and (2) that the purchase would not be approved if they lacked sufficient available funds to cover the transaction. This language was present on Wells Fargo's website (TX 129), on Wells Fargo's Checking, Savings and More brochures from 2001 and 2005 (TX 88, 89), and Wells Fargo's New Account Welcome Jacket from 2004 (TX 82).
224. A few examples of these misleading statements were included in the findings for Ms. Gutierrez and Ms. Walker (Nos. 17, 63). Additional examples of these misleading representations are shown below:
225. This order finds that these misrepresentations were placed in such a wide array of marketing documents and these documents were distributed in such a widespread manner that class members were likely to be misled by them.
226. Furthermore, when a customer used Wells Fargo's online banking service, the bank would display "pending" debit-card purchases in chronological order, leading customers to believe that the processing would take place in that order. In reality, debit-card purchases were never posted in this order during the class period, as the bank well knew. Nothing on the bank's online-banking website warned customers that the pending transactions, displayed chronologically to them, would be resequenced behind the scenes and posted in high-to-low order to maximize overdrafts.
227. Finally, the trial record clearly shows that Wells Fargo encouraged its customers to keep track of their account balances using a register (see, e.g., TX 131 at 48; Tr. 547, 1216). Indeed, this has been a theme of the bank in this litigation, and Wells Fargo account statements included this very recommendation (see, e.g., TX 1). Pam Erwin, who handled Wells Fargo's financial literacy program to educate consumers about financial decision-making, explained at trial when discussing debit-card purchases that "it's the account holder's responsibility to track those transactions against their check register and/or savings register" (Tr. 547). A register, however, is chronological. It fosters the view that items are deducted in chronological order.
228. In this regard, a customer who faithfully uses a register to track items would know the exact point—indeed, the exact transaction—at which her account goes into overdraft. That faithful customer could not reasonably be expected to know that the bank would manipulate the order of her transactions so as to deplete her account balance faster than shown in her register, triggering not one but as many as ten overdraft penalties, all due to the bank's high-to-low bookkeeping scheme. A precise register could never alert a customer who makes a mistake that her one overdraft will be converted into as many as ten overdrafts.
229. This order finds that Wells Fargo not only failed to adequately disclose its high-to-low posting practices but promoted a false perception that debit-card purchases would be deducted from their accounts in the order transacted.
230. The shadow line itself has never been disclosed to customers—its specific
231. The objective of the shadow line was to promote "overdraft via POS," as the bank's internal memos explained, meaning to encourage more overdrafts— and thus more overdraft fees—through point-of-sale debit-card purchases. Instead of declining transactions when funds were too short, as it once did, the bank switched to approving them without warning to the customer standing in line at a coffee shop or elsewhere that an overdraft—and a $34 overdraft fee—was in progress.
232. The available-balance information communicated online to customers was supposed to represent the amount of funds the bank would make available for their next transaction. Wells Fargo defined it as such. This, however, was not true. Wells Fargo allowed customers to spend more than their available balance using their debit cards via undisclosed shadow-line overdrafts.
233. In 2002, prior to adopting the shadow line for debit-card purchases, Wells Fargo put the following notice in its customers' monthly account statements and—for new customers—by an addendum to their customer account agreements (TX 364) (emphasis in original):
234. This was inadequate to fairly warn depositors.
235. Second, the insert stated: "If we authorize the transaction, you may be charged a fee, which will vary depending upon the action taken." In truth, however, it might not be "a fee" but as many as ten overdraft fees (at $35 each). The same applied to the phrase "We may . . . pay the item and create an overdraft to your account." In truth, it might not be "an overdraft" but as many as ten overdrafts. The word "may" was misleading, for it was in truth a certainty—not a maybe—that overdraft fees would be assessed barring
236. This order finds that these disclosures failed to warn or explain to reasonable depositors that the shadow-line extension would amplify the impact of high-to-low posting on overdraft fees (Tr. 1154; TX 90, 341, 408). This omission was likely to deceive class members.
237. Hundreds of customer complaints regarding overdraft fees were admitted at trial. These included many complaints written by California Wells Fargo customers who found themselves on the receiving end of a cascade of overdraft charges caused by high-to-low resequencing. These complaints expressed indignation that transactions were "re-dated," "re-arranged," and "manipulated" from the order in which they had occurred (see, e.g., TX 127, BATES 4583, 4663, 4931, 5241, 5269, 5957).
238. One handwritten complaint from a customer who had been "a loyal Wells Fargo customer for almost 8 years, starting at age 18," stated (id. at BATES 5148):
239. This order agrees that Wells Fargo constructed a trap—a trap that would escalate a single overdraft into as many as ten through the gimmick of processing in descending order. It then exploited that trap with a vengeance, racking up hundreds of millions off the backs of the working poor, students, and others without the luxury of ample account balances.
240. This order further finds that enforcement of plaintiffs' claims would have only an incidental effect on the bank's ability to engage in the business of banking. It would also not "obstruct, impair, or condition" the bank's ability to exercise its deposit-taking powers. For many years, the bank used low-to-high processing and other posting methods. Doing so did not impair the bank's abilities.
241. For years, moreover, Wells Fargo has complied with far more restrictive laws in Nevada barring high-to-low posting of checks (TX 59):
N.R.S. 657.120 (1999).
242. Wells Fargo has complied with this restrictive law by simply posting transactions differently in Nevada than in California (since 2001) (TX 59). This has not disturbed the bank's ability to engage in the business of banking and exercise its deposit-taking powers.
243. Additionally, Wells Fargo has also been posting transactions differently in New Mexico and Washington "for at least a couple of years" (Tr. 1502-08). In these states, the bank has been operating a "pilot program" where all "prior day" credits and debits are posted before credits and debits presented on the "current day." This provides further proof that posting order variances do not more than incidentally affect the bank's ability to engage in the business of banking.
