2:11-bk-28944-DPC Adversary 2:13-ap-00799
NOT FOR PUBLICATION
Daniel P. Collins, Bankruptcy Judge
TABLE OF CONTENTS
I. SUMMARY OF LAWSUIT AND COURT FINDINGS........................................5
IV. POTENTIAL DYNAMIX'S BANKRUPTCY......................................................12
V. ADVERSARY PROCEEDING PROCEDURAL HISTORY...............................15
A. The Complaint and Answer....................................................................................15
B. Amazon's Counterclaims.......................................................................................16
C. Discovery Disputes................................................................................................16
D. Motions for Summary Judgment............................................................................16
E. Motions in Limine..................................................................................................18
VI. THE TRIAL...........................................................................................................19
A. Trial Exhibits..........................................................................................................19
B. Trustee's Witnesses................................................................................................20
1. Timothy Shaffer..................................................................................................20
2. Serena Mor ones..................................................................................................21
a) Morones' Damages Report..............................................................................22
b) Morones February 5, 2021, Declaration..........................................................27
c) Morones' Direct, Cross and Re-Direct Examinations.....................................32
3. Thomas Azzarelli................................................................................................38
4. Justin Ice.............................................................................................................39
5. Tasha Bachand....................................................................................................41
6. Jeff Moore...........................................................................................................43
7. Eric Soder............................................................................................................44
8. Jeffrey Cone........................................................................................................45
9. Stephen Ashworth...............................................................................................46
10. Sean Lawcock.....................................................................................................47
11. Daniel A. Bellino................................................................................................49
C. Motion for Directed Verdict...................................................................................49
D. Amazon's Witnesses..............................................................................................50
1. Tasha Bachand....................................................................................................50
2. E. Weiant Williams.............................................................................................52
a) Williams' Rebuttal Report of September 14, 2020.........................................52
b) Williams Declaration.......................................................................................57
c) Williams' Cross and Re-Direct Examinations................................................58
3. Timothy Shaffer..................................................................................................60
4. Justin Ice.............................................................................................................61
5. Jeff Moore...........................................................................................................63
6. Eric Soder............................................................................................................64
7. Jeffrey Cone........................................................................................................66
8. Stephen Ashworth...............................................................................................67
9. Thomas Reilly.....................................................................................................68
10. Sean Lawcock.....................................................................................................69
11. Daniel A. Bellino................................................................................................70
E. Post-Trial Proceedings...........................................................................................71
F. Tentative Under Advisement Order.......................................................................72
VII. ANALYSIS OF TRUSTEE'S CAUSES OF ACTION.........................................72
A. Count 1: Failure to Turnover Estate Property Pursuant to § 542(a)......................72
1. Legal Analysis....................................................................................................72
2. Conclusion on Turnover Claims.........................................................................73
B. Count 2: Violation of the Automatic Stay Pursuant to § 362................................73
1. Legal Analysis....................................................................................................73
2. Conclusion on the Stay Violation Claims...........................................................77
C. Count 3: Breach of Contract...................................................................................77
1. Legal Analysis....................................................................................................77
2. Burden of Proof...................................................................................................78
a) Burden Shifting...............................................................................................79
b) Conclusions Regarding Burden of Proof.........................................................82
3. Findings Regarding Breach of Contract and Damages.......................................83
a) Calculating Replacement Value......................................................................83
b) Cutoff Date of January 31, 2014.....................................................................84
4. Amazon's Contract Breach.................................................................................86
a) The Settlement Data Was Not Introduced Into Evidence...............................86
b) Plaintiffs Proven Inventory Damages............................................................90
c) The Court's Analysis of Codes Used by Amazon and the Impact on Claimed Damages.................................................................................................................92
(1) Code M.........................................................................................................92
(2) Code D.........................................................................................................94
(3) Code Q.........................................................................................................95
d) Unpaid Refunds...............................................................................................97
e) Sales Proceeds Not Remitted to Debtor..........................................................98
f) Trustee Did Not Carry Its Burden of Proof on the Basis of Amazon's Motivation for Termination of the Contract...........................................................99
g) Unclean Hands.................................................................................................99
h) Money Claimed Owing by Debtor to Amazon...............................................99
5. Conclusion on Breach of Contract Claims.......................................................100
D. Prejudgment Interest on Damages........................................................................100
a) Legal Analysis...............................................................................................101
b) Conclusion on Prejudgment Interest.............................................................103
VIII. SUMMARY OF CONCLUSIONS......................................................................103
I. SUMMARY OF LAWSUIT AND COURT FINDINGS
This Adversary Proceeding involves a nine count Complaint brought by Timothy H. Shaffer, Trustee of the Chapter 11 Debtor, Potential Dynamix, LLC against Amazon Services, LLC on July 9, 2013. Over the course of this Adversary Proceeding, Plaintiffs claims have been boiled down to claims for (1) breach of contract, (2) turnover of property to Plaintiff, and (3) a claim for a violation of the bankruptcy automatic stay.
Defined terms in this Order are summarized in Attachment 1.
DE 1. "DE" means docket entry in Adversary No. 2:13-ap-00799-DPC ("Adversary Proceeding").
Plaintiff alleges Debtor placed roughly 6.4 million units of product in Amazon's hands for sale on Amazon's online platform but was only paid by Amazon for approximately 5.9 million units. Plaintiff contends (and Defendant does not disagree) that the average retail price of these products was $22 per unit from which Debtor was paid $16.31 per unit and Amazon received $5.69 per unit. The heart of the parties' disputes concern Amazon's accounting related to the inventory the Debtor placed with Amazon over the course of their eight-year relationship. Plaintiff contends Amazon has the burden of demonstrating where all its inventory went, and that Plaintiff is entitled to recover $16.31 for every unaccounted unit placed in Amazon's hands. Plaintiff seeks breach of contract and turnover damages of $2,869,663 plus interest at 12% per annum from the midpoint of their relationship. Amazon contends Plaintiff carries the burden of proving all matters pertinent to the claimed breach of contract and Plaintiffs damages and that Plaintiff failed to do so. The Court resolves the burden of proof issues in favor of Defendant.
Plaintiffs final claim is for damages Debtor sustained when Amazon violated the bankruptcy automatic stay by terminating Debtor's access to Amazon's online sales platform on April 11, 2013. Years ago, this Court found that Amazon's actions willfully violated the automatic stay. Plaintiff seeks stay violation damages of $732,000 plus interest from the date of judgment.
Following years of discovery and motion practice, these matters finally came to trial in 2021. The trial was conducted virtually on the Court's ZoomForGovernment platform. Once the four-day trial concluded, the parties briefed certain issues. The Court held oral argument on those briefs. The Court issued a tentative ruling and invited the parties' critiques. On February 24, 2022, the Court heard oral argument on the parties' critiques. After oral arguments, the Court took this matter under advisement.
Counsel for both sides (and their staff) masterfully conducted this virtual trial at about the time COVID infections were reaching their 2021 post-holiday peak.
After reviewing the contracts between the parties, the vast amount of Amazon's inventory control data, the expert reports and other exhibits submitted at trial and after hearing and/or reading the testimony of numerous witnesses as well as considering the parties' pleadings and 122-page Joint Pretrial Statement, this Court finds that Amazon is liable to Plaintiff in the amount of $668,484 on account of damages it caused to Debtor when it violated the stay by terminating Debtor's access to Amazon's online platform. The Court also finds Amazon breached its Contract with Debtor by failing to pay or turnover to Debtor 20,405 units of product which were placed in Amazon's hands by Debtor but, as of March 31, 2015, had never been returned to Debtor or paid for by Amazon. Nothing produced to this Court indicates these units were unsellable. Debtor should have been able to realize $16.31/unit from the sales of these units. Amazon breached its Contract with Debtor with respect to the 20,405 units and is liable to Debtor for the sum of $332,806. Judgment shall be awarded against Amazon and in Plaintiffs favor on these claims in the aggregate amount of $1,001,290. Attorney's fees will not be awarded to either party as neither sought fees in connection with this Adversary Proceeding. Debtor's stay violation damages were unliquidated so this Court will not award prejudgment interest on the stay violation damages but Debtor's breach of contract damages were liquidated as of April 1, 2015, so interest shall accrue on the breach of Contract damages of $332,806 at 12% per annum from April 1, 2015, until paid.
Attachment 2 is the list of Exhibits admitted at trial.
20,405 x $16.31/unit.
Under the parties' contract, Washington law governs the contract. Washington's statutory judgment rate is 12% per annum.
This Order constitutes the Court's findings of fact and conclusions of law pursuant to Rule 7052 of the Rules of Bankruptcy Procedure.
Debtor was formed in 2006 in a basement in Springfield, Missouri by Bellino and Schmidt. Later that same year, the company moved to Phoenix, Arizona. Bellino and Schmidt were each 50% members of the Debtor. The business is sometimes referred to as "DAB."
Admin DE 219, pg. 4; Admin DE 80, pg. 1. "Admin DE" references a docket entry in the Debtor's administrative bankruptcy proceeding, 2:11-bk-28944-DPC.
Admin DE 219, pg. 4. Even though the company moved to Phoenix, Arizona, it was organized under the laws of Nevada.
The initials of Bellino.
The Debtor primarily sold health and personal care items and did so exclusively online via amazon.com through two programs offered by Amazon: (1) Amazon's Merchant Fulfilled Network and (2) Amazon's Fulfillment by Amazon Program The relationship between the parties was governed by the Merchants@Amazon.com Program Agreement and the Amazon Services Business Solutions Agreement. The FBA Program was governed by of the ABSA which is referred to as the FBA Agreement. The FBA Agreement incorporates applicable "Program Policies." Included in these "Program Policies" are the "Inventory Adjustment" codes which were identified in the Seller Central "Help Page."
Admin DE 219, pg. 4. See also DE 332, pg. 14, ¶ 6.
DE 332, pg. 17, ¶ 24. Exhibit 1.
DE 332, pg. 18, ¶ 2.
Trial Ex. 1 at Bates page No. 1.0001.
These Adjustment Codes are found at Trial Ex. 2. These Adjustment Codes were revised by Amazon's Bachand in 2018, well after the Amazon-Debtor relationship terminated. The Court will not consider these 2018 revisions. The 2018 revisions are found at Trial Ex. 261.
Through the Merchant Program, the Debtor offered items for sale on amazon.com but shipped the items itself and handled their own customer service. Through the FBA Program, Debtor shipped inventory to Amazon and offered that inventory for sale on amazon.com and paid Amazon to receive, pick, pack, ship, and provide customer service for the Debtor's products.
DE 332, pg. 14, ¶ 7.
DE 332, pg. 15, ¶ 8.
Debtor started its relationship with Amazon in 2008 using the Amazon Merchant Program. In 2010, the Debtor switched from the Amazon Merchant Program to the FBA Program. Debtor was unprepared for the problems that would follow with respect to Debtor's internal systems and the new cost structures that came with the FBA Program. As a reseller of health and personal care products, Debtor sold 31,166 distinct products.It is the Court's understanding that Debtor was one of Amazon's biggest customers in terms of number of products listed for sale on amazon.com.
Admin DE 219, pg. 5. This Disclosure Statement incorrectly notes Debtor moved from an Amazon fulfillment to a Merchant fulfillment. It was just the opposite, but the implications were the same: chaos ensued. See Reilly's deposition testimony described below.
DE 332, pg. 18, ¶ 1.
See DE 396, pg. 65.
Amazon assigned a unique number to each product. These numbers are known as "Amazon Standard Identification Number" Each product also had its own "Stock-Keeping Units" barcode. Because other customers of Amazon also sold some of these exact same products (same ASIN, same SKU), Amazon mixed some of Debtor's inventory with inventory owned by third parties. This "comingling" was approved by the Debtor and became a part of the process for fulfilling customer orders. Since Amazon has fulfillment centers (warehouses) all over the world, if a customer of Debtor ordered a particular product, that sale might be fulfilled not by sending the customer a unit shipped by Debtor to Amazon at fulfillment center A, but rather, by a unit shipped to Amazon by a third party to fulfillment center B, which was closer to the customer's desired delivery point.
DE 332, pg. 18, ¶ 1.
"Comingling" is a "virtual" concept in that it refers to Amazon fulfilling a customer's order with the most readily available product regardless of whether that product was shipped to Amazon by the Seller from which the customer bought the product. See DE 397, pg. 27-8. Physical comingling may also occur, but the Court understands that, in Debtor's relationship with Amazon, comingling is more a "virtual" concept.
Over the course of their eight-year relationship, the Debtor placed 6,316,429 product units into Amazon's hands, 5,978,113 of which were sold. 63,445 of the sold units were the subject of refunds/sale reversals requested by the customer. Through January 31, 2014, 344,906 units were "Removed," i.e., either destroyed by Amazon at Debtor's request or shipped back to Debtor at Debtor's request. Debtor's expert, Morones, noted that Debtor's inventory was reduced by an additional 17,524 units through "Adjustments." This series of events led Debtor's damages expert to set the following table in her Damages Report:
On October 22, 2013, Amazon finally terminated this relationship. See Ex. 144.
A unit might be destroyed, for example, if it was a consumable product whose useful life had expired.
Amazon contends that, by May 2015, the number of units reduced through "Adjustments" were 16,567. See DE 397, pg. 29 and Trial Ex. 5, ¶ 33, Table 2.
(1) Expected Ending Inventory
Customer Returns 63,445
As of January 13, 2014. After that date, an additional 19,890 units were Removed. See § V(D) below for further discussion on this point.
These inventory related facts are uncontested. What is contested is Morones' damages amounts referred to in her Report as (2) Reimbursable Adjustments, (3) Unpaid Refund Reimbursements, and (4) Sales Proceeds not remitted to Debtor. Amazon also challenges Trustee's claim for prejudgment interest.
Amazon was paid $128,224,650 from the sale of Debtor's units and, in turn, paid Debtor $95,048,175 after taking its storage and handling fees of $33,176,475. Overall, the average sales price was $22 per unit of which the Debtor was paid $16.31 per unit and Amazon kept $5.69 per unit. These facts are uncontested.
The complexity of this case centers around the massive volume of data pertaining to these units and whether the many, many coded transactions can be relied upon as accurately describing what happened to those units. For example, did Debtor ship fully sellable units to Amazon? Did Amazon accurately input those units into Amazon's electronic system? How many units were lost, stolen, damaged, expired, etc. while in Amazon's possession? If a customer was refunded their purchase price did Amazon receive that unit back from the customer? Did Amazon pay Debtor the full amount due to the Debtor? . . . and on and on.
As the parties' relationship carried on, Amazon produced inventory, sales and payment data and posted it for Debtor to review on an electronic platform known as Seller Central. The information disseminated on Seller Central is referred to as the Seller Central Data. Within the body of Seller Central Data is a subset of information focused upon sales transactions. This is referred to as the Settlement Data. While Amazon had a duty to provide data to Debtor contemporaneous with the event affecting Debtor's inventory,Amazon did not preserve all this data, nor did it apparently have the ongoing contractual obligation to do so. For its part, Debtor apparently never challenged the Seller Central Data although Azzarelli testified that, prior to the Petition Date, Debtor did ask Amazon questions about Debtor's perception that its inventory in Amazon's hands was reported to be lower than Debtor thought it should be.
See DE 396, p. 61, ll. 4-13. See also Ashworth Deposition at 43:13-24.
Williams' Testimony at trial.
Plaintiff has not suggested that Amazon has committed evidence spoilation, but Plaintiffs lawyers were obviously displeased in this Adversary Proceeding that Amazon did not (could not?) produce a complete Seller Central Data set from the outset of the parties' business relationship. During the discovery phase of this Adversary Proceeding, Amazon did produce the M15 Data in May 2015, but the M15 Data does not include Settlement Data. By the time this Adversary Proceeding was filed the Seller Central Data, according to Amazon employee Bachand, did not include a full or accurate set of the Settlement Data. The Court does not find, nor does the Trustee suggest, that the Debtor's or Trustee's challenges to Amazon's inventory control or sales data was time barred.
Azzarelli Deposition at page 81.
As various events came to affect a given product unit, Amazon would enter a code applicable to the subject transaction or event. For example, a Code M indicates the unit is misplaced, Code D means the unit was destroyed and Code F notes the unit was found. Attached to this Order as Attachment 3 are two trial exhibits reflecting the Amazon "Inventory Adjustment" codes and a description of the meaning of those codes. One Inventory Adjustment code in particular was heavily used by Amazon, Code Q. Code Q signaled "Damaged - Miscellaneous." Amazon assigned this code to reflect "a decrease of [Debtor's] sellable inventory when damage cannot be attributable to a source." Yet the Inventory Adjustment codes in effect through 2013 (when the parties' relationship terminated) do not reflect that Amazon disclaimed responsibility for Code Q units. At trial, Amazon did not take responsibility for Code Q units. 166,279 units were given a Code Q designation. Much of the debate between the parties relates to Code Q units, especially the 147,968 sellable Code Q units. Some Inventory Adjustment codes called for Debtor to bear the loss or cost or impact of a code designation to that unit. Other Inventory Adjustment codes called for Amazon to reimburse Debtor for the unit. Amazon's Reimbursement to Debtor could be effectuated by paying the Debtor or adding that exact same kind of unit to Debtor's inventory.
Trial Exs. 2 and 147.
Morones Declaration, page 10, ¶ 26.
See Attachment 3.
At $16.31 per unit this is a matter of $2,413,470.
Exs. 131 (the M15 Data) and 161 (the Settlement Data).
Same ASIN and/or SKU.
During the course of the discovery phase of this Adversary Proceeding, Amazon created a May 2015 transaction data report which came to be known as the M15 Data.The M15 Data is contained in gigantic Excel spreadsheets but is incomprehensible to all but the deepest data divers. To make matters more interesting, the M15 Data, while identified by Amazon as "the most complete information known to be available to account for Debtor's units in the Fulfillment by Amazon Program," nevertheless does not contain all the data pertaining to all of Debtor's inventory nor does the M15 Data perfectly match the Seller Central Data. In particular, the M15 Data does not contain any unit sales data. That information is referred to as the Settlement Data. The Settlement Data is a subset of the Seller Central Data. Plaintiff and Defendant hired experts to make sense of the colossal volume of inventory data with an eye towards establishing (or refuting) Plaintiffs breach of contract damages (and, therefore, turnover damages) caused by Amazon, if any.
Ex. 5, PDF page 8 of 87.
Ex. 5 (Williams' Report), PDF page 8 of 87, ll. 1-4, citing at n.4, Amazon's Amended Responses to Trustee's Third Set of Discovery Requests, dated September 20, 2016, at 6. This document can be found at Trial Ex. 118 at Bates No. 118.0036. This document is defined as "Amazon's September 20, 2016 Discovery Response." See also Ex. 7 (Morones' Damages Report), page 5 of 267, ¶ 9.