The legal analysis will proceed in two steps: First, this order will address plaintiffs' claim that Wells Fargo's high-to-low posting practice was "unfair" under Section 17200. Second, this order will address whether that practice was "fraudulent" under Section 17200. With respect to Wells Fargo's defenses on these claims, most will be addressed with the merits of each claim. This order, however, will separately address Wells Fargo's preemption defense and "no standing" defense. Finally, this order will consider the impact of the release in Smith v. Wells Fargo Bank, N.A.
Section 17200 of the California Business and Professions Code prohibits business acts or practices that are "unlawful," "unfair," or "fraudulent." Each of these three restrictions constitutes a separate and independent claim. Cel-Tech Commc'ns, Inc. v. Los Angeles Cellular Tel. Co., 20 Cal.4th 163, 180, 83 Cal.Rptr.2d 548, 973 P.2d 527 (1999) (citations omitted). Only "unfair" and "fraudulent" business practices are at issue in this litigation.
For an "unfair" business practice under Section 17200, the undersigned judge previously held herein that plaintiffs had to (1) identify an unfair policy or practice tethered to a legislatively declared policy or (2) show an actual or threatened impact on competition (Dkt. No. 246). Id. at 185-88, 83 Cal.Rptr.2d 548, 973 P.2d 527. The claim that Wells Fargo's "resequencing" practices are "unfair" under Section 17200 is properly "tethered" under Cel-Tech to the legislative comment expressed under California Commercial Code Section 4303(b), as now detailed.
Adopted in 1963, California Commercial Code Section 4303(b) is California's version of Section 4-303 of the Uniform Commercial Code. It addresses the relationship between the bank and presenters of items for payment and states:
The California legislature, however, expressly disavowed such unfettered discretion. As part of the 1992 amendments to the California Commercial Code, a legislative comment was added to Section 4303 that stated:
Cal. Com.Code § 4303(b), Calif. cmt. 7 (emphasis added).
The same legislative comment also made clear that "discretion" as used in Section 4303(b) meant item-by-item discretion:
Cal. Com.Code § 4303(b), Calif. cmt. 7 (emphasis added). In other words, the legislature recognized that a bank may exercise individualized discretion to prioritize
While on its face this comment only refers to checks, it plainly covers debit cards as well. The comment was approved when depositors only used checks in lieu of a cash payment. Since then, debit cards have arrived as another way to make such payments. The comment applies with full force to both. Indeed, Wells Fargo itself argued that Section 4303(b) applied to both checks and debit cards when it raised this section as a shield to liability—albeit without expecting anyone to read the legislative comment. Wells Fargo does not argue to the contrary even now, but this point should be made lest the question suggest itself on appeal. (That the bank once relied so vigorously on the California Commercial Code, moreover, belies its supposed faith in federal preemption.)
Apart from this provision of the UCC, California's civil law governs the relationship between a bank and a customer. In Perdue v. Crocker Nat'l Bank, 38 Cal.3d 913, 923, 216 Cal.Rptr. 345, 702 P.2d 503 (1985), a decision specifically addressing overdraft fees, the California Supreme Court held that "where a contract confers on one party a discretionary power affecting the rights of the other, a duty is imposed to exercise the discretion in good faith and in accordance with fair dealing."
Throughout trial, Wells Fargo stated that the CAA was its "contract with the customer" that "govern[ed] the relationship" between the bank and deposit-account holders. The CAA itself contained such language (see, e.g., TX 3 at 4). Unlike the simple one-page agreements that were used at the time of Perdue, the Wells Fargo CAA weighs in at over 60 pages, each page containing single-spaced text in ten-point font. Buried deep within this document in substantially the same form in all versions was the statement that Wells Fargo "may, if it chooses, post items in the order of the highest dollar amount to the lowest dollar amount" (see, e.g., TX 13 at 24-25, TX 79 at 23). Passing over momentarily the fact that even the bank's own expert conceded that few consumers would ever read or understand it, this phrasing called for the exercise of discretion, as the bank's own internal memos recognized (TX 48). And, having delegated itself discretion, Perdue imposed on the bank the duty to exercise that discretion in accordance with the covenant of good faith and fair dealing. This precludes exercising that discretion to turn what would ordinarily be one overdraft into as many as ten.
Citing Carma Developers v. Marathon Dev. Cal., Inc., 2 Cal.4th 342, 373-74, 6 Cal.Rptr.2d 467, 826 P.2d 710 (1992), Wells Fargo argues that this language allowed the bank to exercise its discretion to adopt high-to-low posting—even if in bad faith. In Marathon, the California Supreme Court held that while the implied covenant of good faith and fair dealing was read into every contract, it could not create obligations different from or inconsistent with the express terms of the contract. That said, Marathon emphasized that express grants of discretion were still subject to "the reasonable expectations of the parties." Id. at 374, 6 Cal.Rptr.2d 467, 826 P.2d 710.
At least with the freely negotiated lease-termination provision in Marathon, both sides reasonably expected and understood the impact of the discretion. Here, even Wells Fargo's own expert, Itamar Simonson, admitted that it was "difficult for consumers to understand," based upon the CAA disclosure, how the exercise of discretion by a bank to post transactions from highest-to-lowest dollar amount would impact them (id. at 982).
An important factor in determining the reasonable expectation of the parties in the context of adhesion contracts is "the extent to which the contract affects the public interest." Allan v. Snow Summit, Inc., 51 Cal.App.4th 1358, 1375-76, 59 Cal.Rptr.2d 813 (1996). The legislative comment to Section 4303(b) of the California Commercial Code evidences a policy of California to ensure that banks act with good faith when re-ordering transactions during the posting process and not post in a way to ramp up overdraft fees. Here, the contractual language regarding posting order was ambiguous at best. The bank stated in every version of its CAA that it "may, if it chooses, post items in the order of the highest dollar amount to the lowest dollar amount" (see, e.g., TX 79 at 23) (emphasis added). Even a sophisticated customer could reasonably have understood this to mean that the bank had not yet chosen to implement such a change or would exercise discretion on an item-by-item basis as expressly contemplated by the legislative comment to Section 4303. Ambiguities should be resolved against the drafter, especially where contracts of adhesion are involved. See Cal. Civ.Code § 1654; see also Goddard v. South Bay Union High School Dist., 79 Cal.App.3d 98, 105, 144 Cal.Rptr. 701 (1978).
Finally, the same section of the CAA also noted that "[t]he Bank may post Items presented against the Account in any order the Bank chooses, unless the laws governing your Account either requires or prohibits [sic] a particular order" (see, e.g., TX 79 at 23) (emphasis added). The "good faith" limitation found in the legislative comment of Section 4303(b) imposed such a restriction. That section (and the comments thereto) prohibited certain posting orders—namely, those implemented for the purpose of increasing the number of overdraft penalties inflicted on customers. Thus, by the bank's own terms, it subjected itself to the "good faith" limitations of Section 4303(b).
In sum, customers reasonably expect, of course, that if they overdraft their accounts, they will have to pay an overdraft fee. They do not, however, reasonably expect that they will have to pay up to ten overdraft fees when only one would ordinarily be incurred. This severe result is so pernicious that it should be allowed, if at all, only upon a showing that it was
Based on the legal standards above, this order concludes that the revenue initiatives implemented by Wells Fargo in 2001 and 2002—all of which ran rampant during the class period and indeed through at least the end of trial— violated the "unfair" restriction of Section 17200. The trial record and findings herein show that these initiatives had an immediate, intended, and material impact on increasing the number of overdraft fees that Wells Fargo could assess customers. All three initiatives were motivated by avarice at the depositor's expense. No credible evidence—indeed, no written evidence—points to any other rationale.
To recap these facts, the first initiative was the high-to-low switch. Before, Wells Fargo used low-to-high posting, which minimized overdrafts. The switch to high-to-low posting maximized overdrafts. This was mathematically guaranteed. No notice was given that the switch had occurred.
None of the bank's proffered justifications are credible. First, as found above, the bank did not act out of solicitude for customers and any supposed belief that customers would prefer high-to-low posting. That is a post-hoc rationalization. Among other reasons, the vast majority of debit-card purchases are "must pay" transactions. As such, posting order would make no difference to the customer—the bank is required to honor a debit-card purchase whether it posts first, last, or somewhere in between. Second, as found above, the bank did not base its decision on the factors set forth in 12 C.F.R. 7.4002. Wells Fargo considered nothing other than increasing overdraft revenue when deciding to post debit-card transaction in high-to-low order.
The initiatives challenged herein required nine to twelve months of planning and implementation. Many emails and reports would have been generated over this extended time period. The trial record, however, is devoid of written documents showing that Wells Fargo implemented the high-to-low change for any purpose other than profiteering. While Wells Fargo asserted that it did not have document retention policies at the time for these types of documents, its retention policies were never offered at trial despite an express invitation to present them. And, if Wells Fargo had truly believed that its initiatives were subject to federal oversight under 12 C.F.R. 7.4002 and other federal regulations, it would normally have prepared and retained written records to show compliance. The absence of any such records speaks volumes.
In sum, Wells Fargo's decision to post debit-card transactions in high-to-low order was made for the sole purpose of maximizing the number of overdrafts assessed on its customers, exactly what Section 4303 and Perdue bar. Accordingly, the decision was not made in good faith.
The second initiative, implemented in December 2001, was the commingling of debit-card transactions with checks and
Before, all debit-card transactions were posted ahead of checks, and all checks were posted ahead of ACH transactions. There was a purpose behind this separation. Unlike debit-card transactions, checks and ACH transactions were not "must pay" transactions. In other words, the bank had discretion to return them unpaid. Because of this, Wells Fargo had a legitimate business reason to pay "must pay" items first (e.g., process checks and ACH transactions after all debit-card transactions). On this point, the legislative comment to Section 4303(b) states that "[t]he bank has the right to pay items for which it is itself liable ahead of those for which it is not."
This, however, was thrown aside in December 2001. The commingling change had the intended effect of multiplying the number of debit-card transactions for which the bank could assess an overdraft fee. Through commingling, the typically larger checks and ACH items used up the accounts faster so that there would be more overdrafts when the typically smaller value debit-card items were paid. Wells Fargo's own words summed this effect up best: the commingling change was designed to "more-closely mirror true High-to-Low sort order" (TX 36), i.e., to eat the account up as soon as possible and thus yield more overdraft fees.
The internal Wells Fargo documents quoted in the findings show that overdraft revenue was the sole motivating factor, predicting a "lift" in overdraft revenues by $40 million. The April/May 2002 email exchange between Mr. Zimmerman and Mr. Biller graphically confirmed this profiteering motive.
Given the evidence, this order concludes that the decision to commingle debit-card transactions with checks and ACH transactions was made solely to gouge the customer, exactly what the legislative comment barred.
The third initiative was "overdraft via POS"—the extension of the shadow line to debit-card purchases in May 2002. As a result of this change, Wells Fargo began authorizing debit-card purchases even though the account was already overspent. Before, if an account holder had insufficient available funds to cover a debit-card purchase, the bank would decline the transaction, thereby protecting the customer from further unintended overdrafts. After, the bank authorized the transaction without informing anyone that an overdraft was in progress.
Profiteering was the sole motive behind this revenue initiative. On this point, BSE documents clearly showed that the extension of the shadow line to debit-card purchases in May 2002 was—like the commingling change—expected to increase fee revenues by $40 million. (In other words, this $40 million increase was above and beyond the separate $40 million in revenue expected from commingling.) Mr. Zimmerman called this initiative "overdraft via POS"—a telling label capturing the true motive. No credible or written evidence showed that this initiative was implemented for any other reasons.