Amazon's September 20, 2016 Discovery Response. This document makes clear that the M15 Data tracks the movement of "units." It does not purport to deal with sales or payments. The Debtor and Trustee knew or should have known sales information was contained in the Settlement Data.
The parties stipulated to this Court's jurisdiction to enter final orders adjudicating and disposing of all core and non-core claims contained in the Trustee's Complaint.Additionally, this Court finds it has jurisdiction over this Adversary Proceeding pursuant to 28 U.S.C. § 1334 and 157(B)(2).
DE 60. See also DE 332, the Joint Pretrial Statement, Section I(c), page 16 of 122.
IV. POTENTIAL DYNAMIX'S BANKRUPTCY
Disagreements between Debtor's owners, Bellino and Schmidt, arose along the way. This resulted in Schmidt filing a lawsuit in Arizona Superior Court, Maricopa County against the Debtor, Bellino, and Tara Bellino.
Admin DE 219, pg. 6 (case no. CV2011-014287). Bellino and Schmidt eventually settled the State Court lawsuit. Admin DEs 183 & 197.
After borrowing on its line of credit until it was 100% utilized and with sales plummeting and the Bellino versus Schmidt fight brewing, the Debtor quickly ran out of money. Debtor filed its chapter 11 bankruptcy petition on October 13, 2011.
The Bankruptcy Proceeding which was filed in this District at Case No. 2:11-bk-28944.
Admin DE 219, pg. 5-6; Admin DE 1. This date is referred herein as the Petition Date. The Bankruptcy Proceeding was initially assigned to Judge Charles G. Case. Upon his retirement, the case was reassigned to the undersigned on January 18, 2013.
Animosity between Bellino and Schmidt eventually made its way into this Court.Schmidt filed a Motion to Dismiss the Bankruptcy Proceeding asserting his approval was not obtained prior to filing Debtor's chapter 11 bankruptcy petition, as required by the Debtor's operating agreement. Schmidt claimed he did not ratify the bankruptcy filing. In its Response, Debtor alleged Schmidt forged Bellino's signature on the Debtor's operating agreement and that management rested with Bellino because he was the Debtor's majority member.
Admin DE 219, pg. 6.
Admin DE 80.
Admin DE 80.
Admin DE 86.
The ongoing dispute between Bellino and Schmidt was disruptive in the bankruptcy proceeding to the point that it caused the Official Committee of Unsecured Creditors to request the appointment of a chapter 11 trustee to oversee the Debtor's bankruptcy case. Although Bellino and Schmidt disagreed on many things, they agreed that the appointment of a trustee was the best course for this Bankruptcy Proceeding.The Court approved the appointment of a chapter 11 trustee pursuant to Fed.R.Bankr.P. 2007.1(c) and § 1104(a). Subsequently, the Trustee was appointed as the Debtor's chapter 11 trustee to administer the affairs of the Debtor's bankruptcy estate.
The United States Trustee appointed the Unsecured Committee on November 9, 2011, and later amended it due to the resignation of creditor Bias Distributors. Admin DEs 51 & 64. Prior to his judicial appointment, Paul Sala was hired as counsel for the Unsecured Committee. Admin DEs 72 & 77.
Admin DE 89.
Admin DE 102.
Admin DE 104.
Admin DE 110. The Trustee was selected by the UST.
The Trustee filed a Disclosure Statement and Plan which provided for the assumption of the Contract between Amazon and the Debtor. A hearing on approval of the Disclosure Statement was held on April 29, 2013. Although that hearing was continued to May 28, 2013, the hearing was vacated because, on April 12, 2013, Amazon unilaterally terminated the Contract, revoking the Debtor's ability to sell its products on Amazon's online platform. Amazon contended the Debtor was using Amazon's platform to sell "restricted items," as defined in Section 1.1 of the Contract. The Trustee sought the Court's intervention. The Court ordered Amazon to immediately reinstate the Contract and Debtor's access to Amazon's online platform. Additionally, the Court found Amazon's conduct constituted a willful violation of the bankruptcy automatic stay. Due to Amazon halting Debtor's access to Amazon's online platform, Debtor was unable to sell its products on that platform from April 11, 2013 to April 26, 2013, a total of 15 days.
Admin DE 219. Trial Ex. 222 introduced by Amazon is actually a disclosure statement filed in error at Admin. 218. The Court takes judicial notice of the correct Disclosure Statement filed at Admin DE 219. The Court also takes judicial notice of all other pleadings filed in this Adversary Proceeding and on the administrative docket in Debtor's Bankruptcy Proceeding to the extent those pleadings were not otherwise admitted into evidence at trial.
Admin DE 217.
Admin. DE 265.
Admin DE 243. Trustee's Trial Exhibit 144 includes a copy of the April 11, 2013, letter terminating the Contract.
Admin DEs 243 and 244.
Admin DE 262.
Admin DE 262.
DE 332, pg. 16, ¶ 20.
After reinstating Debtor's access to its online platform, Amazon filed a Motion for Relief from Stay seeking to terminate the Contract pursuant to § 362(d)(1). Debtor, in turn, filed a Motion to Assume the Contract pursuant to § 365(a). The Trustee sought summary judgment granting the Debtor's Motion to Assume the Contract and denying Amazon's Motion for Relief from Stay. Amazon sought summary judgment opposing the assumption of the Contract and seeking relief from the automatic stay so it could terminate the Contract. The Court granted summary judgment in favor of Amazon, holding the Debtor could not assume the Contract. Amazon was granted stay relief to send the Debtor a letter terminating the Contract pursuant to Section 10.1(y) of the Contract. On October 22, 2013, Amazon terminated its Contract with the Debtor.Conceding defeat, in January 2014 the Debtor ceased operations.
Admin DE 253. The Unsecured Committee filed a Joinder in Support of Debtor's Motion to Assume the Contract. Admin DE 276.
Admin DE 266.
Admin DE 318.
Admin DE 325.
Admin DE 364.
Admin DE 398. The Court's order granting summary judgment in favor of Amazon was subsequently amended to revise the language in footnote 1, which now reads:
Solely for the purpose of deciding these motions for summary judgment, the Trustee, to streamline the issues and advance the litigation process, does not dispute the facts recounted in this order. However, the Trustee states that "if this matter proceeds to trial, [he] may present evidence to the contrary." (Trustee's Controverting Statement of Facts, Dkt. No. 333). Accordingly, the findings of fact and conclusions of law set forth in this decision are made for the purpose of deciding these cross-motions for summary judgment only and, except to the extent such findings are necessary predicates for the Court's ultimate rulings on the cross-motions for summary judgment, shall not serve as res judicata, collateral estoppel or law of the case in other proceedings before this Court.
Ex. 144. Also, Trustee's Trial Exhibit 152 includes a copy of the Termination Letter.
DE 332, pg. 17, ¶ 23. The most recent monthly operating report shows the Debtor's total cash on hand to be $581. Admin DE 660.
To this day, neither the Disclosure Statement nor the Plan have been approved or rejected by the Court. Nevertheless, Debtor remains in a chapter 11 proceeding.
V. ADVERSARY PROCEEDING PROCEDURAL HISTORY
A. The Complaint and Answer
Prior to commencement of this Adversary Proceeding, Debtor alleged that Amazon failed to account for a considerable amount of the Debtor's inventory. This contentious issue together with Amazon's automatic stay violation drove the Trustee to file the Complaint. The Court eventually dismissed Counts 4, 5, 6, 7, 8, and 9 of the Complaint (discussed further in Subsection D below).
Admin DE 243, pg. 4, ¶ 12.
DE 91 dismissed counts 4, 6, 7, 8, and 9, and DE 348 dismissed count 5.
Only counts 1, 2, and 3 remain from the Trustee's Complaint. Count 1 alleges Amazon failed to turnover estate property pursuant to 11 U.S.C. § 542(a). Count 2 alleges Amazon violated the automatic stay pursuant to 11 U.S.C. § 362. Count 3 alleges Amazon breached the Contract by terminating the Contract without sending a Notice of Material Default to the Debtor, by denying Debtor access to the Amazon platform, by not accounting and compensating Debtor for inventory placed in Amazon's hands, by failing to store, maintain, preserve, and account for Debtor's inventory and by failing to compensate for that inventory.
Unless indicated otherwise, statutory citations refer to the U.S. Bankruptcy Code ("Bankruptcy Code"), 11 U.S.C §§ 101-1532, and to the Federal Rules of Bankruptcy Procedure, Rules 1001-9037.
Amazon filed an answer that included two (2) counterclaims against the Trustee and asserted the Debtor had "unclean hands" because it first breached the Contract. Amazon's Answer was later amended. Amazon's counterclaims allege that, to the extent Amazon demonstrates any liability to Amazon by the Debtor, the Trustee, and/or the Debtor's bankruptcy estate, Amazon must be allowed to setoff or recoup such liability against any liability which Amazon owes to the Debtor, the Trustee, or the Debtor's bankruptcy estate.
DE 5, ¶¶ 11-14.
B. Amazon's Counterclaims
DE 332 at pg. 9 and DE 397 at pgs. 5-6.
C. Discovery Disputes
Over the course of this Adversary Proceeding, the Court had occasion to address numerous discovery disputes between the parties. Those disputes were generally resolved without motions to compel or motions for protective orders. No attorney's fees or sanctions were awarded by the Court as a part of those discovery hearings nor did the Court reserve fee or sanctions awards for disposition at the end of this Adversary Proceeding.
By this Court's count, it heard nine discovery disputes. See DEs 123, 156, 173, 182, 188, 209, 251, 335 and 358.
The Court's procedure page and Local Rule 9013-1 require the parties meet and confer before calling the Court for a resolution of discovery issues on the spot, without the filing of motions to compel or for protective orders.
D. Motions for Summary Judgment
Amazon moved for summary judgment on Counts 2, 3, 4, 6, 7, 8, and 9 of the Complaint. The Trustee filed his Response and Amazon filed its Reply. On September 16, 2015, the Court issued an order granting summary judgment in favor of Amazon dismissing Counts 4, 6, 7, 8, and 9. Additionally, the Court granted partial summary judgment on Count 2 (stay violation) to the extent it sought punitive damages.
DE 91. Those counts were as follows: Count 4 - Breach of Implied Covenant of Good Faith and Fair Dealing; Count 6 - Injunction Against Further Stay Violations; Count 7 - Specific Performance; Count 8 - Negligence; and Count 9 - Intentional Interference with Business Expectancy.
On November 20, 2020, the Trustee filed his motion for summary judgment, requesting the Court find that, in determining Amazon's liability, the proper cutoff date is January 31, 2014 and the proper methodology requires an assessment of both "expected ending inventory" and "inventory shrinkage." The Trustee also requested the Court find that Amazon, at minimum, is liable to pay the Debtor the replacement value of 111,688 units of damaged, lost, or missing inventory. Amazon filed its Response. The Trustee filed his Reply. At its February 11, 2021 hearing, the Court denied the Trustee's motion for summary judgment in all respects. The Court specifically noted that it would not set a January 31, 2014 hard stop on the calculations of damages. This is important in that Amazon "Removed" 19,890 units to Debtor after January 31, 2014.
The report from Amazon's expert Williams contains a table reflecting unit totals through May 2015. See Trial Ex. 5, ¶ 33, table 2.
Amazon filed an additional motion for summary judgment seeking: (1) dismissal of Count 5 of the Complaint (accounting for the inventory); (2) dismissal of Count 3 (breach of contract), contending the Debtor failed to provide accurate information to Amazon thereby materially breaching the contract and discharging further performance by Amazon; and (3) enforcement of the Contract's limitation of liability clause. The Trustee filed his Response in opposition. Amazon filed its Reply.
The Court heard oral arguments on February 11, 2021. At that hearing, the Court (1) granted Amazon's request to dismiss Count 5, (2) denied Amazon's motion for summary judgment as to the contention that Debtor failed to provide it accurate information, and (3) took under advisement the issue pertaining to enforceability of the Limitation Clause. On February 15, 2021, the Court issued an under-advisement order denying Amazon's request to enforce the Limitation Clause because (1) the Limitation Clause is unclear and unreasonable, (2) at the time the Contract was executed damages were not unascertainable, and (3) enforcing the Limitation Clause would violate public policy.
E. Motions in Limine
The Trustee filed his Motion in Limine to Preclude the Admission of "Deep Dive" Documents into Evidence. Amazon filed its Response. The Trustee filed his Reply. At the July 8, 2015, hearing, the Trustee briefly discussed the First Motion in Limine and advised it would not be argued at the hearing. The hearing was then vacated, subject to call. The First Motion in Limine was never rescheduled.
DE 49 filed April 13, 2015.
DE 89. Interestingly, at trial the Trustee sought to admit into evidence excerpts of the "Deep Dive" data and Amazon objected, asserting the excerpts were attached to Justin Ice's deposition and he failed to establish foundation or authenticity. DE 332, Ex. A, pg. 3. The excerpts of the "Deep Dive" data were not admitted into evidence.
On November 20, 2020, Amazon filed its Motion in Limine to Exclude Unpleaded and Untimely Claims and Damages Theories. On the same day, Amazon filed its Motion in Limine to Exclude the Testimony of Expert Serena Morones. The Trustee filed his Responses to the Second Motion in Limine and the Third Motion in Limine.Amazon filed its Replies in Support of the Second Motion in Limine and the Third Motion in Limine.
At the January 21, 2021, hearing, the Court placed its findings and rulings on the record and denied the Third Motion in Limine because among other things, the Court concluded Morones possessed the requisite specialized knowledge called for under Rule 702(a) of the Federal Rules of Evidence, that her expert report was the product of reliable principles and methods, and that she reliably applied those principles and methods. If the Defendant believed Morones was not armed with sufficient data to form her expert opinions, the Court held that would be a matter best revealed through well-crafted cross-examination by Defendant's counsel at trial. Subsequently, at the February 11, 2021, hearing, the Court placed its findings and rulings on the record and denied the Second Motion in Limine.
VI. THE TRIAL
Beginning on February 16, 2021, and concluding on February 19, 2021, the Court conducted a four-day trial. Only four witnesses gave live testimony. The bigger end of witness testimony was introduced via video depositions and designated portions of deposition transcripts from 11 witnesses. The parties' closing arguments were heard on February 24, 2021.
A. Trial Exhibits
The Parties' Joint Pre-Trial Statement stipulated to the admission of Joint Exhibits 1-12. Plaintiffs exhibits were marked 101-183 and were admitted by stipulation or order of the Court or denied admission by the Court, as described in Attachment 2 to this Order. Defendant's exhibits were marked 201-264 and were admitted by stipulation or order of the Court or denied admission, as described in Attachment 2 to this Order, beginning at page 5 of 7.
332-1, p. 1 of 15. See also p. 1 of Attachment 2.
See DE 332, p. 124 of the PDF.
B. Trustee's Witnesses
1. Timothy Shaffer.
(February 16, 2021, 1:17 pm through 2:24 pm).
The Trustee's testimony was live, but some deposition designations were also supplied.
The Trustee testified that when he was appointed in January 2012 the Debtor was not profitable but became so by July 2012. The Debtor needed to achieve gross revenue of $1.5 million per month. That was happening by the 4th quarter of 2012. Not every month but on average. Debtor filed its Plan in early 2013 calling for 100% payment to its creditors. The Plan was going to be, in the Trustee's estimation, widely supported by the Creditor's Committee and the Debtor's creditors generally.
Once Amazon terminated Debtor's use of its online platform in April 2013, Debtor never again reached gross sales of $1.5 million in any given month. Rather, Debtor's gross sales thereafter failed to consistently reach $1.0 million per month. Amazon terminating Debtor's access to Amazon's online platform devastated Debtor's working capital. Bellino and others reported to the Trustee that this termination severely spooked Debtor's trade creditors who thereafter refused to supply product to the Debtor. Amazon eventually permanently terminated Debtor's access to Amazon's online channels in October 2013. This effectively put Debtor out of business because then Debtor had no other online channel on which it sold its products. After January 2014 Debtor no longer maintained a warehouse.
In 2012, the Trustee was told by Azzarelli, Debtor's contract CFO, that some of Debtor's inventory was missing. Bellino told the Trustee that if the Debtor made inventory claims against Amazon it would hurt the Debtor. After Debtor made a demand upon Amazon for allegedly unaccounted inventory, Amazon eventually terminated Debtor's access to the Amazon online sale platform.
On cross-examination the Trustee acknowledged directing Bellino on December 15, 2012, to make a $1 million demand on Amazon for lost or unrecorded or unaccounted inventory. Instead, on December 17, 2012, Bellino made a $10.5 million claim against Amazon for the Debtor. Until his deposition, the Trustee was not aware of the magnitude of Bellino's demand or that it conflicted with the Trustee's direction.
In a May 7, 2013, email from Bellino to the Trustee in which the Trustee asked when Bellino sent a $1 million demand to Amazon, Bellino simply said "12-17-2012" without mentioning that the actual demand he sent to Amazon was for $10.5 million.
At trial Amazon focused at length on an email trail between Azzarelli and Bellino on April 23, 2013 concerning actual and projected revenue. While the Court admitted these Exhibits, the Court finds it provides no probative value to Amazon's defense of Plaintiffs claims.
On cross-examination the Trustee re-confirmed that Amazon's 15-day shutdown of Debtor's access to the Amazon online platform shattered the confidence of Debtor's creditors and damaged Debtor's cash flow and capital position. Moreover, this shut down by Amazon caused three members of Debtor's Creditor's Committee to withdraw their earlier support of Debtor's Plan. The Creditor's Committee's counsel confirmed this fact to Trustee.
Trial Transcript, February 16, 2021, at 1:22 p m.
Plaintiffs counsel did not offer a re-direct examination of the Trustee.
2. Serena Morones
(February 16, 2021, 2:37 pm to February 17, 2021, 1:03 pm).
Except for a portion of her direct examination being supplied to the Court via her Declaration, Morones' testimony was live.
Morones' direct testimony began with her 13-page Declaration but additional direct and then her cross-examination and re-direct examination were live on the first and second days of trial. Morones was called as Plaintiffs expert on Plaintiffs damages. Morones' Declaration pointed to her "Affirmative Expert Report" dated May 10, 2019, which is hereafter referred to as Morones' Damages Report.
Ex. 172, the "Morones Declaration."
The Court will first address Morones' Damages Report, then Morones' Declaration, and finally her examinations at trial.
a) Morones' Damages Report
Morones' Damages Report is a PDF document totaling 267 pages. She charged Plaintiff at the rate of $475 per hour. The Report itself is 28 pages but contains 6 schedules (including numerous sub-schedules) and 3 attachments. Attachment A is Morones' curriculum vitae. Attachment B is the list of documents she considered in preparing her Damages Report. Attachment C is a 26-page description of the basis for the Settlement Data and M15 together with 190 pages of schedules. Attachment C is a document describing how Morones came to sort the vast amount of data contained in the Settlement Data and the M15 Data and how these two heaps of data were sorted and merged to enable her to reach the conclusions made in Morones' Damages Report. By way of an analogy, if the Damages Report was instead to answer the question, "what time is it?" then Attachment C is an answer to the question, "how was that clock built?"