In sum, Wells Fargo used bad faith to deploy high-to-low posting of debit-card transactions, commingle debit-card transactions with checks and ACH transactions, and extend the shadow line to debit-card purchases. The commingling and shadow-line decisions were closely allied with the bank's high-to-low bookkeeping scheme. Together, they formed a "one-two-three" punch to maximize the overdraft-multiplying effect of a high-to-low posting order— all at the expense of customers.
Wells Fargo's counter-arguments are unpersuasive. First, Wells Fargo argues that it did not seek to "maximize" the number of overdraft items for the "sole purpose" of assessing overdraft fees. To support this argument, the bank points to its overdraft-protection services, its limitation of ten overdraft items per day, and its one-dollar overdraft "courtesy threshold." In the same vein, Wells Fargo also argues that by posting a customer's credits first (rather than last), this reduced the number of overdraft items it could have potentially assessed customers.
These arguments are rejected. That Wells Fargo could have gouged even worse than it has hardly alters the fact that it has gouged badly. Plaintiffs do not need to prove that the bank mistreated depositors in every way possible in order to show that they were mistreated. The presence of an extraordinarily high ten-item cap on daily overdraft fees, as well as a minuscule one-dollar courtesy threshold, do not change the fact that Wells Fargo deployed the challenged practices for the sole purpose of multiplying overdrafts to increase fee revenue. Additionally, the mere continued availability of overdraft-protection services—another profit center for the bank that benefits from increased overdraft activity—does not change the fact that high-to-low posting had the intended effect of squeezing more overdrafts out of its customers.
Second, this order rejects Wells Fargo's argument that because class members could have avoided overdraft fees if they simply had managed their accounts with precision, its practices cannot be deemed "unfair." Even the most precise register could not have predicted that what appeared to be a single overdraft would be converted to ten by the artifice of high-to-low resequencing. Of course it is true that we should all live within our means and avoid overdrafts. That is a given. And, of course it is true that when we overdraft, whether by accident or not, we must expect to pay a fee. That is also a given. This, however, cannot justify turning what would ordinarily be one overdraft into as many as ten. Such a severe result should only be tolerated after clear-cut warnings and consent.
Third, contrary to Wells Fargo's characterization of this litigation, plaintiffs are not challenging a simple "price" increase. Nor are plaintiffs arguing that "making money" is unlawful. Rather, plaintiffs' claims are properly tethered to a "good faith" limitation set forth by the California legislature in a comment to California Commercial Code § 4303(b) limiting a bank's ability to increase its revenues through resequencing schemes.
To establish liability under the "fraudulent" restriction of California Business and Professions Code Section 17200, "it is necessary only to show that members of the public are likely to be deceived." In re Tobacco II Cases, 46 Cal.4th 298, 312, 93 Cal.Rptr.3d 559, 207 P.3d 20 (2009) (internal quotations and citations omitted). Unlike common law fraud, which requires proof of falsity, scienter, and actual reliance, it is not necessary for a plaintiff to prove that a fraudulent deception under
A claim that a business practice is (or was) "fraudulent" under Section 17200 can be based upon representations that deceive because they are untrue as well as representations that may be accurate on some level but nonetheless tend to mislead or deceive. As such, a perfectly true statement couched in such a manner that it is likely to mislead or deceive the consumer, such as by failure to disclose other relevant information, is actionable under Section 17200. Id. at 1255, 99 Cal.Rptr.3d 768 (citation omitted). As now detailed, Wells Fargo's deceptive practices were and continue to be in violation of the "fraudulent" restriction of Section 17200.
Fees associated with individual overdraft items were disclosed to Wells Fargo customers through a document called the Consumer Account Fee and Information Schedule. This document, which was provided to customers when they opened an account, not only explained the per-item fee that would be assessed per overdraft item, but set forth the ten-item daily cap on overdraft and NSF fees. Wells Fargo used this document to disclose all relevant information for account-related fees, and reasonable consumers would have understood this document to contain all relevant information pertaining to overdraft fees.
Significantly, the Consumer Account Fee and Information Schedule did not disclose the high-to-low posting order or its impact. It did not tell customers that frequent use of a debit card for small-valued purchases could result in an avalanche of overdraft fees for each of those purchases due to the high-to-low posting order. It also failed to cross-reference any information on posting order in the CAA.
In shaping the reasonable expectations of its customers, the bank should have prominently disclosed its high-to-low scheme and its ability to transform a single mistake into as many as ten overdrafts. The bank's "disclosures" on posting order, however, were buried 20-or-so pages into a 60-plus-page document of single-spaced, ten-point font text. As stated, Wells Fargo's own trial expert, Itamar Simonson, affirmatively testified on direct examination that customers would not and could not be expected to read the lengthy document (Tr. 977-78). And, even if it was read, the statements therein on posting order were "difficult for consumers to understand," as the bank's own expert confirmed (id. at 982).
For example, the October 2004 CAA stated that "the Bank may, if it chooses, post [transactions] in the order of the highest dollar amount to the lowest dollar amount" (see, e.g., TX 79 at 23) (emphasis added). This statement—if it was ever read—would have been understood by an ordinary consumer to mean that Wells
The closest Wells Fargo got to an accurate and understandable disclosure of its posting practices was after the fact and, even then, only to those who complained (see, e.g., TX 127, BATES 4387). Wells Fargo provided its tellers and phone-bank employees with a clear script to respond to customers who protested after receiving multiple overdraft fees caused by high-to-low resequencing. These explanations were in plain English. At least for those angry few lucky enough to qualify for these responses, they provided a comprehensible disclosure of the bank's actual posting practices—after the fact.
Wells Fargo directed misleading propaganda at the class that likely led class members to expect that the actual posting order of their debit-card purchases would mirror the order in which they were transacted. For example, the "account activity" information provided to customers through online banking—a service made available to all Wells Fargo depositors—displayed "pending" debit-card transactions in chronological order (i.e., the order in which the transactions were authorized by Wells Fargo). When it came time to post them during the settlement process, however, the same transactions were not posted in chronological order but were posted in high-to-low order.
Misleading marketing materials promoted the same theme of chronological subtraction. A number of Wells Fargo marketing materials, including the Wells Fargo Welcome Jacket that was customarily provided to all customers who opened a consumer checking account, contained misleading representations regarding how debit-card transactions were processed. Specifically, these various materials—covered in detail in the findings of fact—communicated that debit-card POS purchases were deducted "immediately" or "automatically" from the user's checking account (see, e.g., TX 116, 129, 131, 143, 501, 559). Such representations would lead reasonable consumers to believe that the transactions would be deducted from their checking accounts in the sequence transacted.
This deception was then exploited by the shadow line. As Wells Fargo Senior Vice-President Karen Moore testified at trial, "[t]he available balance is created for the bank to understand the amount of money that we would make available to the customer for their next transaction" (Tr. 593). Wells Fargo's own website defined the term "available balance" to customers as "the most current picture of funds you have available for withdrawal" (id. at 1220-21). Once the shadow line was extended to debit-card purchases in May 2002, Wells Fargo began using it to approve into overdraft purchases that were previously declined. Customers who relied
Finally, Wells Fargo encouraged the use of registers in its marketing materials. This promoted the idea that transactions would be deducted chronologically.
In sum, Wells Fargo affirmatively reinforced the expectation that transactions were covered in the sequence made while obfuscating its contrary practice of posting transactions in high-to-low order to maximize the number of overdrafts assessed on customers.
Plaintiffs seek only injunctive relief for their claims of negligent misrepresentation and fraud. Given that such relief will be ordered under plaintiffs' Section 17200 claims, it is unnecessary for this order to reach these claims. As for plaintiffs' false advertising claim, since liability under the "fraudulent" restriction of Section 17200 has been established, liability for plaintiffs' false advertising claim under Section 17500—which is based upon the same deceptive conduct—has also been proven. See Tobacco II, 46 Cal.4th at 312, 93 Cal.Rptr.3d 559, 207 P.3d 20 ("A violation of the UCL's fraud prong is also a violation of the false advertising law.").
Wells Fargo argues that the National Bank Act and the implementing regulations of the Office of the Comptroller of the Currency (OCC) preempts all of the foregoing (Dkt. Nos. 38, 323). 12 U.S.C. 24 (Seventh); 12 C.F.R. §§ 7.4002, 7.4007, 7.4009. The undersigned rejected this argument on two prior occasions (Dkt. Nos. 98, 337). Based on a full trial record, this order again rejects it.
The Act vests national banks such as Wells Fargo with authority to exercise "all such incidental powers as shall be necessary to carry on the business of banking." 12 U.S.C. 24 (Seventh). The Supreme Court has "repeatedly made clear that federal control shields national banking from unduly burdensome and duplicative state regulation." Watters v. Wachovia Bank, N.A., 550 U.S. 1, 11, 127 S.Ct. 1559, 167 L.Ed.2d 389 (2007). That said, "[s]tates are permitted to regulate the activities of national banks where doing so does not prevent or significantly interfere with the national bank's or the national bank regulator's exercise of its federal powers." Id. at 12, 127 S.Ct. 1559. However, "when state prescriptions significantly impair the exercise of authority, enumerated or incidental under the [NBA], the state's regulations must give way." Ibid.
The OCC has the primary responsibility for the surveillance of the "business of
The Act and OCC regulations do not "preempt the field" of banking in its entirety. "Federally chartered banks are subject to state laws of general application in their daily business to the extent such laws do not conflict with the letter or the general purposes of the [Act]." Watters, 550 U.S. at 11, 127 S.Ct. 1559 (citations omitted). The Ninth Circuit recently emphasized this limitation on preemption, stating that "[s]tate laws of general application, which merely require all businesses (including national banks) to refrain from fraudulent, unfair, or illegal behavior, do not necessarily impair a bank's ability to exercise its [federally-authorized] powers." Martinez, 598 F.3d at 555.
As the Martinez decision explained, "the OCC has specifically cited [Section 17200] in an advisory letter cautioning banks that they may be subject to such laws that prohibit unfair or deceptive acts or practices." Ibid. (citing OCC Advisory Letter, Guidance on Unfair or Deceptive Acts or Practices, 2002 WL 521380, at *2, *7 n. 2 (Mar. 22, 2002)). The Ninth Circuit then listed a number of decisions exemplifying this limitation on preemption, including White v. Wachovia Bank, N.A., 563 F.Supp.2d 1358 (N.D.Ga.2008), which held that a claim under the Georgia Fair Business Practices Act that a bank engaged in unfair or deceptive business practices by manipulating the posting of transactions to an account in order to impose overdraft fees was not preempted by the Act or OCC regulations.
Wells Fargo's preemption defense focuses on three OCC regulations: 12 C.F.R. 7.4002, 7.4007, and 7.4009. This order will first address 12 C.F.R. 7.4002, which targets "National bank charges." Section 7.4002—formerly Section 7.8000 prior to 1996—was most recently amended in 2001.
Wells Fargo says that the resequencing practice challenged by plaintiffs is a component of "pricing" (i.e., is part of establishing "non-interest charges and fees" under this regulation) under 12 C.F.R. 7.4002. This is rejected.
There is a material difference between the bank's authority to establish overdraft fees—which includes the authority to set the amount of each fee and the method of calculating each fee—versus bank practices aimed at multiplying the number of overdrafts during the posting process. The recent Martinez decision illustrates the former situation. In Martinez, the plaintiffs challenged the alleged "overcharging of underwriting fees and the marking up of tax service fees" by Wells Fargo Home Mortgage. The plaintiffs argued that the bank's underwriting fees were excessive because they were not reasonably related to the bank's actual costs of performing the underwriting. The bank's tax service fees were argued as being "excessive" for analogous reasons. The Ninth Circuit held that both challenges were preempted by 12 C.F.R. 7.4002 because plaintiffs were "[i]n essence,... argu[ing] that these fees [were] too high, and [were] ask[ing] the court to decide how much an appropriate fee would be." Martinez, 598 F.3d at 556.