Morones' bills through June 14, 2019, forward at Ex. 202.
Attachment C indicates the Settlement Data includes (1) 747 text files containing 624,370,359 records where each record row describes a single type of transaction for an order, (2) 696 Excel files containing 28,426,216 records with each row describing a single type of transaction for an order and (3) two "Comma Separated Values" containing 1,804,115 records and 1,109,682 records, respectively. Some of the Excel and text files contain over 100,000 lines. Altogether, these data mountains describe 4,471,104 customer orders for 5,827,309 units which produced revenue totaling $128,224,650 of which $33,176,475 was paid to (retained by) Amazon to cover its fees and expenses. Morones calculates that the average sales price per unit was $22, of which Amazon retained $5.69 per unit. Debtor, in turn, was paid $95,048,175 or, on average, $16.31/unit.
$128,224,650 sales revenue ÷ 5,827,309 units sold.
$33,176,475 paid to or retained by Amazon ÷ 5,827,309 units sold.
$95,048,175 paid to Debtor ÷ 5,827,309 units sold. These calculations appear in Ex. 7, Attachment C, Schedule 1b at PDF page 31 of 267.
The first ten pages of Morones' Damages Report discusses the parties and their litigation. Beginning at § V, ¶ 28, she digs into her analysis. As with her Declaration, she summarizes her damages formula as
(1) Expected January 31, 2014, Ending Inventory
+ (2) Reimbursable Inventory Adjustments
+ (3) Unpaid Refund Reimbursements
+ (4) Sales Proceeds Not Remitted to Debtor
- (5) Reimbursements Paid by Amazon
= (6) Total Inventory Damage. These six calculation components are broken down in order below.
1. Expected January 31, 2014, Ending Inventory
To determine the January 31, 2014, ending inventory held by Amazon, Morones applied the inventory reconciliation formula described in Amazon's Seller Central website. That formula is as follows:
+ Customer Returns
-Removals to Debtor
Expected Ending Inventory
Looking at the M15 Data, Morones pegged the starting inventory as the total number of units placed by Debtor in Amazon's hands (6,316,429) and reduced that by the Sales (5,978,113), Removals (344,906) and Adjustments (17,524) and added to that sum Customer Returns (63,445) for a grand total of 39,331 units. This amount was multiplied by the average net price/unit ($16.31) to reflect an amount of $641,521 due to Debtor.
Customer Orders are referred in Morones' Damages Report as "Sales."
See Ex. 7, Schedule 1b.
2. Reimbursable Inventory Adjustment Transactions
Morones counts as "Reimbursable Inventory" the amount of inventory damaged, lost or stolen while in Amazon's custody. Paragraph F-4 of the Contract specifies that Amazon is liable to Debtor for such events. Morones reviewed the M15 Data and concluded 72,357 units were reimbursable to Debtor, calculated as follows:
Ex. 1 PDF page "31 of 38
See Ex. 7, Schedules 2a, 2b and 2c.
(from M15 Data)
(from M15 Data)
(from M15 Data)
72,357 x $16.31/units = $1,180,202
See the Amazon Code descriptions at Exs. 2 and 147 which are also attached hereto as Attachment 3.
Units in Morones' Lost/Found category were placed there because these units had an assigned Code of M (Misplaced) or F (Found). M and F Codes were offset against one another because M is a negative unit and F is a positive unit.
Units in the Mis-Received category have a Code X designation reflecting that a unit was incorrectly added to inventory upon receipt. This would be a Non-Reimbursable Adjustment. The M15 Data showed 28,363 such units but Morones ascertained that 4,695 of those units were not correctly designated with Code X but, rather, were items lost by Amazon. Morones came to this conclusion because 4,102 of Code X designations occurred over 30 days after the unit was received by Amazon and Amazon's 30(b)(6) witness, Jeff Moore ("Moore"), said Code X designations should be promptly made when one Amazon employee checks in a unit and soon thereafter another Amazon employee does the same. A Code X designation corrects the first person's work. Morones also found 593 units with a Code X where there was no receipt of such unit. Morones assumes such units were lost in transit therefore obligating Amazon to pay for such lost units.
3. Unpaid Refund Reimbursements
When Amazon refunded a customer's purchase price the customer was generally supposed to return the item so Debtor could re-sell the unit. If the unit was never returned to Amazon, Morones claims it should not be back charged for the refund granted to the customer. Morones found that Amazon reimbursed Debtor for some but not all the unreturned refunds, leaving 27,255 units where Debtor was charged for refunds but not credited for returned inventory.
4. Sales By Amazon, Not Returned to Debtor
See Schedule 4 to Ex. 7.
5. Amazon Payments for Reimbursable Inventory Adjustments
Based on testimony from Amazon's employee Bachand, Morones concluded the M15 Data was not reliable on the question of the amount Amazon reimbursed to Debtor. Morones, therefore, looked at Settlement Reports to quantify these Reimbursements. Morones concluded that Amazon reimbursed Debtor $304,276 for damaged or lost units. This reduces Morones' final calculated damage amount.
See Schedule 5 to Ex. 7.
6. Interest Damages
Morones added prejudgment interest at 12% per annum from the midpoint of the Amazon/Debtor relationship to arrive at a prejudgment interest damage amount of $2,906,22 through January 31, 2014.
7. Stay Violation Damages
Morones notes this Court found Amazon violated the automatic bankruptcy stay when, without Court authority, it terminated Debtor's access to Amazon's online sales platform on April 11, 2013, and did not turn access back on until 21 days later. Morones charts Debtor's sales activity for the 15 months prior to April 11, 2013 and through October 22, 2013, the date of Amazon's final shut down of Debtor's access to the online sales platform. Morones calculated average monthly sales from November 2012 through March 2013, a time period of Debtor's stabilized income performance. To that average Morones applied a .2% per month sales growth factor for the months of May 2013 through October 2013. October was then trimmed back because of the October 22, 2013, platform termination. Next, Morones looked at variable expenses from November 2012 to March 2013. Those expenses were deducted from revenues to deduce a contribution margin profit of 9.7%. Morones applied this 9.7% contribution margin profit to her projected lost revenues from May 2013 through October 22, 2013, to arrive at a lost profit damage amount of $732,000.
Ex. 7, page 22 of 267. Morones' Declaration later corrected this time frame to correctly note the shutdown was 15, not 21, days.
Variable expenses "include Cost of Sales, Outside Services - Lab or, Outside Services - Fulfillment, and Packaging Supplies."
See Ex. 7, Schedule 6a.
See Schedule 6.
All told, Morones' Damages Report reflected Debtor's inventory damages at $2,911,648, its stay violation damages at $732,000 and its prejudgment interest on inventory damages at $2,906,224 for a grand damage total of $6,549,871.
b) Morones February 5, 2021, Declaration
Morones had 30 years of accounting experience including 23 years of work analyzing commercial damages and conducting forensic accounting investigations. She is based in Portland, Oregon. She was engaged by Plaintiffs counsel to analyze whether Amazon owes Debtor for inventory entrusted to Amazon by Debtor during the course of its business relationship from January 1, 2008 to January 31, 2014. Much of Morones' Declaration rehashes the information contained in Morones' Damages Report. She bases her damage calculations on the following formula:
Ex. 7, page 14 of 267, ¶ 28.
Expected Ending Inventory on Hand January 31, 2014
+ Reimbursable Inventory Adjustment Transactions
+ Unpaid Refund Reimbursements
+ Sale Proceeds Not Remitted to Debtor
- Reimbursements Paid by Amazon
Total Inventory Damage
Ex. 7, page 15 of 267, ¶31.
1. Expected Ending Inventory:
See Ex. 7, Schedules 1 and 1a.
Inventory Units Received by Amazon from Debtor
(per M15 Data)
- Inventory Units Sold
(Per M15 Data)
- Inventory Units Returned by Customers
(per M15 Data)
(per M15 Data)
- Inventory Units Adjusted
Total Expected Inventory Units as of January 31, 2014
In analyzing "Expected Ending Inventory" Morones notes she was asked by Trustee's counsel to investigate "Removals Back to Debtor" from October 23, 2013through January 31, 2014. "Removals" occur generally when Debtor requested units be destroyed by Amazon or shipped back to Debtor. In either event, the units are Removed from Debtor's inventory Debtor Amazon's hands. Debtor might make these requests, for example, when a unit's expiration date has come and gone. Counsel also asked her to review the amount of unsellable inventory as of January 31, 2014.
The day after Amazon finally terminated its Contract with Debtor.
Reviewing the M15 Data, Morones shows that Debtor's "Removal" requests from October 1, 2013 to January 31, 2014 totaled 88,405 units but that Debtor canceled 23,995 of those requests and that, after October 22, 2013 to November 31, 2014, Amazon shipped or destroyed 62,103 units.
88,405 removals requested less 23,995 canceled removal requests = 62,103 units.
This exercise becomes interesting when Morones looked at the Seller Central Report of December 2013 which indicated that Debtor had 28,390 units in Amazon's hands of which only 3,216 units were identified as unsellable and only 2,464 of those unsellable units were "expired." Morones assumed this Seller Central Report accounted for inventory through December 31, 2013. Amazon employee Bachand testified that this report was a snapshot as of December 13, 2013. Bachand also testified that, with respect to removals, there is a "creation date and a completion date of the removal. And the units can actually have been removed at any point during that time frame."
The December 2013 Seller Central report was apparently the last report which Morones received for review.
February 18, 2021, Trial Transcript at 153:11-12.
February 18, 2021, Trial Transcript at 153:14-19.
Unfortunately, Bachand's testimony does not aid the Court in pinpointing how many units were removed after December 13, 2013 and before January 31, 2014. The Court finds the Debtor had 28,390 units in Amazon's hands as of December 13, 2013. It is not clear, nor can the Court find how many units were removed between December 13, 2013 and January 31, 2014. The M15 Report shows Debtor's total inventory in Amazon's hands was reduced in January 2014 by 12,055 units (10,850 of which were Removed).
Williams' Rebuttal Report shows Amazon Removed 19,890 units after January 31, 2014. Whether some of those removals were from a "creation date" (as Bachand would say) before January 31, 2014, is not clear. However, the Court finds the completion of those removals had to occur after January 31, 2014, because Williams' Report tells us 19,890 units were removed after January 31, 2014, i.e. the creation date and/or the completion date necessarily occurred after January 31, 2014. Units necessarily remained in Amazon's hands until the completion date. The Court finds that Amazon held at least 19,890 units of Debtor's products as of January 31, 2014, and that the completion of Removal of all those units occurred after January 31, 2014.
2. Reimbursable Adjustments:
Ex. 7, Tables 1 and 2a, 2b and 2c.
(per M15 Data)
(per M15 Data)
(per M15 Data)
Total Reimbursable Inventory Adjustments
As reflected in Amazon inventory codes D, E, 5, 6 and 7.
As reflected in Amazon's inventory codes M and F.
As reflected in Amazon's inventory code X.
This damage component is more fully discussed above where the Court reviews the Morones' Damages Report.
$195,721 ÷ $16.31/unit = 12,000 units.
The Seller Central Report is sometimes referenced as the Settlement Data.
This damage component is more fully discussed above where the Court reviews Morones' Damages Report.
$1,156,495 ÷ $16.31/unit = 70,907 units.
(Per M15 Data and Seller Central Report)
Morones indicates this figure represents the difference between sale orders identified in the M15 Data as compared to the lesser amounts referenced in the Settlement Data. She assumes Amazon had not reported these sales to Debtor for these increased sales. Morones does note, however, that after delivering her Damages Report, Amazon accounted for the six 48-hour periods where the Settlement Data had not been produced, which time period gaps largely accounted for this $1,756,495 differential. Armed with this new Settlement Data to fill in the Data Gap, Morones concedes her damage amount for the line item "Sales Proceeds Not Remitted" from Amazon to Debtor should be reduced to $172,851 from the $1,156,495 reflected in her Damages Report, so long as Amazon "satisfactorily authenticated the additional data." This Court finds Amazon "satisfactorily authenticated the additional data" at trial. At best, Debtor's damages for the "Sales Proceeds Not Remitted" total $172,851.
This is referred to herein and in Williams' Rebuttal Report as the "Data Gap."
This equates to 10,598 units ($172,851 ÷ $16.31/unit).
$304,276 ÷ $16.31/unit = 18,656 units.
This damages component is more fully discussed above where the Court reviews Morones' Damages Report.
$2,869,663 ÷ $16.31/unit = 175,945 units.
This amount is down from the inventory damage of $2,911,648 reflected in Morones' Damages Report.
To her inventory damage amount Morones adds prejudgment interest of $3,460,578 through January 31, 2021. This prejudgment interest calculation is made at 12% per annum at the "mid-term convention" between January 1, 2008 and January 31, 2021. From February 1, 2021, Morones would add a per diem rate of $943 to the Plaintiffs inventory damages.
Or $617 per day if Amazon "satisfactorily authenticated the" Gap Data.
Next, Morones analyzes what would happen if her net price of $16.31/unit attributable generally to Debtor's products entrusted to Amazon instead also disallowed Amazon's presumed fees and costs of $5.69/unit. If Debtor were to be paid the full $22 average sales price on "Expected Ending Inventory" Debtor's damage number would increase by $223,999. The Warehouse Adjustment damage number would increase by $161,647. The Lost/Found Adjustment would increase by $223,350. The Mis-Received Adjustments would increase by $26,715.
Morones "assumed a net cost per unit of $22.00 reduced by $5.69 per unit to account for Amazon's fees and costs." This Court presumes that by "cost per unit" Morones means $22 is the price per unit paid by Debtor's customers and that the $5.69 per unit reflects the amount Amazon would be entitled to retain from the $22 per unit sales price, all pursuant to the parties' Contract.
39,331 units x $5.69/unit.
Morones say the increase would be $161,740 but 28,409 units x $5.69/unit = $161,647.
Morones says the increase would be $223,478 but 39,253 units x $5.69/unit = $223,350.
Morones says the increase would be $26,730 but 4,695 units x $5.69/unit = $26,715.
Morones' Declaration next turns her attention to perhaps the most hotly contested element of the Plaintiffs case - Code Q Adjustments. Morones notes the Trustee is claiming inventory damages for all inventory listed with a Code Q, where that inventory is also listed as sellable. While Morones notes the M15 Data shows 166,279 units as bearing a Code Q, only 147,968 units were reflected as sellable. At $16.31/unit damage to Debtor, Morones calculates the Code Q damages at $2,413,470.
Also referenced as "Damaged-Miscellaneous."
147 968 units at $16.31/unit is $2,413,358 not $2,413,470 as Morones indicates in her Declaration.
Morones' Declaration takes umbrage with Amazon's damage expert's Rebuttal Report. Morones criticizes the September 14, 2020, report of Williams for allegedly (1) failing to factor in downward inventory adjustments over the duration of the parties' relationship, (2) not acknowledging Unpaid Refund Reimbursements, (3) ignoring unpaid sales, and (4) counting inventory transactions after January 3, 2014.
Morones contends Williams deducts reimbursements from ending inventory rather than first adding adjustment transactions. This, she claims, results in a double deduction of liability transactions. Moreover, since a large number of Amazon's unit reimbursements show no money paid for those units, she quantifies reimbursements in dollars, not units, in contrast to what Williams does. Morones notes that, although Amazon says it removed 19,980 units after January 31, 2014, the M15 Data shows that, as of March 31, 2015 (the last month of data reflected in the M15 Data), Amazon still held 20,405 units.
Ex. 172, ¶ 28(b).
Finally, Morones' Declaration criticizes Williams on his stay violation damages calculation, goes on to affirm her pre-judgment interest calculations and then notes her corrected inventory per diem rate.
$617/day per Morones' Declaration, down from $1,057 referred in Morones' Damages Report, Schedule 1, Ex. 7.
Ex. 1 72 at page 6
Ex. 172 at page 4.
c) Morones' Direct, Cross and Re-Direct Examinations
At the outset of Day 2 of the Trial, the Court addressed the admissibility of Exhibits 247-252 and the Declaration of Amazon's custodian of records, Ershad Junaid. Through hearing the arguments of counsel and comments by Morones, the Court learned that Exhibits 247 - 252 are six authenticated Amazon business records, each containing a vast amount of data covering approximately 87% of the payment data which was missing from the payment data Morones reviewed in preparing her Damages Report. Amazon's expert Williams refers to this payment Data Gap in his Rebuttal Report. Exhibit 253 contains highlighted records pertaining to the payment Data Gap. Junaid's Declaration is Exhibit 262. The Court overruled Trustee's objection to admitting these exhibits into evidence and granted Amazon's request to admit Exhibits 247 - 253 plus Junaid's Declaration (Ex. 262). Since these exhibits were authenticated as Amazon business records which fill 87% of the Data Gaps, Morones conceded that the damage component described as "Sales Proceeds Not Remitted to Debtor" is $172,851 not the $1,156,495 figure referenced in Morones' Damages Report.
The six documents collectively contain 15,845 pages of coding related to the Data Gaps of payment data missing from the Settlement Data reviewed by Morones.
See Ex. 172, Morones' Declaration, ¶ 11, page 6, segment 4.
Morones noted in her live supplemental direct examination that the Seller Central Data reflected less inventory on hand than the M15 Data. She noted the Debtor could only seek Removal from Amazon's inventory if it knew from the Seller Central Data that such inventory units were in Amazon's possession. Her Exhibit 178 showed Seller Central Data indicated Amazon was holding 156,919 of Debtor's inventory as of August 13, 2013, yet the M15 Data showed that between September 1, 2013 and January 31, 2014, there were requests to remove 164,629 units. Exhibit 179 showed M15 Data ending inventory figures at month end (August through December 2013) consistently showed more inventory than was reflected in the Seller Central Data. Exhibit 180, a document prepared by Morones, shows 3,555 more units ultimately Removed by Amazon than Seller Central indicated were held by Amazon on December 31, 2013.
As noted above in the section discussing Morones' Declaration, testimony from Amazon's Bachand indicates why Exhibit 180 is flawed.