By contrast, plaintiffs in the instant case are neither challenging Wells Fargo's right to establish overdraft fees nor claiming that the amount of each individual overdraft fee is somehow excessive. Rather, plaintiffs target a practice whereby debit-card transactions are manipulated immediately prior to being posted so as to skyrocket the number of overdraft items posted against a customer's account. In other words, plaintiffs challenge bookkeeping practices that convert what is really one overdraft to count as ten overdrafts. This is a classic state law issue. Indeed, it was Wells Fargo that first raised California Commercial Code Section 4303 in its defense herein—only to learn that the legislative comment thereto barred the very practice at issue.
Equally fatal to this defense is an utter failure of proof. For preemption to possibly apply, 12 C.F.R. 7.4002 requires national banks to actually use a four-factor decision-making process when establishing non-interest charges and fees. Specifically, to invoke preemption under this regulation, a national bank must consider: (1) the cost incurred by the bank in providing the service, (2) the deterrence of misuse by customers of banking services, (3) the enhancement of the competitive position of the bank in accordance with the bank's business plan and marketing strategy, and (4) the maintenance of the safety and soundness of the institution.
As stated, however, Wells Fargo did not consider any of these factors when it decided to resequence debit-card transactions from highest-to-lowest dollar amount, commingle debit-card transactions with checks and ACH transactions, and extend the shadow line to the authorization of debit-card purchases. That no documentary evidence supported the bank's supposed consideration of these factors shows it was
Even if the practices challenged herein were somehow "pricing changes" under 12 C.F.R. 7.4002, a number of OCC interpretive letters recognize that state law still regulates conduct like that at issue here. For example, in OCC Interpretive Letter # 916 (May 22, 2001), the OCC expressly stated that the certain acts of misrepresentation by a bank could warrant a finding that the "good faith" limitation set forth in the comment to Section 4303(b) of the California Commercial Code was violated:
Similarly, in OCC Interpretive Letter # 997 (April 15, 2002), the OCC noted that "a relevant factor in evaluating good faith [under Texas Commercial Code § 4.303] may be whether a bank's actions were inconsistent with the practices it had represented to its customers that it would follow."
Both of the above interpretive letters expressly noted that state laws based upon UCC 4-303, including California Commercial Code § 4303, were "consistent with" and therefore did not conflict with federal laws governing national bank fees. Additionally, a number of OCC advisory letters militate against preemption. In Advisory Letter AL 2002-3, the OCC warned federal banks of the ramifications of engaging in unfair or deceptive practices that were in part violations of state law:
The letter went on to explain that the consequences of engaging in practices that may be unfair or deceptive under federal or state law can include litigation, enforcement actions, monetary judgments, and harm to the institution's reputation. The letter also stated that "[a] number of state laws prohibit unfair or deceptive acts or practices, and such laws may be applicable to insured depository institutions." Thus, the OCC itself contemplated that such suits will occur from time to time. In sum, the state-law restrictions applied herein are not preempted by 12 C.F.R. 7.4002.
Similarly, Wells Fargo has not proven preemption under 12 C.F.R. 7.4007 or
Furthermore, as held in a prior order (Dkt. No. 98), 12 C.F.R. 7.4007 only relates to a bank's deposit-taking powers and not to a bank's ability to reorder transactions in order to maximize overdraft fees during the posting process. None of the evidence presented at trial warrants a departure from this ruling. Additionally, the trial record is devoid of any credible evidence demonstrating that the state-law claims asserted in this action would impose more than an incidental effect on the exercise of Wells Fargo's deposit-taking powers or— as covered by 12 C.F.R. 7.4009—any other power authorized under the Act or OCC regulations.
Due to the passage of Proposition 64 in November 2004, a plaintiff must prove "actual reliance" to have standing to bring fraud-based Section 17200 claims on behalf of absent class members. See Cal. Bus. & Prof.Code § 17204. As explained by the California Supreme Court in Tobacco II, a plaintiff must show that the defendant's deceptions were "an immediate cause of the plaintiff's injury-producing conduct." The deception, however, does not have to be "the sole or even the predominant or decisive factor influencing [the plaintiff's] conduct." Instead, "it is enough that the representation ... played a substantial part, and so had been a substantial factor, in influencing [the plaintiff's] decision." Additionally, where a plaintiff has been exposed to "an extensive and long-term advertising campaign," proof of individualized reliance on specific misrepresentations or false statements is not required. Tobacco II, 46 Cal.4th at 326-28, 93 Cal.Rptr.3d 559, 207 P.3d 20 (citations and internal quotations omitted).
As explained in the Tobacco II decision, Proposition 64 was directed at "unscrupulous lawyers who exploited the generous standing requirement of the UCL to file `shakedown' suits to extort money from small businesses." Id. at 316, 93 Cal.Rptr.3d 559, 207 P.3d 20 (citation omitted). As such, the "intent of California voters [in passing Proposition 64] was to limit such abuses by `prohibiting private attorneys from filing lawsuits for unfair competition where they have no client who has been injured in fact.'" Id. at 316-17, 93 Cal.Rptr.3d 559, 207 P.3d 20 (citation omitted) (emphasis added). Proposition 64 amended Section 17204 of the California Business and Professions Code to authorize suits by any person "who has suffered injury in fact and has lost money and property as a result of such unfair competition." This changed the prior language, which granted standing to any plaintiff "acting for the interests of itself, its members or the general public[.]" In sum, the intent and effect of Proposition 64 was to "prevent uninjured private persons from suing for restitution on behalf of others." Id. at 314, 93 Cal.Rptr.3d 559, 207 P.3d 20 (citation omitted) (emphasis added).