In Amazon's cross-examination of Morones she acknowledged that, before working on this engagement, she had never worked on issues concerning Amazon's inventory system. Morones acknowledged that a considerable amount of information that she considered in preparing Morones' Damages Report was not listed in Attachment B of that Report. She also confirmed she failed to timely reveal to the Oregon Board of Accountancy a lawsuit filed against her. For this omission she was disciplined and required to pay a $250 fine. Morones also used a program created by Cone to help make sense of the Seller Central Data made available to Cone. That program was never produced by Plaintiff to Defendant. She also acknowledged the Seller Central Data does not necessarily reflect month ending inventory figures but the M15 Data does. 39,331 units are reflected in her Report as Ending Inventory for January 31, 2014. Morones confirmed Sales Proceeds Not Remitted is $172,851 not the $1,156,495 units reflected in her Damages Report. This difference is largely attributable to the Data Gaps in the Seller Central Data. Morones also acknowledged her original Damages Report failed to include an analysis of Code Q designated units. While Morones' Declaration remedied this omission, she indicates that, of the 166,279 units designated as Code Q in the M15 Data, 147,968 of those units are identified as sellable and should be included in Plaintiffs damage amount.
See Ex. 5, Fn. 8.
See Ex. 172, page 4.
See Exs. 247 - 252.
Morones admitted not factoring Code N or O transactions into her Damages Report. Morones confirmed the M15 Data does not contain data concerning payments by Amazon to Debtor. Such payment information is contained only in the Settlement Data subset of the Seller Central Data.
Cross-examination concerning Morones' stay violation damages revealed that, while her Damages Report twice indicated Amazon shut down Debtor's access to the Amazon platform for 21 days, the correct shut down time frame was 15 days. Cross-examination also delved into Debtor's sales activity prior to November 2012, revealing to this Court that Debtor's sales were stabilized, not as of November 2012, but instead back to July 2012. Cross-examination suggested some of Debtor's May to October 2013 sales may have been forbidden to be sold on Amazon's platform and that some of its products were priced far below normal prices.
Ex. 7, ¶¶ 21 and 56.
In her re-direct examination, Morones indicated that, after reviewing information from Williams' Rebuttal Report which filled in a big portion of the Data Gaps, she revised her Amazon Reimbursements figure to ($305,611), from Morones' Damages Report number ($304,276). This reduced her damages calculation by $1,335.
Morones discussed Schedule 5 to Morones' Damages Report explaining that the Seller Central Data concerning Reimbursements paid by Amazon over the course of the parties' relationship show Miscellaneous Adjustments and Balance Adjustments of $304,276 while the M15 Data for the time frame show such adjustments as totaling $347,323. She said she would expect such M15 Data to be lower than the Seller Central Data because the Seller Central Data is complete, but the M15 Data is not. She was suggesting the M15 Data was incorrect in this regard.
Ex. 7, page 39 of 267.
Morones reviewed Exhibit 9, a March 1, 2019, email from Trustee's counsel to Amazon's lawyer enclosing draft language to be included in the Disclosure Statement to the Trustee's proposed chapter 11 Plan. The Disclosure language at page 8 of 16 mentions Amazon shut down Debtor's access to its online platform on April 11, 2013, for 21 days. This was apparently picked up by Morones in her Damages Report but is incorrect as the Debtor was shut out for 15 days, not 21.
Ex. 7, page 12, ¶ 23 and page 22, ¶ 56.
Morones next testified concerning Williams' Rebuttal Report where he says Code E inventory is removed from inventory as compared to Williams' Declarationwhere he said Code E inventory is not removed from inventory but, instead, destroyed (Code D) or moved out of inventory (Code O) and Reimbursed to Debtor. She understands Removals to be units returned to Debtor or destroyed. Many Removable units are unsellable. She understands Code E is a change from sellable to unsellable which item would later likely be Removed (destroyed or returned to Debtor). This change, however, does not reduce the overall number of inventory units. Morones had not observed an Amazon policy consistent with Williams' revised opinion nor did she observe data which would bear out his revised opinion in Code E treatment.
Ex. 5, ¶ 40.
Ex. 259, ¶ 32, Williams' Declaration dated February 4, 2021.
Morones was asked about Amazon's demonstrative exhibit which suggests the maximum number of unaccounted units is 36,972. She disagreed with this. Among other things she indicates "Removals" should total 10,284 more units so the unaccounted units would be 47,256 units, not 36,972 units.
Morones noted that inventory units can be destroyed via Code D or Removals. Williams, she says, indicates Code Q units are Removed not destroyed under a Code D as suggested by four of the Defendant's demonstrative exhibit(s).
Morones discussed the list of Amazon's Fulfillment Reports, which also provides an inventory formula from which Morones calculated Plaintiffs damages. One of the Fulfillment Reports is described as Inventory Reports. One of the Inventory Reports is a report called Inventory Adjustments. This is where Amazon's inventory codes are described and defined. These codes are important as there is much controversy regarding application of appropriate codes and what happens when a unit is coded one way or the other. The Inventory Adjustment Report is broken into the following code categories: (1) the adjustment reason codes, (2) the misplaced (M) and found (F) inventory codes, (3) the damaged inventory codes, (4) the unrecoverable inventory codes, (5) the inbound shipment received adjustments, (6) the software connections, (7) the transferring ownership codes and (8) the catalogue management codes.
Ex. 147. See also Ex. 2. These are both attached to this Order collectively as Attachment 3.
Morones discussed Codes F, M, D, E, K, U, Q. She never saw an explanation that Code Q (damaged-miscellaneous) or Code E (damaged at Amazon Fulfillment Center) units would eventually be given a Code D (destroyed). She noted that 168,526 units were given Code Q.
This number is inconsistent with her Declaration which at ¶ 26, says these were 166,279 Code Q units referenced in the M15 Report.
Morones turned to her Declaration at ¶ 18 where she shows that, as of December 31, 2013, Seller Central Data showed 28,390 units of which 3,216 were unsellable, most of which were expired not damaged.
Morones read Amazon's discovery responses where it twice states that the M15 Data "contains all of the relevant data necessary to evaluate and calculate the Debtor's inventory over time in the Fulfillment by Amazon Program."
Ex. 111, page 18, line 14 and Ex 116, page 11 at question 10, a October 30, 2018 letter from Amazon's counsel to Plaintiff's counsel.
According to Morones, the Settlement Data, not the M15 Data, provided the best data on Reimbursements. She saw discrepancies in data produced in M15 Data as compared to the Settlement Data. Amazon did not produce to Plaintiff a full set of Settlement Data and the M15 Data does not provide all data concerning all the parties' transactions, especially financial transactions.
Finally, Morones discussed units destroyed after January 2014. Her one-page Table shows starting inventory per the M15 Data as 39,331 units, that in September 2014 4,741 units were destroyed and as of March 2015 Amazon continued to hold 20,405 units of Debtor's inventory.
According to Morones, before taking into consideration Amazon's Reimbursements to the Debtor or the 87% Data Gaps, which has now been filled by Amazon's information, the Debtor's inventory damages total $2,869,663. Morones applied the following formula to determine this total:
Morones Declaration, page 4.
Expected Ending Inventory
Unpaid Refund Reimbursements
Sale Proceeds not Remitted
Total Inventory Damage
Per Morones, after subtracting payments made by Amazon to the Debtor for Reimbursements and after subtracting the Data Gaps, which have now largely been filled, the Debtor's net inventory damages total $1,875,208, before application of prejudgment interest. Morones applied the following formula to determine this total:
Morones Declaration, page 6.
Expected Ending Inventory
Unpaid Refund Reimbursements
Sale Proceeds not Remitted
Net Inventory Damage
Attachment 4 is a list of all Plaintiff's deposition designations from the depositions of Azzarelli, Bachand, Ice, Moore, Lawcock, Bellino, Ashworth, Soder, Cone, Reilly and Shaffer.
3. Thomas Azzarelli
(February 17, 2021, 1:50 pm to 1:58 pm)
Azzarelli's testimony was given by videotape from his April 18, 2019, deposition played at trial plus additional designations. Although Exhibits were discussed in the Azzarelli deposition, those Exhibits were not attached to the Court's copy of the written deposition transcript.
Those exhibits were, however, included in the List of Trial Exhibits. See DE 373.
Azzarelli was at some point an independent contractor hired to be the chief financial officer of Debtor. He was deposed on April 18, 2018. Azzarelli each month reviewed the Debtor's inventory numbers and came to understand that Debtor's inventory in the hands of Amazon was shrinking faster than sales of its units. Lost inventory and damaged inventory were discussed internally with employees of the Debtor. He did not know how much inventory was lost by Amazon, but it was "a lot."
Plaintiff introduced Azzarelli's direct testimony by designating portions of his deposition conducted on April 18, 2019. Prior to filing bankruptcy, Debtor retained Azzarelli as a part-time consulting Chief Financial Officer ("Consulting CFO") to the Debtor, and he stayed in that capacity through Debtor's bankruptcy. His post-petition role as Consulting CFO was to provide consultation to Debtor's management and to ensure the availability of experienced accounting and record maintenance during the pendency of the Chapter 11 proceedings.
In his deposition, Azzarelli testified that he was informed by Bellino of Amazon losing Debtor's inventory and failing to properly reimburse Debtor. Azzarelli, however, never received a report regarding the allegedly lost inventory or reimbursements. Azzarelli spoke to Bellino's character, noting Bellino had the desire and ability to grow Debtor into a successful company. Azzarelli also spoke to Shacklock's character, stating she was very good at her position. Shacklock was the accounting manager for Debtor and she, along with Lawcock, Peeples, and Reilly, reviewed Debtor's inventory each month. After determining what the inventory was, Shacklock adjusted the Debtor's financial statements according to the value she determined it to be for financial reporting. According to Azzarelli, Shacklock stated Debtor's inventory seemed to be shrinking in value, as reflected in the financial statements, which meant the inventory was going down faster than she imagined it would be based on revenues generated.
The Court found Azzarelli's testimony credible but not particularly revealing because he knew little about Debtor's relationship with Amazon or Debtor's inventory management.
4. Justin Ice
(1:59 pm to 2:34 pm on February 17, 2021, and then from 9:10 am to 9:53 am on February 18, 2021.)
Ice's testimony was via videotape of his July 18, 2017, deposition clips played at trial plus additional designations from his 171-page deposition.
Ice was an Investigation Specialist for Amazon. He worked specifically as a reimbursement specialist with the FBA Program. Ice testified about Amazon's inventory processes and accountings - specifically, the Deep Dive and the M15 Data. The Deep Dive is what Amazon refers to as a thorough review of the Debtor's top selling products. Ice indicated that Debtor's inventory was stored in Amazon's fulfillment center or warehouse in totes, bins, and shelves. He pointed to Ex. 154 as a graphic representation of how one can determine the amount of its inventory in Amazon's possession. Ice discussed the fact that Amazon needed to re-write programs so its computer data would provide better data to Amazon and its customers. Ice discussed the M15 Data but noted it does not attempt to reconcile Debtor's inventory at Amazon. Among other things, the M15 Data does not contain all data concerning the parties' sales transactions. Ice discussed ASIN flips or mergers, which happen when units with a certain ASIN # contain more than one SKU code. An ASIN # links product referenced on amazon.com to products in inventory in Amazon's hands.
See also the last page of Ex. 148. This exhibit is attached hereto as Attachment 5.
Ice also discussed Amazon's reimbursement process and the beginnings of the FBA Program. Ice detailed how and when Amazon would process reimbursements and explained different scenarios, which would trigger a reimbursement, like prohibited or unsellable products. With respect to the FBA Program, Ice acknowledged the FBA Program was started on the fly and improvements were necessary as the FBA Program grew because of gaps in information provided to sellers.
Ice further testified about different adjustment codes used to help determine a seller's inventory. Specifically, Code D is an adjustment code for destroyed products, which is used when a product is so badly damaged it cannot be shipped back to the seller. Code E stands for a change in the disposition of an item and would be used if a product was damaged while waiting to be sold at the Amazon warehouse. Code Q is a miscellaneous damage adjustment code used when Amazon is unable to determine the cause or source of the damage. Code M is used when a product is missing from its designated location. Often missing product is placed in a holding account. Generally, within 30 days, Amazon would reimburse sellers for any missing products. A Reimbursement could take the form of payment to Debtor or new units credited to Debtor's inventory at Amazon. Amazon would sometimes credit a seller's account if there was inventory with the same FNSKU as the damaged product in the holding account. Transferring a product out of the holding account to a seller's account is referred to as a Code N adjustment.
However, Ice testified that he did not know the criteria Amazon used to determine whether to apply Code Q to a given unit. Ice Deposition at 133:14-15. Code Q is discussed at length in § VII(c)(4) below.
Ice further discussed the Deep Dive and the reconciliation process he went through to try to determine why there were discrepancies between Debtor's data and Amazon's. Some information Ice looked at was not available because problems with ASIN flips, out of sync databases, and because the age of the information. Ice went through the M15 Data to validate the information and ensure the data sets were marginally close to the information he had pulled together during the reconciliation process. Ice acknowledged during his deposition that the M15 Data did not include all the information available because of the limited timeline and Debtor's long history. Ice also testified that he recalled seeing reimbursements to Debtor for zero dollars but was unaware of the reason behind this. However, Ice did report to his manager and an Amazon attorney that the reimbursement information included in the M15 Data was underinclusive and was incomplete with respect to transactions prior to 2012.
Finally, Ice testified about Amazon's inventory tracking process and acknowledged that it would be nearly impossible to track an individual product from the time Amazon received it until the time it was sold. To track a product from inception to sale, a seller would have to perform additional research and connect where inventory adjustments occurred to certain time stamps.
The Trustee analogizes this inventory accounting to one where a farmer deposits number of bushels of corn in a cooperative's silo and later can sell that many bushels out of the silo. It will not be the same kernels of corn deposited and sold by the same farmer, but the cooperative should track the volume of deposits and removed corn. See DE 396, p. 62.
5. Tasha Bachand.
Bachand's testimony in Plaintiffs case was live at trial plus via clips from her March 26, 2019, videotaped deposition plus additional deposition transcript designations. Bachand was designated by Amazon to testify as its Fed.R.Civ.P. 30(b)(6) witness.
Bachand started working for Amazon in 2015, about two years after this Adversary Proceeding commenced. Bachand is an operations manager at Amazon. She was not called as an expert witness in this case although she knows more about the M15 Data than anyone at Amazon. Bachand first became involved in the Amazon/Trustee dispute in August 2016. She has knowledge of Debtor's inventory issues since 2011 but not before then. The M15 Data is Amazon's best accounting of Debtor's inventory, but it does not contain payment information. Payment information would be in the Seller Central Data, specifically in the Settlement Data.
Bachand described Seller Central as an internal web portal where sellers can log in to view information about their account and see the disposition of their inventory which was delivered to Amazon. Seller Central also provided a formula for Debtor, and other sellers, to manage and reconcile its inventory. The formula was: "starting inventory balance
+ received inventory
-customer orders + returns
= ending inventory balance."
Debtor could have independently tracked its starting inventory balance, received inventory, and removals but as to the remaining components, Debtor would have to rely on Amazon to provide such information through Seller Central Data.
Bachand explained the difference between physical and virtual inventory. Physical inventory is the inventory that is on the truck, in the tote, or in the bin. Virtual inventory is what is shown through Seller Central and various Amazon reports. Even though Amazon did regular checks on physical inventory to ensure it matched the virtual inventory records, Bachand stated it was possible for the virtual inventory shown on Seller Central to not always match the physical inventory at Amazon. She confirmed that, with respect to Debtor, there were times this happened.
Bachand explained adjustment codes are a virtual way of tracking the physical inventory. Employees in Amazon's fulfillment centers decide which adjustment codes are used. The purpose of some of the adjustment codes are to make physical inventory and virtual inventory match. For example, software corrections are used to correct inventory discrepancies. She testified that software corrections reflected as Codes 1, 2, and J are not considered in calculating the ending inventory balance.
Bachand explained Code X and Code Q and confirmed that there were 71,900 Code Q adjustments to Debtor's inventory in July 2013. In reviewing the reimbursement chart from the M15 Data, Bachand testified that there is no information showing whether Amazon paid Debtor in product or dollars for inventory reimbursements that show up as zero.
6. Jeff Moore.
(Video deposition shown at trial on February 18, 2021, beginning at 10:09 am and ending at 10:47 am plus deposition designations.)
Amazon designated portions of Moore's December 5, 2017, six-hour videotaped deposition testimony in response to designations offered by the Trustee. Moore's direct examination was introduced through designations from his Fed.R.Civ.P. 30(b)(6) deposition on December 5, 2017.
Moore is an Amazon employee who testified about Amazon's inventory processes. He noted Debtor would not generally be given access to any of Amazon's fulfillment centers. He discussed the M15 Data and indicated he worked on that with Ice until Ice left Amazon. Since then, he has worked on M15 Data issues with Bachand.
Moore testified concerning the FBA Program's inventory check-in process at Amazon's distribution and fulfillment centers. Moore explained an associate scans inventory as it arrives and determines the seller and units included. During this process, an associate counts the units in each shipment and inspects the unit for damage. If a unit is damaged, the associate is to answer programmed questions in the internal software system. Based on the answers, a specific adjustment code is assigned. After the shipment is processed and checked for damage, the unit is placed in a tote (i.e., a plastic bin) and sent to shelving where a second associate will store the unit on the unit's designated shelf. Moore explained if a seller wants to inspect their inventory at an Amazon center, an associate will generally conduct an audit of all the seller's inventory upon that request.
Moore further testified about the inventory reconciliation formula, which represents a seller's ending inventory balance in Amazon's fulfillment centers. Moore explained in detail the removal component of the formula, which is initiated by a seller's unit return request. Any units not removed by a seller or sold to a customer would be returned to the seller when the seller stops selling through the FBA program.
Moore also testified about different adjustment codes, which informs sellers through Seller Central about the status of their inventory at Amazon. Moore described Code X as a minus adjustment to inventory, meaning the number of units received were less than expected. Code X is assigned contemporaneously with the unit being discovered as missing. Moore also described Code M as the code assigned to missing units. If Code M is assigned to a unit, Moore stated Amazon would reimburse the seller within 30 days. Moore testified further on Amazon's process of identifying commingled products and how missing and/or damaged units are handled. If a product is damaged and Amazon is responsible, the seller receives the sale price less the fees. If a customer returns a damaged product to the fulfillment center, an associate will conduct an analysis of the item through a programmed questionnaire in the internal software system. This will tell the associate whether Amazon or the seller is responsible for the damage.
7. Eric Soder.
(Soder's testimony was only by designated portions of his July 27, 2017, videotaped deposition plus designations from his 63-page deposition transcript.)
Soder is a former Amazon employee who testified about Amazon's inventory processes and the Debtor's inventory management. Soder began his career with Amazon in 2008 as an account manager. In this role, Soder assisted sellers outside the FBA Program with Amazon's website platform by helping navigate the website platform so sellers could list their products.
Soder discussed ASIN which are designed to represent a product on Amazon's website but not to track seller's inventory. Soder further discussed Amazon's ownership determination software, which helps determine where a missing unit belongs in Amazon's warehouses. Soder explained the process of identifying missing and misplaced items in Amazon's fulfillment centers and how missing units are assigned a missing adjustment code and reidentified products are assigned a Code F. Even if a missing product is found, a damaged adjustment code will be reported in Seller Central. Soder further testified about Amazon's cycle-count time-period. After a certain time, referred to as a "cycle-count," Amazon employees will rescan every barcode of every product in Amazon's warehouse. The barcode will tell Amazon's internal system what the unit is, but not who owns that unit.
8. Jeffrey Cone.
(Cone's direct testimony was by designated portions of his August 17, 2020, videotaped deposition plus designations from his 140-page deposition transcript.)
The substance of Cone's testimony was limited to foundation for the analysis underlying Morones' Damages Report and to responding as a rebuttal witness.
Cone was employed as a consulting expert witness by the Trustee. He testified by deposition as to his role in helping Morones understand the M15 Data and Seller Central Data used in Morones' Damages Report.
Specifically, Cone was tasked with figuring out what the M15 Data and the Seller Central Data represented and comparing his analysis to Ashworth's, identifying points where they reached different conclusions. To do this, Cone reverse engineered Ashworth's conclusions back to the original data source. Although Cone testified they had different numbers in some instances and could not recall the specifics, he did recall they had the same number for Receipts.
Cone further analyzed inventory status reports and inventory activity reports from Seller Central to understand general background information regarding Debtor's inventory levels and activity. These reports were not the primary source for Morones' calculation but supported Morones and her team's position regarding adjustments. Cone testified that he confirmed the M15 Data and Seller Central Data had not been modified.
Lastly, during his deposition, Cone acknowledged gaps in their analyses were caused by missing Settlement Reports. He stated that, even if those gaps were filled, their conclusion would have remained the same. The calculation for unremitted sales to Debtor in Morones' Damages Report did not consider the gaps because the calculation was based on the M15 Data, not the Settlement Reports. Furthermore, the gaps were only for the time periods of January 2010 and March 2010.
9. Stephen Ashworth.
(Ashworth's direct testimony was by designated portions of his August 19, 2020, videotaped deposition plus designations from his 91-page deposition transcript.)
Ashworth's testimony was limited to foundation for exhibits and to respond as a rebuttal witness. Ashworth was employed by the Trustee as a consulting expert witness and testified as to the documents and data he collected from the Debtor as well as his coordination with Morones and Mr. Cone in the development of Morones's Damages Report.
Ashworth accessed Seller Central starting in October 2013 up until the production of the M15 Data and downloaded all available relevant data related to Debtor's transactions and inventory. Ashworth testified about the specific information he looked for in Seller Central including inventory event detail reports, monthly ending inventory reports, individual transactions, and Settlement Reports.
Ashworth explained the difference between various types of Amazon's reports and discussed how Settlement Reports are exclusive to sales in that they show money exchanged between Amazon and Debtor. Inventory reports include things such as receipts and adjustments and show the inner movement of inventory in Amazon's warehouses. Ashworth further testified about the process for retrieving Settlement Reports from different time periods. Ashworth first began by downloading Settlement Reports between January 2012 and February 2014. Next, to obtain Settlement Reports prior to January 2012, Lawcock and Peeples downloaded the Settlement Reports and produced them to Ashworth. Ashworth noticed gaps only in Settlement Reports Lawcock and Peeples produced.
Lastly, Ashworth testified about the information and various reports he sent to Cone and Morones. Ashworth noted that during the 2014 mediation, Amazon discussed certain sales transactions not reported in Seller Central Data and that they were uncertain whether the Debtor was paid for those transactions. This, in Ashworth's opinion, is what led to the production of the M15 Data.
10. Sean Lawcock.
(Lawcock's direct testimony was by designated portions of his April 23, 2019, videotaped deposition plus designations of his 158-page deposition transcript.)
Lawcock's testimony was limited to foundation for exhibits and to respond as a rebuttal witness. Lawcock was Debtor's former inventory manager. He testified as to the Debtor's inventory management practices, the Debtor's internal inventory tracking difficulties, and the December 2012 inventory audit that serves as the basis for Amazon's fraud claims.
Lawcock began his career with Debtor as a purchaser on the merchant side and switched to a software specialist/inventory manager around 2011-2012.
Prior to Debtor's bankruptcy, while Lawcock was a purchaser, he created an internal tracking system for his own personal inventory. This system was built specifically for the Merchant Program, not the FBA Program. Post-bankruptcy, Lawcock started tracking the Debtor's Settlement Reports so he would know when Amazon reversed a sale. Amazon would reverse sales if a product was unsuccessfully shipped to the customer or returned. The Settlement Reports had a column listed as "Other," which included a lump sum dollar amount for various items such as reimbursements and fees, making it difficult to independently track reimbursements. Eventually the Settlement Reports put reimbursements into its own category. Lawcock testified he had a poor experience in getting reimbursements from Amazon because he was consistently told by Amazon it was against their policy to refund for certain products, but he never received a reference to the actual policy stating this. He also testified that there were hundreds of instances where Amazon would state products were shipped to the Debtor's warehouse even though he had the packing slips, and the products were not there.
Lawcock discussed the Removal process implemented at Debtor and specifically the automated Removal program which Amazon offered to sellers. Lawcock noted the Debtor switched to a manual Removal program from an automated Removal program because, under the automated Removal program, products were randomly sent to Debtor and packaging slips were included less than 25% of the time. Without packaging slips, Debtor was unsure where the original product came from and why it was coming back. The automated Removal program also did not include reporting, which resulted in Debtor having to pull the return report from Seller Central and review each product to see if there was an item related to the order. If Debtor was unable to find a related item, Debtor would contact Amazon seller support group and request a reason for the returned product, a Removal order, and the original SKU. Under the manual Removal program, Debtor was able to implement a process and track which units were coming back to Debtor and why.
In December 2012, Lawcock conducted an historical audit of Debtor's entire inventory based on inventory data available in Seller Central for the years 2008 to 2011 to determine the amount of inventory which should have then existed. As part of the auditing process, Lawcock identified discrepancies and then worked with seller support to resolve the discrepancies by either correcting the inventory count or receiving a reimbursement. At the end of the audit, Lawcock concluded there were 151,871 missing units from 2010 through 2012 and submitted this to Amazon for reimbursement at Bellino's direction. In his deposition, Lawcock explained the 151,871 missing units were "gross missing units" that excluded the positive overages from Amazon. Lawcock believed Amazon misreported 151,871 units of Debtor's inventory and these units were unaccounted for in Debtor's inventory, regardless of the overages being taken into account. When asked how 151,871 units could have gone missing from 2010 through 2012 without the Debtor's knowledge, Lawcock explained it was because the auditing process had not begun until 2012.
11. Daniel A. Bellino
(Bellino's direct testimony was by designated portions of his April 22, 2019, videotaped deposition plus designations from his 265-page deposition transcript.)
Bellino's testimony was limited to foundation for exhibits and to respond as a rebuttal witness. Bellino was the Debtor's former CEO. He testified as to the Debtor's business practices; the Debtor's participation in the FBA Program; the Debtor's lack of inventory tracking, management, and organization; and the Debtor's reimbursement claims.
Bellino was the founder and Chief Executive Officer of Debtor. The designated portions of his depositions include situations in which Bellino could not answer a question based on lack of information or inability to recall the exhibit or fact. This testimony was not particularly revealing.
C. Motion for Directed Verdict
At the conclusion of Plaintiff s case on trial day 3 at 11:35 am, Amazon moved for a directed verdict arguing that since the M15 Data was not admitted into evidence in Plaintiffs case, there was insufficient data to support Plaintiffs damage claims on Counts 1 and 3 of the Complaint. The Court denied Amazon's motion. Exhibits 131 and 161were admitted into evidence.
Seller Central Data.
D. Amazon's Witnesses
1. Tasha Bachand
(February 18, 2021, at 10:45 am to 11:27 am (Video) and February 18, 2021 at 1:05 pm to February 19, 2021 at 1:40 pm (Live))
Plaintiff introduced portions of Bachand's videotaped deposition in his case in chief. Defendant called Bachand as a live witness in its case and designated portions of her March 26, 2019, 76-page deposition. Bachand discussed Amazon's FBA Programing inventory processes, inventory codes used by Amazon, the Deep Dive, the Seller Central Data and the M15 Data.
Attachment 6 is a list of all the deposition designations by Amazon from the depositions of Azzarelli, Bachand, Ice, Lawcock, Moore, Bellino, Ashworth, Soder, and Cone.
Bachand began working for Amazon in 2015, well after Amazon terminated Debtor's access to amazon.com and well after Plaintiff commenced this Adversary Proceeding. Bachand has been an Amazon manager for two years. She first became involved in Debtor's inventory issues in August 2016, over three years after this Adversary Proceeding was commenced.
Bachand described an ASIN Merge as a consolidation of two or more ASIN's into one ASIN. She talked about having done a deep dive on some products. She indicated the M15 Data cannot be used to draw some of the conclusions Morones drew about Adjustments, Reimbursements and amounts unpaid by Amazon to the Debtor. Some Adjustments duplicate the coding of certain units. M15 Data does, however, show the total quantity of units "Adjusted."
In discussing the nature of Code Q designations for a given unit, Bachand confirmed that every single Code Q unit is accounted for somewhere else as a Code D (destroyed), 5 (unrecoverable), M (misplaced) or Removed (destroyed or shipped back to Debtor per Debtor's request). For these reasons, Amazon contends Code Q does not signify Amazon liability to Debtor.
DE 364, February 18, 2021, trial transcript at pg. 112.
See n.17 above. The Court will not factor these 2018 revisions in this Order.
Bachand discussed Ex. 245 and how when the Debtor on October 17, 2013, turned off the auto-Removal setting for its inventory account with Amazon. Any Removal requests would need to be made manually by Debtor thereafter. If a unit is useless for more than 90 days and Debtor has not asked that it be removed, Amazon would thereafter destroy that unit.
Bachand indicated that, while M15 Data stops at month end, Settlement Data does not. This is one more reason these two data sets are not an exact match.
Bachand agrees with the following data in the Morones' Declaration:
1. Expected Ending Inventory
Inventory as of January 31, 2014
What Bachand does not agree with from Morones' Declaration is her numbers concerning:
(a): 2. Reimburseable Adjustments: 72,357 Units
Bachand noted there is some overlap of an unknown quantity between this line item from Morones and Adjustments or Removals referenced in the M15 Data or Seller Central Data.
(b): 3. Unpaid Refund Reimbursements $186,247
This information cannot be accurately surmised from the M15 Data, and
§ VI(B)(2)(c) above.
2. E. Weiant Williams
(February 19, 2021, from 1:43 pm to 4:00 pm)
E. Weiant Williams' ("Williams") direct testimony at trial was via his ten-page Declaration ("Williams Declaration"). Williams' Rebuttal Report dated September 14, 2020, is attached to his Declaration as Exhibit A. The Court will first address Williams' Rebuttal Report then his Declaration and, finally, his trial testimony.
Ex. 5. The "Rebuttal Report."
a) Williams' Rebuttal Report of September 14, 2020
Williams is an Illinois certified public accountant and managing director of The Claro Group. His career has largely been focused on forensic accounting and accounting consultancy, especially in claims and disputes. His curriculum vitae is attached to this 36-page Rebuttal Report as Exhibit 1. He charged $375 per hour for this engagement. A list of the information reviewed and considered in preparation of the Rebuttal Report is attached to that Report as Exhibit 2.
Williams agrees with Morones in several respects. He agrees the data shows Amazon received 6,316,429 units from Debtor of which 5,978,113 units were sold. He also agrees that Morones is essentially right in determining that Returns total 63,445 units. That's about the sum total of their consensus.
A Return is defined as the reversal of a sale upon a customer's refund request. See Attachment 1.
Williams' presents six opinions, two related to unaccounted inventory, two related to proceeds unremitted to Debtor, one concerning prejudgment interest and one concerning stay violation damages. These opinions are discussed in order below.
Williams breaks Unaccounted Inventory Damages down into two categories, Expected Ending Inventory and Reimbursement Adjustments.
Opinion 1 - Expected Ending Inventory.
Morones' Damages Report calls for damages of $641,521 for unaccounted inventory. Williams says the correct measure of such damages is between $0 and $137,516. The difference is, in part, because Morones used January 31, 2014, as the last date she considered inventory transactions. However, after that date, Amazon removed 19,890 units. Williams contends the maximum unaccounted units total 20,405 but that Amazon gave Debtor Reimbursements for 28,541 units leaving Amazon having over-Reimbursed Debtor by 8,136 units. Williams finds Defendant owes nothing to Debtor for unaccounted units. If Adjustments are not considered, Williams says unaccounted units are, at most, 8,431 units which, at $16.31 per unit, brings damages for unaccounted units to $137,516.
This number is the same in both the Morones Report and Morones Declaration.
Valued at $324,406 (19,890 units x $16.31/unit). 12,107 units were shipped back to Debtor while 7,783 units were destroyed per Debtor's request. Exhibit 5 attached to trial Ex. 5 at page 73 is a chart that has two columns marked as "Post 1/31/2014." The Court assumes the first of these two columns was meant to be identified as "Pre 1/31/2014."
Williams' states that the M15 Data shows Amazon actually reimbursed Debtor for 41,911 units but only the "Warehouse Lost/Damage" amount of reimbursed units should be included here.
Opinion 2 - Reimbursement Adjustments.
Continuing with his discussion of unaccounted inventory, Williams' second opinion is that Morones incorrectly added to her damage calculation $1,180,202 for Reimbursable Adjustments. In his view the data does not reliably show how many lost or damaged units Amazon is responsible for and which are not otherwise accounted for by Reimbursements Inventory transferred by Amazon into Debtor's account and units which are already tallied as Unaccounted Inventory.
Adjustments are made through a series of codes such as Code F (unit Found), Code D (unit Destroyed), etc. These Adjustments can increase or decrease the Debtor's inventory balance, or they can change a unit status from sellable to unsellable without a change to the overall inventory balance.
Williams contends Amazon's Adjustments codes cannot be used to determine the quantity of units that have been damaged, lost or stolen and for which Amazon is liable. There is not a 1:1 correlation between Adjustments transactions and discrete units. His reasons for this contention are:
(a) A given unit of inventory cannot be traced from Receipt to disposition through the M15 Data. Morones used the M15 Data to attempt this tracing.
(b) A given unit can have several coding events while in Amazon's possession. That unit could be booked as a Receipt, but later coded as mis-placed (Code M), then found (Code F), then damaged in Amazon's fulfillment center (Code E), and finally Removed (i.e. sent back to Debtor or destroyed). If the total amount of mis-placed inventory is added to the total number of units destroyed there will be some overlap. The extent of that overlap cannot be ascertained through the M15 Data.
(c) Numerous codes do not signify a final disposition which would increase or decrease the number of Debtor's inventory units. Codes P, E, H, K, Q, U, 6 and 7 result in a re-characterization between sellable and unsellable units. Using any of the M15 Data pertinent to units bearing these codes will incorrectly increase or decrease Morones' damage calculations.
(d) Some coded data does not necessarily reveal who is responsible for the changed status of a given unit. A Code D (destroyed unit) could be Debtor's loss (if, for example, it was destroyed because the unit was defective) or could be a loss for which Amazon is responsible (i.e. a unit damaged at Amazon's warehouse).
(e) The final disposition of a unit may not be reflected in Adjustments. If a unit is given a Code E it will likely be Removed but that Removal will not alter the number of units which were given a Code E. Some of those Code E units may remain in Ending Inventory, albeit as an unsellable unit.
(f) Morones assumes that a Lost/Damages Adjustment must give Debtor a reimbursement, but Amazon could simply replace the misplaced unit (Code M) with an identical unit owned by Amazon (Code N). Morones did not account for Code N units. The M15 Data indicates there were 16,585 Code N units.
Morones herself noted this same problem and indicated she worked with total unit counts not individual units. See Trial Ex. 7 at Bates No. 700006.
See DE 310, Williams' December 21, 2020, Declaration at page 6, ¶ 16e which further discusses this scenario.
Williams' Declaration at DE 310 also discussed Code 6 adjustments (damage by inbound carrier). Note that a Code 6 damage could be an Amazon responsibility (if Amazon's carrier was used) or a Debtor responsibility (if a non-Amazon carrier was used). This further highlights the problem with using a given code as applicable to universally lay the damage responsibility on one party or the other. The Trustee reveals that only 2 units were coded 6. See DE 396, p. 62, citing Trial Ex. 5, Williams' Rebuttal Report at Ex. 5.
At bottom, Williams contends that Morones' claim that Amazon is liable to Debtor for 72,357 units of Reimbursable Adjustments is based on unreliable assumptions pertaining to the three components of her Reimbursable Adjustments.
Those three components are (1) Warehouse Damage Adjustments (Codes D, E, 5, 6, and 7), (2) Lost/Found Adjustments (Codes M and F) and (3) Mis-Received Adjustments (Code X).
Williams Rebuttal Report next focuses on two components of proceeds unremitted to Debtor by Amazon. Those components are sales proceeds not remitted to Debtor and Unpaid Refund Reimbursements.
Opinion 3 - Sales Proceeds Not Remitted to Debtor.
Generally Williams asserts that Morones' analysis "regarding the unremitted proceeds are fraught with false assumptions and based on incomplete data, both of which lead to incorrect and unreliable conclusions." Williams notes Morones reliance on the Settlement Data to determine the $1,156,495 amount of sales proceeds not remitted to Debtor failed to account for six data gaps of 48 hours. Morones and Cone were aware of these Data Gaps, but she did not disclose this in Morones' Damage Report. When Amazon retrieved some of the information that could help fill the Data Gaps, it accounted for 87% of the so-called unremitted sales proceeds. Williams did not focus specifically on the $172,851 which Morones' Declaration indicates would remain if the Data Gaps are filled.
Trial Ex. 5, ¶ 46.
Defined by the Court as the Data Gaps. See Attachment 1.
Morones declared that, if 87% of the Data Gaps were filled by this new information, the sales proceeds unremitted would drop to $172,851.
Opinion 4 - Unpaid Refund Reimbursements.
Williams points out that Morones' calculated Unpaid Refund Reimbursements as $237,706 yet she failed to account for $41,985 of refunds reflected as "miscellaneous order adjustments." Seeing this error, Morones's Declaration corrected her damage number down to $195,721. Williams points out that Morones appears to assume all Refund Reimbursements would appear in the Settlement Data as "refund reimbursal" but ignores the data showing some "refund reimbursements" are captured in "miscellaneous order adjustments."
See Morones' Declaration (Ex. 172) at page 4 but then her chart at page 6 states this number should be $186,247.
Opinion 5 - Prejudgment Interest.