In the instant case, it is clear that both Ms. Gutierrez—the class representative— and Ms. Walker have alleged and proven that they were harmed as a result of misrepresentations and omissions by Wells
Additionally, Wells Fargo's argument that the class-wide misrepresentation claims (and the exact documents they pertain to) must be surgically and precisely limited to those asserted by the class representative is not supported by the Tobacco II decision. The California Supreme Court clearly stated that "[r]epresentative parties who have a direct and substantial interest have standing; the question whether they may be allowed to present claims on behalf of others who have similar, but not identical, interests depends not on standing, but on an assessment of typicality and adequacy of representation." Id. at 319-20, 93 Cal.Rptr.3d 559, 207 P.3d 20 (emphasis added). Moreover, "there is nothing in ... Proposition 64 that purports to alter accepted principles of class action procedure that treat the issue of standing only to the class representative and not the absent class members[.]" Id. at 321, 93 Cal.Rptr.3d 559, 207 P.3d 20.
Here, both Ms. Gutierrez and Ms. Walker have standing to bring claims under Section 17200. Both relied upon the bank's deceptive omissions and marketing materials—including the Wells Fargo "Welcome Jacket" and CAA—that promoted the idea that transactions would be deducted chronologically rather than in high-to-low order. Both Ms. Gutierrez and Ms. Walker suffered actual losses of money as a result of the unfair and deceptive business practices discussed herein. As such, having met the "injury in fact" requirement imposed by Proposition 64, both Ms. Walker and Ms. Gutierrez have standing to bring their individual claims, and Ms. Gutierrez may properly present claims on behalf of class members with "similar, but not identical, interests."
Throughout this litigation, Wells Fargo has argued that any claims related to the extension of the shadow line to the authorization of debit-card purchases were foreclosed by the settlement and release negotiated in Smith v. Wells Fargo Bank, N.A. As noted in the findings, the settlement was approved by Judge Ronald S. Prager of the San Diego County Superior Court in November 2007 (Dkt. No. 431-1). Following the close of trial, both sides were invited to submit further briefing on this issue.
This order concludes that the Smith release does not bar any of the claims resolved herein. In their respective briefs, both Wells Fargo and plaintiffs agreed that this case is and has always been about the bank's high-to-low posting practices
Despite conceding that "the jurisdictional requirements of 28 U.S.C. § 1332(d)(2) are satisfied" in its answer (Dkt. No. 16), Wells Fargo recently filed a "notice" of a "potential jurisdictional defect" under the Class Action Fairness Act (CAFA), Pub.L. No. 109-2, 119 Stat. 4 (codified in scattered sections of 28 U.S.C.) (Dkt. No. 465). In its notice, the bank cited and attached the recent Eleventh Circuit decision in Cappuccitti v. DirecTV, Inc., 611 F.3d 1252, 1256 (11th Cir.2010), which held that "in a CAFA action originally filed in federal court, at least one of the plaintiffs must allege an amount in controversy that satisfies... 28 U.S.C. § 1332(a)." Plaintiffs filed a response a few days later (Dkt. No. 471).
Although no motion has been filed challenging subject-matter jurisdiction in this action, and the bank has clarified that it is not challenging jurisdiction herein (Dkt. No. 475), this order nevertheless concludes that the requirements for jurisdiction under CAFA have been met. The applicable code section clearly states:
28 U.S.C. 1332(d)(2). Nowhere in Section 1332(d) are the jurisdictional requirements of Section 1332(a) cross-referenced, except for the removal of "mass actions" from state court. See 28 U.S.C. 1332(d)(11)(B)(i) (defining a "mass action" as any civil action except a class action filed in district court under FRCP 23 in which monetary relief claims of 100 or more persons are proposed to be tried jointly on the ground that the plaintiffs' claims involve common questions of law or fact). The instant action was not removed from state court as a "mass action," but filed originally in district court as a class action under FRCP 23. Jurisdiction is therefore proper.
While this action has always been about Wells Fargo's high-to-low resequencing practices, the shadow line—as this order
The shadow line, however, has met its own fate. The Federal Reserve—through recent amendments to 12 C.F.R. 205 ("Regulation E") that have been effective since July 1, 2010—forced financial institutions like Wells Fargo to provide consumers with a choice regarding whether they wanted debit-card purchases to be authorized into overdrafts. This went far beyond simply requiring banks to provide a clear disclosure of their fees and terms associated with overdrafts. Rather, new Regulation E bars all shadow lines on debit-card purchases unless the customer opts in and affirmatively consents to having her debitcard purchases subject to shadow lines of credit.
The Board of Governors of the Federal Reserve System, who issued Regulation E pursuant to the Electronic Fund Transfer Act (15 U.S.C. 1693 et seq.), explained the need for these changes as follows:
Official Staff Commentary to Regulation E, 74 Fed. Reg. 59033, 59038-39 (Nov. 17, 2009).
In sum, the Federal Reserve recognized nearly all of the abuses found herein and concluded that if consumers truly wanted a service that subjected many customers to costly overdraft fees, they should be given a clear disclosure and the choice to affirmatively opt-in. While Regulation E did not address the high-to-low resequencing practices at issue in this class action, it is instructive for the relief now ordered.
Plaintiffs and the class seek a permanent injunction enjoining Wells Fargo from engaging in the unfair and fraudulent business practice found herein. This order finds that such relief is appropriate and authorized under Section 17203 of the California Business and Professions Code, and necessary to prevent further harm to Wells Fargo's customers. The Effective Date shall be
In addition to injunctive relief, this order finds that an award of restitution is warranted under Section 17203 to restore those overdraft fees paid by class members that were wrongfully extracted by Wells Fargo through its unfair practice as found herein.
Section 17203 states that "[t]he court may make such orders or judgments ... as may be necessary to restore to any person in interest any money or property, real or personal, which may have been acquired by means of such unfair competition."