Williams recounts how much of Morones damage amounts are incorrect. Then he calculates interest from the Complaint date at the same 12% used by Morones. He also criticizes Morones' use of the "mid-point of Debtor's relationship with the FBA Program."
July 9, 2013.
12% is the legal rate under Washington law. The parties' Contract is governed by Washington law.
Opinion 6 - Stay Violation Damages.
Williams Rebuttal Report calculates Debtor's stay violation damages to be, at most, $47,497. He notes the period of the stay violation was 15 days not 21 days as reported by Morones. He criticizes her for her use of a five-month damage period following the stay violation, suggesting there is no proof of such continued damage. Next, Williams contends Morones' use of a baseline period from November 2012 to March 2013 is wrong because the Debtor's sales had stabilized by July 2012. The average sales for those four earlier months were 14% less than the November to March sales. Finally, Williams claims Morones erred by assuming a sales growth rate from April through October 2014, while sales topped out several months before the stay violation. Using a July 2012 to March 2013 baseline period, Williams calculates Debtor's April 2013 revenue at $1,495,357. The contribution margin would be 9.8% or $147,165. Since Debtor's actual April 2013 contribution margin was $99,668, he calculates that Debtor's stay lift damages would not exceed $47,497. Williams finds insufficient support for any damages after Amazon re-opened its online platform to Debtor after April 25, 2013.
See Exhibit 10 attached to his Report.
April 11, 2013 to April 25, 2013.
$147,165 less $99,668 = $47,497.
In summary, Williams finds Plaintiffs damages could not exceed:
(1) Unaccounted inventory - $0, if Adjustments are factored in. If Adjustments are not factored in it would produce a number for damages based on unaccounted inventory, of $137,516. Prejudgment interest on that amount would be $118,602, (2) Proceeds unremitted to Debtor - $0, and
But this seems to omit the $186,247 of Unpaid Refund Reimbursements.
This would become $489,862 if the unremitted proceeds number is $186,247.
b) Williams Declaration
Williams' direct evidence at trial was presented in the form of his ten-page February 4, 2021, Declaration/Direct Testimony of E. Weiant Williams ("Williams Declaration"). Attached as Exhibit A to the Williams Declaration is his Rebuttal Report from September 14, 2020.
See Ex. 5.
Williams Declaration is broken down into the same six opinions stated in his Rebuttal Report. His Declaration does not address changes made to Morones' Damages Report in her February 5, 2021, Declaration because, of course, Williams' Declaration was filed a day before Morones' Declaration.
Williams includes Table 1 which summarizes his calculations with and without Adjustments applied. Williams also corrects his Rebuttal Report to clarify that units designated with a Code E (damaged at Amazon's fulfillment center) are not Removed but, rather destroyed (Code D) or transferred out of inventory (Code O). Next, Williams offers two hypotheticals that explain why his inventory accounting methodology is correct and compares it to how he believes Morones' methodology would produce an incorrect tally of units handled by Amazon. Finally, Williams declares that, from January 2013 through June 2013, the Debtor paid Amazon fees of $2,206,369.
DE 259 at ¶ 31.
Williams does not explain why that time frame is relevant or what data he relied upon to conclude the amount paid to Amazon.
c) Williams' Cross and Re-Direct Examinations
(February 19, 2021, at 1:43 pm to 4:00 pm)
Williams' direct examination was exclusively done by his Declaration. His cross-examination revealed the following:
Williams noted he had never testified in court as an expert, nor had he ever before done work for Amazon. Williams billed his time in preparing his Report at $375 per hour. Through October 2020 he and his group at Claro had billed $382,000 for the Rebuttal Report work.
Williams interviewed none of the 14 people listed by Amazon as likely to have discoverable information. He did, however, interview three Amazon employees (principally Bachand) but none of them worked for Amazon from 2008 to 2013. His first interview was on June 14, 2019.
Williams indicated one must read the Inventory Adjustment document to understand how Amazon reconciles inventory it held for the Debtor. He also pointed to the Seller Central inventory policy, which policy was periodically updated or revised. Williams described the Seller Central Data as going to the financial data of the parties' relationship whereas the M15 Data is about the Debtor's inventory held by Amazon. Williams described the M15 Data as an "extremely, extremely large data set." He acknowledged Amazon had a duty to maintain contemporaneous data records but does not know whether that duty extends to preservation of that data.
Williams acknowledged his Rebuttal Report was incorrect concerning Code E (damaged at Amazon fulfillment center) implications. He corrected this error in his Declaration. Code E is not a code signifying Removal of a unit. Not all unsellable units (like Code E units) are subject to Removal. Not all of the 10,000 Code E units correspond to one unit per Code E designation.
Williams indicated Code O (Out) units are unrelated to Code N (In) units. He says Code Q (Damaged-Miscellaneous) units are unsellable but are not chargeable to Amazon. A "Q" coding does not reduce units in Debtor's inventory count. Amazon is not liable for a Code Q unit he says. Code Q was the most heavily used Inventory Adjustment code in the M15 Data at 168,526 units. Williams thinks the Code Q definition was poorly written by Amazon.
Ex. 5, page 73.
Williams discussed the M15 Data's reporting of software corrections identified as Code 1 (at 125,661 units), Code 2 (at (79,892) units), and Code J (at 30,897 units). These software correction codes do not refer to physical units but, rather are used to plug the difference between actual physical units and the virtual or electronic accounting of units. Williams says Software Code Adjustments are extremely complicated and are not susceptible for use in calculating Debtor's inventory units.
On October 17, 2013, Debtor changed its choice on how its inventory was to be Removed by Amazon. Prior to that date, Amazon automatically returned unsellable units to Debtor so long as the unit was not damaged in Amazon's warehouse. After October 17, 2013, Debtor had to manually request that Amazon return such unsellable units to the Debtor. Bachand testified that if Debtor did request that such units be returned to Debtor, Amazon would destroy the unit 90 days after the unit was identified as unsellable.
February 18, 2021, Trial Transcript at 159:12 - 160:4. See also Ex. 245.
Id. at 160:5-8.
Although Debtor ceased its operations on January 31, 2014, Amazon either destroyed, sent back to Debtor or continued to retain Debtor's units. Of the19,850 units removed after January 31, 2014, 9,500 were sellable as of January 31, 2014. A total of 19,890 units were Removed (destroyed or shipped back to Debtor) after January 31, 2014, 12,107 units were shipped to Debtor, and 7,783 units were destroyed. Additionally, as of March 2015, the M15 Data showed Amazon continued to retain 20,405 units. No evidence was presented to indicate whether, after March 2015, Amazon returned these units to Debtor, destroyed the units, or otherwise compensated Debtor for those units.
DE 310, Williams Declaration of December 21, 2020.
DE 310 page 3 ¶ 12
The Debtor could not, of course, have made a request to Remove these units because this data was not produced to the Trustee until his lawyers received the M15 Data, two years after the commencement of this Adversary Proceeding.
See Ex. 5, page 73, Code D line.
Williams was not subjected to re-direct examination by Amazon's counsel.
3. Timothy Shaffer
(Defendant introduced Shaffer's direct testimony by designating portions of his depositions from April 24, 2019, and July 3, 2019.)
Shaffer was designated as a Fed.R.Civ.P. 30(b)(6) witness for Debtor in the July deposition.
In his Rule 30(b)(6) Deposition, Shaffer testified about Debtor stabilizing its revenues after he became Trustee. Shaffer also reviewed the automatic stay violation damages in Morones' Damages Report. Shaffer acknowledged Morones' Damages Report stated the automatic stay violation period was 21 days while the Complaint alleged 14 days. However, Shaffer was unable to testify as to which timeframe was accurate. Shaffer was also unable to testify as to Morones' source in stating lost profits calculations extended to 21 days. Schaffer stated that his personal damage calculation for the stay violation would have been higher but acknowledged Morones' estimate was reliable.
In his April deposition, Shaffer discussed Bellino's involvement in the bankruptcy proceeding and his expectations that Bellino would continue operating the business and carry out any directives requested by him. Shaffer testified that Bellino seemed straightforward, honest, and provided accurate information to him. Shaffer also testified about his limited relationship with Lawcock during the Debtor's bankruptcy.
At his April deposition, Shaffer reviewed the Complaint and tried to explain the basis of his allegations, including allegations as to why or how Amazon failed to account for inventory worth more than $1.5 million. Shaffer further discussed his understanding of what led Debtor to file for bankruptcy and noted there were lots of unanswered questions about Debtor's inventory. Ultimately, Shaffer believed Debtor's inability to rectify inventory issues led to significant working capital restraints. Schaffer also discussed Debtor's consignment program through the FBA Program post-bankruptcy. The consignment program addressed Debtor's issue with getting products and the vendors' worries about not getting paid once their product was sold. Under the consignment program, a vendor's product remained property of a vendor while that property was in Amazon's fulfillment center. Shaffer also discussed Debtor's inventory control systems prior to filing bankruptcy and explained Debtor relied mostly on Seller Central Data for information. After bankruptcy, however, Schaffer testified Debtor tried to internally implement better practices to track its inventory. Finally, Shaffer reviewed various Reimbursement claims submitted to Amazon by Bellino and Lawcock and discussed the reason and process for submitting the claims.
4. Justin Ice
(Defendant introduced Ice's direct examination through designations from his deposition conducted on July 28, 2107.)
Ice was hired by Amazon as an Investigation Specialist in 2013. In response to a seller's request, Ice would research Amazon's databases for information. Ice testified that he was frequently selected to conduct intensive reviews of high-volume sellers, including Debtor.
Ice testified about Code M, which indicates a product is missing. A Code M automatically starts a 30-day timer. Unless the inventory is offset by a Code F, Amazon will issue a reimbursement within 30 days. If the reimbursement is of high value, upper-level management will review the reimbursement before it is processed. Ice also testified about how Amazon's reimbursement liability can be offset by a Code N. A Code N is used when inventory, previously found and placed in an Amazon holding account, is transferred to the seller's account as a replacement for the missing product. A product is placed in the Amazon holding account when Amazon's ownership determination engine cannot identify where a product belongs.
Ice further testified about the Deep Dive and reconciliation process he and Francisco Quintana performed. The Deep Dive was an attempt to look at Debtor's top 100 ASIN's, review their accuracy, and determine trending issues. Ice testified that he believed the range selected was statistically relevant in terms of valuation and how it impacted Amazon's potential liability. However, he ended up looking at more than just 100 ASIN's.
Ice also detailed what all that went into the Deep Dive. The crux of the Deep Dive involved tracking the potential changes or shifts in identifiers and seeing how many different identifiers existed. Ice explained the importance of knowing all the identifiers associated with a product to look at a product from a reconciliation standpoint. Ice testified that Debtor had multiple FNSKUs and MSKUs tied to a single ASIN, and it was apparent Debtor created multiple identifiers for either the same item or items which were merged into the same ASIN. Once Ice had collected all the identifiers, step 2 involved putting the data together and looking for things that did not make sense. Ice primarily relied on Seller Central Data and other tools that allowed him to track ASIN movement. During this process, Ice noticed Debtor's claim overlooked information in Seller Central Data, including reimbursements that occurred on order IDs.
Ice also testified about Debtor's claim that was submitted to Amazon. Ice acknowledged that he knew Flores worked on the claim. Ice reiterated Flores' conclusions that the claim was conducive to errors and potentially offsetting or compensating events were not being used to resolve discrepancies across inventory items. Ice clarified that he never made conclusions about where Debtor's data supporting its claim came from but was mostly concerned with checking the data's validity.
5. Jeff Moore
(Defendant introduced Moore's direct testimony by designating portions of his depositions from December 5, 2017.)
Moore was designated as a Fed.R.Civ.P. 30(b)(6) witness for Amazon. During his deposition, Moore provided general background information on the FBA Program. Moore also testified about the differences between Amazon's distribution centers and Amazon's fulfillment centers and discussed Amazon's inventory processes. Once a shipment arrived at an Amazon fulfillment center, Amazon associates (employees) would count items one by one and place them into a tote or cart. A tote is a plastic bin Amazon used to transport products to their designated shelf, where another associate would store the products. At the time products are counted associates are also to inspect for damage. If a product is damaged, the product is placed into a damaged tote. If the product arrived damaged, the product would be placed in the seller's inventory and the seller could have the product Removed. Moore next explained Amazon's Inventory Reconciliation formula, which helped sellers understand how much inventory they had on hand. Moore also explained that between 2008 and 2013 Debtor could check the inventory it had at an Amazon fulfillment center at any given time under a website link in Seller Central called Manage Your Inventory.
Moore explained how Amazon allocates inventory events or damages among commingled products. Based on his recollection, Moore believed only one seller's inventory would be in a tote at a time if products in the tote were commingled. Moore further explained how Amazon could ship another seller's product to fulfill a customer order so long as the product was the same as the seller whose product was sold. The seller whose inventory made the sale would be reduced, and the seller whose product was shipped would be credited a unit.
Moore further discussed how Amazon processes Reimbursements if a customer returns a product. When a customer returns a product, Amazon will run questionnaires and determine who is at fault for the return-the Debtor or Amazon. If Amazon is at fault, a Reimbursement is issued to Debtor, which is typically the sales price minus Amazon's fees associated with the transaction. Moore also explained what happens when a product is not sold because of damage caused by Amazon. Amazon uses an Inventory Valuation System that searches a product's sale history to determine what the replacement/reimbursement value is. Lastly, Moore discussed Code M and how Amazon has an obligation to reimburse Debtor typically within 30-days for missing products.
6. Eric Soder
(Defendant introduced Soder's direct examination through designations from his deposition conducted on July 27, 2017.)
Soder was a Product Manager for Amazon in 2009. Soder testified about Amazon's fulfillment center processes in 2009. When a package arrived at the Amazon fulfillment center it was placed on the receiver line. Next, an associate would look for a label identifying the seller and a unit barcode identifying the number of units in the shipment. The associate would ensure the seller and unit count matched the virtual shipment notification already in Amazon's system. If Amazon's system did not recognize the shipment, the shipment was flagged for a "problem solve" and set aside. Once a product was successfully "checked-in," it was placed in a tote or on a designated shelf.
Soder testified about how and when missing codes and found codes were assigned. A Code M is typically assigned when units are cycle-counted, and a product is missing from its designated location. A Code F is only assigned if a problem-solving agent is unable to figure out where the missing unit came from. Whether a product gets a Code M or Code F, it is assigned a damaged adjustment code, which is reported in Seller Central. To determine what specific code should be assigned, a problem solver would offer different inputs into Amazon's system. Based on the summation of inputs, Amazon's system would determine the specific code. Soder also explained the hierarchy of Amazon warehouse employees. The basic pyramid structure, starting from the bottom, included receivers, pickers, packers, and problem-solving agents. Receivers simply received the shipments as they arrived and performed the check-in process. Pickers would search and find inventory when it was requested for removal. Packers packed inventory for shipment, and problem-solving agents evaluated inventory if there were issues.
Lastly, Soder testified about his time visiting Debtor in Phoenix and Debtor's operations. Soder went to Phoenix because Debtor was experiencing listing errors, meaning Debtor had products in Amazon's fulfillment center that were not offered on the amazon.com for sale. Soder believes Debtor experienced listing errors because it would create an inbound shipment for Amazon to fulfill, send the inventory to Amazon, but then change the inventory back to merchant fulfilled and ship the product from Debtor's warehouse. Soder testified about other issues Debtor faced, like products not showing as listed on Amazon's website and a product's ASIN listed on Amazon's website resulting in a 404-page error. Soder also stated Debtor had issues with listing the same product at different price points and struggled to have inventory returned to Debtor's warehouse. Soder testified that he thought Debtor's overall operations were unorganized and that there was internal confusion.
7. Jeffrey Cone
(Defendant introduced Cone's direct examination through designations from his deposition conducted on August 17, 2020.)
Cone testified regarding the scope of his engagement with Morones and his analysis, which informed Morones' Damages Report. Specifically, Cone took data from different Seller Central Data, including Inventory Activity Reports and Inventory Status Reports, and imported the data into database tables to prepare summaries. However, Cone's summaries were not the basis of Morones' Damages Report. Cone also testified regarding the software programs he created to process the Excel file settlement data. Although he preserved the codes that the software program relied on, he did not preserve the software in a way in which Amazon could run it.
Cone gave detailed testimony on information he relied on in his analysis and information he used for general background knowledge. For his analysis, Morones instructed Cone to focus only on the M15 Data and Seller Central Data. However, he did incorporate information from Amazon's Amended Responses and Objections to the Trustee's third set of discovery request and Amazon's adjustment reason codes, which explains how Amazon defines the M15 Data adjustment codes. Cone also clarified he used all raw data for his analysis and not Ashworth's analysis. Because Cone could not get a clear understanding from Ashworth on how he reached his conclusions, Cone used Power Builder to reverse engineer Ashworth's work back to the original data source. Cone's reverse engineering process was not reflected in Morones' Damages Report.
Cone further testified in detail about Settlement Reports he reviewed from 2003 through 2013 and confirmed he tested the Settlement Reports for completeness at some point in 2018. This is how he noticed gaps in time where nothing was reported. Cone acknowledged that the gaps (Data Gaps) in the Settlement Reports were not disclosed in Morones' Damages Report but confirmed he did notify Morones about the gaps. Cone testified no one looked for the missing information because they were previously told they had all the data that was available. Thus, Cone treated the gaps in his analysis as if Debtor did not receive any cash from Amazon during those periods. Although Cone testified he reviewed one of Debtor's bank deposit statements and it matched cash transfers in the Settlement Report for that period, Cone claimed Debtor's bank statements were unavailable for the periods of the Data Gaps.
8. Stephen Ashworth
(Defendant introduced Ashworth's direct examination through designations from his deposition conducted on August 19, 2020.)
Ashworth is a technology consultant and the owner of Ashworth Consulting, LLC. Ashworth was originally engaged by the Trustee to assist in the perseveration of Debtor's 2013 records. Ashworth discussed how he preserved the Debtor's records, and how he used a cloning tool to convert Debtor's PCs into VHDs. Ashworth testified that he believed, "electronic-record wise," all Debtor's records were preserved. He also confirmed he personally oversaw the collection of Debtor's PCs.
Defined in Attachment 1 as Virtual Hard Drives.
Ashworth explained what he accessed and downloaded from Seller Central including Settlement Reports, Inventory Event Detail Reports, Inventory Transaction Reports, and Monthly Ending Inventory Reports. The purpose of searching Seller Central was to track the final disposition of Debtor's 6.5 million product units which were submitted to Amazon.
Ashworth specifically testified about the Settlement Data in Seller Central. He was unable to retrieve the Settlement Data prior to January 2012 because Seller Central only had a 2-year look-back time limit. However, a former Amazon employee, Dewberry, instructed Debtor towards the end of 2011 and before Debtor's full audit to download the Settlement Data from before 2012. Ashworth testified that he did notice gaps in the Settlement Data the Debtor's employees downloaded and was told the gaps occurred because of problems with Seller Central timing out. He informed Morones and Cone of the gaps (the Data Gaps).
Lastly, Ashworth testified about his initial engagement with the Trustee, whether he knew he would possibly be called as a testifying expert, as well as his relationship with Morones and Cone. Ashworth acknowledged that he thought he might provide expert testimony about inventory claims. However, once Morones became more involved, Ashworth became more of a secondary expert.
9. Thomas Reilly
(Defendant introduced Reilly's direct testimony by designating portions of his deposition from May 1, 2019.)
Reilly was a former employee of Debtor and testified about his professional relationship with Bellino and Debtor's operations. Reilly formed a professional relationship with Bellino while working at United Natural Foods and Lone Star Distribution. Bellino and Reilly's professional relationship led to Reilly working for Debtor. After six to nine months of working for Bellino, Reilly quit working for Debtor. Reilly explained that he quit for multiple reasons, including his unhappiness with Debtor's business practices. Reilly also testified that Bellino's integrity played a role in his decision to leave Debtor. Reilly could not recall the specific point and time he started questioning Bellino's integrity and business choices.
Reilly testified about Debtor's general operations and practices while he worked there. Regarding Debtor's inventory tracking system, Reilly stated Debtor's tracking was "not what [he] thought it would be" and it was "extremely low quality." Further, Reilly testified that there seemed to be a disconnect between the buyers managing the inventory in Seller Central and the warehouse team managing the physical inventory in the building. Reilly further testified about Amazon's FBA Program at the time he worked with Debtor, which FBA Program was still evolving. Reilly testified Amazon reached out to Debtor and helped Debtor get set up with the FBA Program.
During his deposition, Reilly reviewed the Debtor's Disclosure Statement and disagreed/agreed with certain statements. Reilly disagreed with the statement that Debtor "changed its operating model from an Amazon fulfillment method to . . . merchant self-fulfillment." He recalls while he was working for Debtor, the Debtor only shipped smaller, lighter items through the Merchant Program. However, Reilly agreed with statements blaming Debtor's inventory tracking and internal processes for causing difficulty.
Lastly, Reilly testified about Debtor selling expired products and how Bellino would sell products below Debtor's cost because he wanted to win the "buy box." By clicking the "buy box," a customer automatically purchases from the seller with the cheapest price. Furthermore, Reilly described issues the Debtor had with tracking its products because employees would create new listings on Amazon without creating the item information in OMX, Debtor's inventory tracking system.
10. Sean Lawcock
(Defendant introduced Lawcock's direct testimony by designating portions of his deposition from April 23, 2019.)
As part of his deposition, Lawcock testified Debtor did not entirely transition its operating model in 2010 from the FBA Program to the Merchant Program. Lawcock explained that under the Merchant Program, Debtor kept low inventory on the shelf because it could get most items directly from the merchants and still get orders out on time, rendering high volumes of inventory pointless. Because of this policy, Debtor did not have a procedure in place to keep track of products in its warehouse. Lawcock described the Debtor's storing and tracking process of products in the warehouse as a "quick visual," meaning a walk through the stored items would easily show what was available and what needed to be ordered.
Debtor also did not track its sold inventory through the FBA Program prior to bankruptcy because the Debtor had faith in Amazon's tracking systems. However, Debtor did use Seller Central Data to view the amount of inventory in Amazon's fulfillment centers, what was available for sale, and the amount of inventory inbound to Amazon. Lawcock testified that he was not sure Debtor recorded any information Amazon provided to Debtor. However, he maintained his own records through Microsoft Excel from 2010 through 2011. After filing bankruptcy, Debtor created its own internal audit processes for tracking inventory.
During his deposition, Lawcock testified about an email chain between Soder, Reilly, and himself with the subject line "Item Not In Catalog" regarding a 404 error message. In one email response, Soder explained the reason for the 404 error was because Debtor had "2 new SKUs mapped to the new ASIN." Lawcock explained that this sometimes occurred when Amazon would merge two ASIN listings that seemed to be the same product ASIN Merge. The result of the ASIN Merge was multiple SKUs being assigned to the same ASIN, which impacted Debtor's ability to track inventory. Responding to Soder's email, Lawcock recommend that one of the SKUs be deleted. However, Lawcock clarified in his deposition that a SKU could never be deleted.
Lawcock also testified about other situations that made Debtor's ability to track its inventory difficult. For example, Amazon incorrectly formatted SKUs and there were "Dot Missing SKUs." A "Dot Missing SKU" happened when Amazon would just choose a SKU for inventory it received if it did not know which SKU it belonged to. Without reaching out to Amazon, Debtor could not track its inventory because it was uncertain what was listed on the "Dot Missing SKU."
11. Daniel A. Bellino
(Defendant introduced Bellino's direct testimony by designating portions of his deposition from April 22, 2019.)
Bellino was an owner and founder of the Debtor. Bellino testified about the beginnings of Debtor's operations in 2006 and what led the Debtor to file for bankruptcy. When the Debtor first used the FBA Program in 2007, Amazon's reporting and data generation capabilities were limited. However, Amazon's reporting improved over time. Bellino also testified about Debtor's process for shipping products to Amazon. The Debtor would receive products from distributors, label the products with printed stickers provided by Amazon, and ship the product to Amazon. Bellino further discussed Debtor's consignment program process, where vendors would ship the product directly to Amazon, rather than Debtor.
A large portion of Bellino's deposition was dedicated to discussing his understanding of Debtor's operations, the FBA Program, Amazon's processes, and tracking inventory through Seller Central. Debtor relied on Seller Central Data and an internal SQL program to track inventory. Bellino explained the roles Lawcock, Reilly, Peeples, and Schmidt played in managing Debtor's operations. Bellino also discussed the Debtor's process for managing Aged Inventory, which is inventory that has not been sold within 90 days. Debtor had different processes for managing Aged Inventory under the FBA Program and the Merchant Program. Under the FBA Program, Debtor would try to either lower the price or call it back for a return. Bellino further testified about Debtor's reorganization process and the steps Debtor took to improve its operations, like restructuring management to be more efficient. Debtor also hired SP Express, a third-party company, to outsource its supplier receiving and product shipment process to Amazon.
During his Deposition, Bellino reviewed multiple emails between Bellino and Debtor's other managers and emails from Amazon. The emails regarded Debtor's operations through the FBA Program and primarily pertained to issues with inventory. However, Bellino did not remember any of the email communications and could not answer substantive questions. Bellino also reviewed multiple documents related to Debtor's bankruptcy and operations during the deposition and, similarly, could not remember anything of substantive value.
E. Post-Trial Proceedings
Post-trial briefs were submitted by the Trustee on issues concerning burden of proof and pre-judgment interest. Amazon filed its briefs on the burden of proof issues and pre-judgment interest issues. Oral argument on these briefs was conducted on April 19, 2021, after which this matter was taken under advisement.
DEs 371 and 383.
F. Tentative Under Advisement Order
The Court assumed both parties would find fault with the Court's Tentative Under Advisement Order. The parties did not disappoint.
By stipulation of the parties comments were required by January 12, 2022, and responses by February 9, 2022. Oral argument was rescheduled. The Trustee's comments were timely filed as were Amazon's. Amazon's response and Trustee's reply were also timely filed. Oral argument was held via Zoom.gov on February 24, 2022, after which the Court again took this matter under advisement.
VII. ANALYSIS OF TRUSTEE'S CAUSES OF ACTION
A. Count 1: Failure to Turnover Estate Property Pursuant to § 542(a)
1. Legal Analysis
Section 542(a) states:
Except as provided in subsection (c) or (d) of this section, an entity, other than a custodian, in possession, custody, or control, during the case, of property that the trustee may use, sell, or lease under section 363 of this title, or that the debtor may exempt under section 522 of this title, shall deliver to the trustee, and account for, such property or the value of such
property, unless such property is of inconsequential value or benefit to the estate.
The United States Supreme Court held that turnover was appropriate only "when the evidence satisfactorily establishes the existence of the property or its proceeds, and possession thereof, by the defendant at the time of the [turnover] proceeding." § 542(a) does not have a "present possession" requirement.
In re Newman, 487 B.R. 193, 201 (B.A.P. 9th Cir. 2013).
Plaintiffs turnover claim rises or falls with Plaintiffs Count 3 breach of contract claim. If Plaintiffs breach of contract claim is not proven, there would be nothing for Defendant to turnover to Plaintiff.
2. Conclusion on Turnover Claims
Plaintiffs turnover cause of action is closely intertwined with Plaintiffs Count 3 breach of contract claim. These two causes of action will be discussed together in the breach of contract section below.
B. Count 2: Violation of the Automatic Stay Pursuant to § 362
1. Legal Analysis
Under the Bankruptcy Code, "[a]n individual injured by any willful violation of a stay … shall recover actual damages, including costs and attorneys' fees, and in appropriate circumstances, may recover punitive damages." While a corporate entity can be a person, it cannot be an individual for purposes of § 362 because "individual" is not synonymous with "person" under relevant provisions of the Bankruptcy Code.Nonetheless, a corporation may be entitled to recovery for a stay violation under § 105(a) as a sanction for civil contempt. Under § 105(a), the Court may impose civil sanctions sua sponte for a stay violation.
Johnston Environmental Corp. v. Knight (In re Goodman), 991 F.2d 613 (9th Cir. 1993). See also In re Chateaugay Corp., 920 F.2d 183 (2d Cir. 1990) (Corporation could not recover compensatory damages for creditor's willful violation of automatic stay; Bankruptcy Code provision authorizing award of damages was applicable only to natural persons).
In re H Granados Commc'ns, Inc., 503 B.R. 726, 733 (B.A.P. 9th Cir. 2013) (Citing Johnston Envtl. Corp. v. Knight (In re Goodman), 991 F.2d 613, 620 (9th Cir. 1993)). See also In re Dyer, 322 F.3d 1178, 1191 (9th Cir. 2003) (for civil contempt purposes, the automatic stay under § 362 "qualifies as a specific and definite court order.").
§ 105(a) ("No provision of this title providing for the raising of an issue by a party in interest shall be construed to preclude the court from, sua sponte, taking any action or making any determination necessary or appropriate to enforce or implement court orders or rules, or to prevent an abuse of process.").
Generally, monetary civil sanctions are imposed to either compensate the complainant for their losses caused by the contemptuous conduct or to coerce the contemnor's compliance with a court order. When the purpose of sanctions is compensatory, a fine, payable to the complainant, must be based on evidence of actual loss.
The Trustee need not prove by a preponderance of the evidence that Amazon's actions caused a specific amount of damages or lost profits. It need only show Amazon's actions caused the Debtor to lose profits. Here, Amazon "should not profit from the difficulty in proving exact damages."
HSS Enters., LLC v. Amco Ins. Co., 2008 WL 1787127, at *13 (W.D. Wash. Apr. 16, 2008).
Milgard Tempering, Inc. v Selas Corp. of America, 902 F.2d 703, 710 (9th Cir. 1990).
Plaintiff is not required to prove Debtor's lost profits with "mathematical certainty." Rather, Plaintiff must demonstrate lost profits with reasonable certainty, in other words the "existence of damages must be taken out of the realm of speculation."Moreover, Plaintiff is not required to prove that Amazon's stay violation was the exclusive cause of Debtor's lost profits. Rather, in proving Debtor's lost profits, "all that is required is a probability" that Amazon's actions caused Debtor's loss.
In re Visser, 660 Fed.Appx. 553, 538 (9th Cir. 2016) quoting the Idaho Supreme Court case of Trilogy Network Systems, Inc. v Johnson, 172 P.3d 1119, 1121 (2007). This Court presumes the 9th Circuit cites to the Idaho Court in general approval of this proposition as well as the proposition referenced in the next footnote below.
Id. Quoting the Idaho Court in the case of Anderson & Nafziger v GT New comb, Inc., 595 P.2d 709-716 (1979).
Ranger Enterprises, Inc. v Leen & Associates, 1998 WL 668380 at *2 (9th Cir. Sept 21, 1998).
At a hearing on April 26, 2013, the Court found Amazon willfully violated the automatic stay. At trial, the Court heard evidence from the Trustee and Azzarelli about the fallout that occurred following the 15 days in April 2013 where Amazon violated the stay by denying Debtor access to its only sales outlet, namely the amazon.com online platform. Even after this Court ordered Amazon to open its platform to Debtor, Debtor's sales never recovered or even came close to pre-stay violation levels. Moreover, Debtor's vendors and Plan supporters were "spooked" by Amazon's actions and Debtor's vulnerability to the actions of Amazon, Debtor's sole online access provider. Debtor had no other platform where it could sell its products besides amazon.com.
Admin DE 262.
See Ex. 7, Table 6a.
This Court finds that reviewing testimony by Azzarelli and the Trustee as well as reviewing the Debtor's monthly operating reports and Morones' Damages Report reveals that Amazon's stay violations did cause Debtor to lose significant profits. Even if other market factors were at play in Debtor's business between April 2015 and October 21, 2015, this Court finds Amazon's stay violation was far and away the precipitating cause of Debtor's lost profits. Evidence of Debtor's track record of profitability and precipitous decrease in profitability during and after Amazon's stay violations support this Court's conclusions. Amazon suggests other factors created headwinds against Debtor's financial performance but has not persuaded this Court that such challenges were either material or should reduce Debtor's lost profits in any event.
See Trial Ex. 160 as well as the Debtor's monthly operating reports for July 2012 (Admin DE 179), August 2012 (Admin DE 180), September 2012 (Admin DE 198), and October 2012 (Admin DE 199).
Amazon's stay violation not only denied Debtor's ability to sell products from April 11 through April 26, 2013, but Debtor's sales were dramatically diminished thereafter until October 22, 2013, when Amazon once and for all shut down Debtor's access to its platform. Williams calculated Debtor's stay violation damages only in the month of the stay violation. The Court rejects this approach as Debtor's damages caused by Amazon's violation of the stay extended through to the date Amazon terminated Debtor's access to Amazon's online platform. The Court finds Amazon is liable to Debtor for the contribution margin on Debtor's reduced sales from April 11, 2013 to October 22, 2013.
DE 332, page 16, ¶ 20.
While cross-examination also revealed that some of Debtor's sales may have been of products which Amazon prohibited to be sold on its platform, the Court will not reduce Plaintiffs stay violation damages due to those sales because, prohibited by Amazon or not, those sales were a part of Debtor's revenues and those revenues dropped precipitously when Amazon wrongfully denied Debtor access to Amazon's platform for 15 days in April 2013. The fact that Debtor also reduced sales prices of some of its products from April 2013 to October 2013 does not aid Amazon's defense to the stay violation damages. Debtor's reduced prices highlight Debtor's desperate cash flow predicament occasioned by Amazon's stay violation. Amazon may not now profit by reducing Debtor's stay violation damages by the difference between (1) what those units would ordinarily sell for and (2) what those units sold for during this April to October 2013 cash crisis caused by Amazon.
Morones testified that the stay violation damages suffered by Debtor should be measured by expected revenues from April 2013 through October 22, 2013, multiplied by a 9.7% contribution margin less the actual contribution margin realized by Debtor during this time frame. Morones gauged expected revenues by using average monthly sales from November 2012 through March 2013 to which she then added a 0.2% expected monthly sales growth. The April 2013 expected revenue number was $1,591,877 but every month thereafter Morones increases expected revenue by the 0.2% projected sales growth.
Exhibit 240. Here, Amazon's counsel tells Morones to use November - March 2013 numbers for stabilized gross income.
Ex 7, Tables 6 and 6A contain Morones' calculations of Debtor's stay violation damages.
This Court discerns two flaws in Morones' projected revenue numbers. First, the Court finds Morones' 0.2% projected sales growth was not supported by the data regarding Debtor's sales from July 2012 to March 2014. The Court will not apply a growth factor to the revenue projections from May through October 2013. Second, the Trustee testified that Debtor's revenues were stabilized beginning in July 2012. Morones should have used average monthly sales from July 2012 through March 2013, not November 2012 through March 2013. This produces an average gross monthly sales amount of $1,495,357 over this nine-month period, not the $1,591,877 amount used by Morones to project April 2013 sales. Using this nine-month period also produces a 9.8% contribution margin percentage.
See Attachment 7 to this Order.
Applying the margin of 9.8% to the stabilized monthly gross sale number of $1,495,357 produces a contribution margin of $147,165 per month. From this, the actual contribution margin must be subtracted, and the month of October 2013 must be prorated. The Court's calculations produce total lost profits of $668,484.
The aggregate monthly lost profits from May 2013 through October 22, 2013. See Attachment 8.
2. Conclusion on the Stay Violation Claims
See 28 U.S.C. § 1961
C. Count 3: Breach of Contract
1. Legal Analysis
The Parties' Contract specifies application of Washington law in the event of a dispute. Under Washington law, a breach of contract is actionable only if the contract imposes a duty, the duty is breached, and the breach proximately causes damage to the claimant. The Trustee asserts in Count 3 of the Complaint that Amazon breached the FBA Agreement by: (1) not sending a Notice of Material Default to the Debtor before terminating the FBA Agreement, as required by the terms of the FBA Agreement; (2) improperly and wrongfully purporting to terminate the FBA Agreement and denying the Debtor access to the Amazon Platform; and (3) failing to store, maintain, preserve, and account for the Debtor's inventory and failing to compensate the Debtor for "Lost Inventory."
Ex 1 at PDF page 10 of 38.
DE 1, ¶¶ 39-47.
2. Burden of Proof
Under Washington law, the claimant must prove damages with reasonable certainty, meaning by a preponderance of the evidence. The "doctrine respecting the matter of certainty, properly applied, is concerned more with the fact of damage than with the extent or amount of damage." Once the claimant establishes the fact of loss with certainty, uncertainty regarding the amount of loss will not prevent recovery. Thus, a plaintiff "will not be required to prove an exact amount of damages, and recovery will not be denied because damages are difficult to ascertain… Generally, whether the plaintiff has proved his loss with sufficient certainty is a question of fact." Any doubts about certainty are generally resolved against the party who breached the contract.
Larsen v. Walton Plywood Co., 390 P.2d 677 (Wash. 1964), adhered to, 396 P.2d 879 (Wash. 1964); Northwest Land & Inv., Inc. v. New West Federal Sav. and Loan Ass'n, 786 P.2d 324 (Wash.Ct.App. 1990); Restatement (Second) of Contracts § 352, comment a. See also Moore v. Health Care Auth., 332 P.3d 461, 468 (Wash. 2014), en banc (The breaching party bears "the risk of the uncertainty which his own wrong has created.").