That said, an award of restitution must be limited to "return[ing] money obtained through an unfair business practice to those persons in interest from whom the property was taken, that is, to persons who had an ownership interest in the property or those claiming through that person." Korea Supply Co. v. Lockheed Martin Corp., 29 Cal.4th 1134, 1144-45, 131 Cal.Rptr.2d 29, 63 P.3d 937 (2003) (citation omitted).
While a wide range of restitution scenarios were presented at trial by both sides, this order concludes that the proper measure of restitution must be measured from a posting order that most closely tracks a chronological posting order for debit-card transactions. Rejected is plaintiffs' argument that restitution should be based on the low-to-high ordering used by Wells Fargo prior to the high-to-low switch in May 2001. True, low-to-high posting is more favorable to depositors and was the system used before the bank switched to high-to-low resequencing. A main theme of plaintiffs' case, however, is that the bank promoted the expectation that debit-card transactions would post chronologically. As such, restitution based on chronological posting of these transactions will most closely track depositors' expectations.
This order finds that plaintiffs' expert Arthur Olsen has convincingly shown that it is entirely practical to re-run the computerized data in storage for each class members' account and determine how many overdrafts were added by the high-to-low practice for debit-card transactions during the class period. Indeed, he has already done so, using various alternate posting sequences. This has been done by him on an account-by-account, day-by-day, and transaction-by-transaction basis, using the bank's own real-world data. Court orders were needed to provide him access to this data, but—after much work and time—this order finds that Expert Olsen has done a professional and careful job in laying out the impacts of various alternative posting protocols. This work has not only demonstrated the enormous impact of the high-to-low scheme, but it has demonstrated that it is possible, in considering relief and restitution, to add back to depositors' specific accounts the amounts that were wrongfully taken by Wells Fargo, using posting protocols that this order finds would have tracked the ordinary and reasonable expectations of depositors.
In the Court's judgment, restitution should be made based on the following sequence:
This corresponds exactly to Olsen Scenario No. 2A (TX 212-G; Tr. 911-12).
This order finds that Olsen Scenario No. 2A properly posted credits and priority debit transactions in their usual and customary order—namely, first and second, before all debit-card transactions, checks, and ACH transactions (id. at 74, 287, 885-86; Dkt. No. 448-1). This order also finds that Wells Fargo has never posted checks and ACH transactions ahead of credits, or checks and ACH transactions ahead of all debit-card transactions within California (Dkt. No. 448-1). Rather, checks and ACH transactions were customarily posted after debit-card transactions prior to commingling, as done in Olsen Scenario No. 2A (ibid.). The only reason that Wells Fargo now suggests radical rearrangements and manipulations of credits, checks, and ACH transactions is to artificially minimize the restitution awarded to the class.
With respect to the fourth item above, it is worth noting that the bank did not maintain date/time information for a small percentage of its customers' debit-card transactions.
In his analysis, Expert Olsen calculated the differential between the number of overdraft fees that were actually assessed on customers during the class period and the number of overdraft fees that would have been assessed on customers using the above sequence. Overdraft fees that would have been assessed on customers no matter what ordering was used by the bank were excluded from the differential. For example, overdraft fees solely attributable to purchases authorized by the shadow line and not caused by high-to-low resequencing were excluded. Additionally, Expert Olsen considered uncollectibles— meaning overdraft fees that the bank was unable to collect from a particular account—in reducing restitution awards. Class members who had opted out were also excluded from his analysis (Tr. 894). This order finds that these adjustments were accurate and proper.
Expert Olsen also took into account overdraft fee reversals credited against each individual account using two different methods: (1) a last-in-first-out (LIFO) method, where any fee reversal was credited against the last overdraft fee assessed prior to the reversal, and (2) a 30-day
With respect to fee reversals, this order finds that a 30-day reversal period provides the fairest and proper method of restitution. Of the two methods employed by Expert Olsen to account for fee reversals, it is the most generous to the bank and, as the bank's own expert testified on direct examination, is the "superior" approach to crediting fee reversals against an award of restitution (id. at 1748).
Finally, this order finds that the shadow line per se has been properly ignored in Expert Olsen's analysis. While Wells Fargo argued at trial that plaintiffs' restitution studies failed to take into account the impact of the shadow line on returned-item fees, this is a red herring. First, given the secrecy of the shadow line, plaintiffs and their expert were never provided with the proper tools to apply the shadow line in their restitution studies. Instead, they were provided with a bare algorithm— without any accompanying instructions— that would take years to decipher and many guesses along the way. Even the bank's own restitution expert was not up to the task of programming the shadow line from scratch. Instead, the bank's expert relied upon Wells Fargo's internal risk management group and pre-programmed computer system to perform his forensic shadow-line calculations—a luxury denied to plaintiffs' expert. Second, all of the bank's analyses were based upon a mere ten days worth of customer data (as compared to the 43-months of data analyzed by Expert Olsen). This strips the bank's arguments of all weight and credibility.
Restitution must be paid by the bank on a class-member-by-class-member basis, based upon the difference between the foregoing posting sequence versus the result of the high-to-low system used during the class period. The Olsen study demonstrates that this amount, in total, will be close to $203 million.
The next question is how procedurally to refund the money. For accounts due restitution and still open at Wells Fargo, the bank shall simply post a credit to the accounts. For accounts that have been closed but whose owner still banks with Wells Fargo in any capacity, the bank shall credit one of the owner's active accounts. For class members no longer doing business with Wells Fargo and to whom restitution is owed, the bank shall provide class counsel with their last known U.S. mail address, last known email address, and last known telephone number.
Counsel shall meet and confer and recommend a detailed plan of distribution and notice, including a plan to track down class members for whom the available contact information is inadequate, such plan to include use of the National Change of Address Registry and published notice. Counsel shall also meet and confer and produce a computerized listing of each class member and the amount of restitution due to each using the foregoing posting protocol, reserving all objections as to substance. Counsel should consider whether restitution should be extended up to the date of the injunction's Effective Date rather than merely up to the end of
The Court will retain jurisdiction to enforce this order.