Although the exact amount of damages need not be shown with mathematical certainty, a claimant must come forward with sufficient evidence to support a damages award. Evidence of damages "is sufficient if it affords a reasonable basis for estimating loss and does not subject the trier of fact to mere speculation or conjecture." The damages must be susceptible of ascertainment in some manner and by reference to some definite standard, such as "market value, established experience, or direct inference from known circumstances." The factfinder does not commit speculation when, "once the fact of damage is established, it is permitted to make reasonable inferences based upon reasonably convincing evidence indicating the amount of damage." On a challenge as to the sufficiency of evidence, "all evidence must be taken in the light most favorable to the plaintiff, and he is entitled to all reasonable inferences." The amount of damages generally is a question of fact.
Fed. Signal Corp. v. Safety Factors, Inc., 886 P.2d 172, 188 (Wash. 1994); Mut. of Enumclaw Ins. Co. v. Gregg Roofing, Inc., 315 P.3d at 1150 (citing O'Brien v. Larson, 521 P.2d 228 (Wash.Ct.App. 1974)).
Gaasland Co. v. Hyak Lumber & Millwork, Inc., 257 P.2d 784, 788 (Wash. 1953) (emphasis supplied) (quoting 15 Am. Jur., Damages, 414, § 23)).
Gaasland Co. v. Hyak Lumber & Millwork, Inc., 257 P.2d 784, 788-89 (Wash. 1953) ("Once such a prima facie showing is made, there is sufficient evidence in the record to permit reasonable inferences to be drawn therefrom as to the extent of damage.").
Mut. of Enumclaw Ins. Co. v. Gregg Roofing, Inc., 315 P.3d at 1150.
a) Burden Shifting
Generally, a plaintiff bears the burden of proving all elements of their claim. A plaintiff establishing a prima facie case does not shift the burden of proof or require the defendant to prove the negative by a preponderance of the evidence. Instead, it merely requires the submission of the issue to the factfinder to determine the preponderance of the evidence. Additionally, a plaintiffs lack of proof on a vital fact may not be cured by the defendant's failure to prove the negative.
Cedar River Water & Sewer Dist. v. King Cnty., 315 P.3d 1065, 1073 (Wash. 2013).
Gillingham v. Phelps, 119 P.2d 914, 919 (Wash. 1941) ("[T]he burden of proof, meaning the obligation to establish the truth of the claim by a preponderance of the evidence, rests throughout upon the party asserting the affirmative of the issue, and unless he meets this obligation upon the whole case [,] he fails.").
Gillingham v. Phelps, 119 P.2d 914, 919 (Wash. 1941).
Emerick v. Bush, 220 P.2d 340, 342 (Wash. 1950) ("The lack of affirmative proof of a vital fact may not be cured by the opposing litigant's failure to prove the negative thereof").
Trustee asserts that once he has met his burden, the burden then shifts to Amazon to persuade the Court to the contrary with admissible evidence. Specifically, the Trustee asserts: (A) once the Trustee met its burden of proving inventory shrinkage through the inventory adjustment reports, the burden shifted to Amazon to disprove such; (B) for Amazon to avoid liability for Code Q units, Amazon had the burden to present admissible evidence showing that Code Q units had not been damaged while being stored and its failure to do so means Amazon is liable for them; (C) Amazon failed to meet its burden showing Morones double-counted inventory under certain adjustment codes through the use of Bachand's testimony and the demonstrative chart illustrating three hypothetical scenarios; (D) Amazon failed to meet its burden to provide sufficient admissible evidence supporting their maximum liability was limited to the loss of 36,792 units;and (E) Amazon failed to meet its burden to prove that some of the Debtor's inventory fell into one of the four exceptions outlined in the FBA Reimbursement Policy that limit Amazon's duty to pay.
DE 370 pg 3 lines 11-13. See also pg. 6, lines 24-25, "Amazon had the burden to produce competent evidence to refute the trustee's evidence."
DE 370, pg. 10, lines 21-22
DE 370, pg. 9, lines 6-8.
DE 370, pg. 12-13.
DE 370, pg. 13, lines 21-24.
DE 385, Trustee's Reply, pg. 1, lines 26-28, pg. 2, lines 1-8.
If Amazon were raising an affirmative defense, then the burden-shifting and preponderance of the evidence standard would be required. However, Amazon is not raising an affirmative defense; it merely argues the Trustee failed to meet its burden of proof on the element of damages being shown by reasonable certainty and with sufficient evidence. An affirmative defense is distinguishable from an attack on a plaintiffs case-in-chief: "A defense which demonstrates that plaintiff has not met its burden of proof is not an affirmative defense." Likewise, a defense that negates an element that a plaintiff was required to prove is not an affirmative defense. Such defenses are merely rebuttal against the evidence presented by the plaintiff
Lake Hills Invs. LLC v. Rushforth Constr. Co., Inc., 472 P.3d 337, 344 (Wash.Ct.App. 2020), review granted, 481 P.3d 546 (Wash. 2021) (explaining a defendant bears the burden of proving an affirmative defense and that an affirmative defense pleads matters extraneous to the plaintiff's prima facie case, which deny plaintiff's right to recover even if all allegations set forth in the complaint are true) (first quoting Erickson v. Biogen, Inc., 417 F.Supp.3d 1369, 1386 (W.D. Wash. 2019); then citing Fed. Deposit Ins. Corp. v. Main Hurdman, 655 F.Supp. 259, 262 (E.D. Cal. 1987) ("[A]n affirmative defense puts the plaintiff on notice that matters extraneous to his prima facie case are in issue and ordinarily allocates the burden of proof on the issue.")).
DE 376, Amazon's Response on Burden of Proof, pg. 2, lines 23-24, "The Trustee is trying to shift the burden of proof because the evidence does not support his claims …."
Zivkovic v. S. California Edison Co., 302 F.3d 1080, 1088 (9th Cir. 2002) ("[Defendant's] attempt to prove that it provided a reasonable accommodation merely negates an element that [Plaintiff] was required to prove and therefore was not an affirmative defense …").
Trustee asserts Amazon was in possession of the inventory and had a duty to keep track of that inventory. Because Amazon was in a better position to know and produce facts concerning Debtor's inventory and what happened, Trustee controls the burden shifts to Amazon.
DE 370, pg. 6, lines 9-13, "The same principle applies here: because Amazon was in possession of the inventory and had a duty to keep track of it, it was in a better position than the trustee to explain what happened. The Court should thus place the burden on Amazon to prove the precise amount of inventory that was lost, damaged, or destroyed, and if Amazon can't do so, then the Court should accept the trustee's evidence of damages, even if it's only approximate." See also pg. 7, lines 16-18, "Amazon had the duty under the contract to keep track of [Debtor's] inventory, and Amazon was 'in position to know and to produce the most probative facts concerning that inventory." See also pg. 2, lines 4-6, "[I]n which the defendant - Amazon - had sole access to and control of the facts relating to [Debtor's] inventory."
Plaintiff cites to the King County case for the proposition that a party in exclusive control of damage evidence bears the shifting burden to disprove by a preponderance of the evidence that Plaintiff has not been damaged. King County holds that when information necessary to proof is "exclusively within the knowledge of one or the other of the parties, the burden would be upon the party possessed of that knowledge to make the proof" Exclusive means "shutting out all others from a part or share." Unlike King County, here Debtor had access to information regarding its inventory through the Seller Central Data, at least prior to October 22, 2013. Furthermore, even if Amazon had exclusive knowledge of relevant information regarding Debtor's inventory, this was largely cured when Amazon produced the M15 Data, depositions of its employees, and other discovery requests submitted by the Trustee. Because this Court finds that relevant information regarding the Debtor's inventory was not exclusively in the hands of Amazon, the Court also finds the burden of proof does not shift to Amazon.
Cedar River Water & Sewer Dist. v. King Cnty., 315 P.3d 1065, 1073 (Wash. 2013) (quoting Jolliffe v. N. Pac. Ry., 100 P. 977 (Wash. 1909)).
Exclusive, Dictionary.com, https://www.dictionary.com/browse/exclusive (last visited May 17, 2021).
Defined in Attachment 1 as the Termination Date.
The Washington Constitution assigns to the factfinder the ultimate power to weigh the evidence and determine the amount of damages to be awarded, if any. If the Trustee fails to meet its burden of proof, the failure cannot be cured by the Court placing the burden on Amazon to prove the negative and Amazon's failure to do so. Likewise, if the Trustee does meet his burden of establishing a prima facie case, the burden does not shift to Amazon, nor does it require Amazon to prove the negative by the preponderance of the evidence.
Columbia Park Golf Course, Inc. v. City of Kennewick, 248 P.3d 1067 (Wash.Ct.App. 2011) (trial court properly submitted damages claim to jury).
DE 370, pg. 6, lines 11-13, "The Court should thus place the burden on Amazon to prove the precise amount of inventory that was lost, damaged or destroyed, and if Amazon can't do so, then the Court should accept the trustee's evidence of damages, even if it's only approximate."
b) Conclusions Regarding Burden of Proof
An important component of the theory of Plaintiff s damages claim is that Debtor's Contract with Amazon, in effect, called for a "Black Box" into which Debtor shipped its inventory. Plaintiff contends that only Amazon controlled the inventory in the Black Box and any reporting on what was happening in the Black Box was controlled by Amazon. The Trustee suggests that if he could not fully or accurately prove damages then the master of the Black Box (Amazon) should bear the burden of demonstrating all of what Amazon did with Debtor's inventory. In effect, the Trustee suggests Amazon had the responsibility to show it was not liable to Debtor and did not cause damage to Debtor rather than Plaintiff having the burden of proving Amazon's breach of contract and the damages flowing from the breach.
Plaintiffs Black Box theory reminds the Court of the tort concept of res ipsa loquitur where "the mere fact of an accident's occurrence raises an inference of negligence that establishes a prima facie case." Plaintiff has not cited a case applying this concept to a breach of contract case and the Court will not do so now to either find a breach by Amazon or impose upon Amazon a burden to disprove Plaintiffs damages. Among other things, Amazon did not alone control all the data in the Black Box. The data was always shared with Debtor. Testimony at trial suggested that Debtor maintained poor inventory records and Debtor often operated in a chaotic state. It is not clear to this Court that the Debtor preserved all the Seller Central Data or that the information stored by the Debtor was accurate. However, even if this Seller Central Data or the Settlement Data was fully preserved by the Debtor, it was not introduced into evidence at trial by the Trustee. The Court rejects the Trustee's "Black Box" theory of shifting the burden it bears to prove Amazon's breach and/or Plaintiffs damages.
A Latin phrase which translates as "the thing speaks for itself."
Black's Law Dictionary, 9th Ed.
For example, testimony by Soder.
Bachand testified the Settlement Data portion of the Seller Central Data regularly rolled so that the initial Settlement Data presented to Debtor may not be accurate at a later date. See § VII(c)(4).
3. Findings Regarding Breach of Contract and Damages
a) Calculating Replacement Value
Under the Contract, Amazon is to pay Debtor the Replacement Value of a product unit for which Amazon destroys, loses, etc. The Replacement Value calculated by Morones is consistent with the FBA Lost and Damaged Inventory Reimbursement Policy. This Contract provision specifically requires Amazon to reimburse Debtor the "estimated proceeds of the sale of that same item." The "estimated proceeds" is the amount of money that the Debtor would have received if someone had purchased the item.Amazon considers several factors when determining the reimbursement amount, including "your sales history, the average FBA selling price on Amazon, the sales history of the specific ASIN." If there is not enough information to calculate the reimbursement amount for an item, Amazon uses default reimbursement values broken down by product category.
FBA Lost and Damaged Inventory Policy.
FBA Lost and Damaged Inventory Policy.
FBA Lost and Damaged Inventory Policy.
FBA Lost and Damaged Inventory Policy.
Morones' calculation of the Replacement Value looked at the Debtor's sales history and the average FBA selling price on Amazon, both of which are factors Amazon uses to determine the reimbursement value. However, the FBA Lost and Damaged Inventory Reimbursement Policy is unclear whether the "average FBA selling price" refers to all products or the specific product that was lost or damaged. Morones used the average sales price of all Debtor products. The Court agrees with Morones' approach concerning Replacement Value.
In calculating the Replacement Value, Morones relied on the M15 Data and Settlement Reports. Morones stated the M15 Data may not be complete or accurate with respect to every Debtor unit because not all inventory records were maintained in the currently used data warehouse. Amazon contends the M15 Data is the most complete set of data it could produce to account for units in the FBA Program.
Morones' Damages Report, ¶ 10. In concluding the M15 Data is incomplete or inaccurate, Morones relied on the deposition of Bachand, a FRCP Rule 30(b)(6) witness for Amazon, where she states Amazon did not keep reliable records regarding the exact purpose or amounts of reimbursements and that some of the data was backfilled from older data sources. Morones' Damages Report, ¶ 51; Deposition of Bachand, pg. 92, lines 6-16, pg. 142, lines 6-12, pg. 145, line 23, pg. 147 line 12.
Relying on the Settlement Reports and M15 Data, Morones calculated the average gross unit sales price for all Debtor's products sold to determine an average sales price of $22/unit. Morones then subtracted Amazon's fees, costs, and other credits and charges to determine an average price of $5.69/unit paid to Amazon from sales of Debtor's units. Finally, Morones arrives at the net average sales price to Debtor of $16.31. Morones uses the net average sales price of $16.31 as the Replacement Value, which is the amount Amazon is required to pay the Debtor under the FBA Agreement. Amazon does not meaningfully challenge these numbers. The Court finds these average figures are appropriate under the circumstances of this case. These numbers shall be applied by the Court to the units for which Amazon must pay the Plaintiff.
Morones Damages Report, ¶ 32 & Schedule 1b.
Ex. 1, § F-4.
b) Cutoff Date of January 31, 2014
In determining damages, Morones used a cutoff date of January 31, 2014 because the Trustee or his lawyers told her that was when the Parties' business relationship terminated. Amazon terminated the FBA Agreement with the Debtor by sending the Termination Letter. At the end of the Termination Letter, there were instructions to the Debtor, which stated:
Morones' Damages Report, pg. 19 n.37 & Schedule 1.
Termination Letter, Trustee's Trial Exhibit 120.
Please create a Removal Order for your existing Fulfillment by Amazon inventory as soon as possible. Under the FBA Service Terms, we may elect to dispose of these products at your expense if you do not submit a Removal Order within 90 days of the date your account was terminated. Information on how to submit a Removal Order is located in the Create Removal Orders section of Seller Central.
The 90-day period mentioned in the Termination Letter ended on January 20, 2014. The Trustee states an additional 11 days was given to Amazon to complete the requested removals the Debtor made through submission of the Removal Orders, allegedly making the cutoff date to determine damages January 31, 2014.
DE 266, Trustee's Motion for Summary Judgment, pg. 13.
The Trustee's contention that the Termination Letter creates a cutoff date of January 31, 2014, is incorrect. The language of the Termination Letter explicitly states that the Debtor had to submit a Removal Order within 90 days, not that Amazon had 90 days to complete the requested Removals. FBA Agreement, § F-7.2, states, "We may dispose of any Unit we are entitled to dispose of in the matter we prefer." § F-7.1 of the FBA Agreement says, "We may return Units to you for any reason, including upon termination of these Service Terms." Pursuant to the Termination Letter and the FBA Agreement, Amazon was under no obligation to complete the requested Removals within the 90-day period mentioned in the Termination Letter.
In Williams' Rebuttal Report, he states the data which Amazon produced shows that Amazon removed 19,890 units after January 31, 2014. These units, and other transactions occurring after January 31, 2014, are omitted from Morones' calculations and her Damages Report. Before the trial commenced, Morones' January 31, 2014, inventory analysis end point was found by this Court to be improper because Amazon continued to return inventory to Debtor (or to destroy it as the case may be) well after January 31, 2014. Amazon proved to the Court's satisfaction that 19,980 units were returned or destroyed after January 31, 2014. The Court finds that Amazon must be given credit at $ 16.31/unit for the 19,890 units Removed after January 31, 2014. That sum is to be applied against any inventory breach of Contract damages owed by Amazon to Debtor through January 31, 2014. However, as noted below, the Court is not awarding Plaintiff breach of Contract inventory damages for time periods prior to February 1, 2014.
Williams' Rebuttal Report, ¶ 11.
4. Amazon's Contract Breach
The Contract calls for Amazon to care for and keep track of inventory placed in its hands by Debtor. If Amazon fails to do so, it is in breach of the Contract and is liable to Plaintiff for unreimbursed damages. Debtor apparently never challenged Amazon's accounting until the Trustee in December 2012 directed Bellino to make a demand for $1 million of inventory received by Amazon but for which Debtor then claimed Amazon never paid for or returned to Debtor.
However, Azzarelli testified that, prior to the Petition Date, Debtor did ask Amazon questions about Debtor's perception that its inventory in Amazon's hands was reported to be lower than Debtor thought it should be. See Ice Deposition, page 129, lines 1-18.
a) The Settlement Data Was Not Introduced Into Evidence.
It is worth mentioning a few items gleaned by this Court in reviewing Attachment C to Morones' Damages Report. First, the Seller Central Data was produced from frequent, or perhaps daily, reports Amazon created and provided to Debtor over the eight-year course of their relationship. The Settlement Data was a subset of the Seller Central Data and included such information as sales of inventory, payments to Debtor, fees paid to Amazon, etc. This information was made available to Debtor via Amazon's Seller Central website. Amazon did not preserve all the Seller Central Data and the Trustee did not introduce it at trial. This Court finds nowhere in the Contract that Amazon was contractually bound to retain all the Seller Central Data. Had the Trustee introduced all the Settlement Data at trial it presumably could have fully and accurately identified at trial all elements of its claims against Amazon and with precision could have demonstrated the amount of damages it suffered due to Amazon's alleged breach of the Contract. Instead, the Trustee's discovery demands resulted in Amazon producing the M15 Data, a gigantic data dump which Amazon has confirmed to be "the most complete information known to be available to account for the Debtor's units in the Fulfillment by Amazon Program." However, all acknowledged that the M15 Data is not a complete set of the data once generated by the Seller Central Data. The M15 Data, while massive, is nevertheless incapable of fully explaining all the parties' transactions or all of the damages allegedly suffered by Debtor. This is particularly true because the M15 Data does not include sales information. The Trustee knew or should have known of this information gap as it was spelled out in Amazon's September 20, 2016 Discovery Response. The Trustee's counsel also apparently understood that the M15 Data did not include the Settlement Data and the Trustee would need to review Settlement Data to understand all sales transactions. While this Court finds Amazon's September 20, 2016 Discovery Responses adequately alerts the Trustee to the limitations of the M15 Data, Amazon further clarified the M15 Data's limitations when it declared the "M15 [D]ata are the most complete information known to be available to account for [Debtor's] 'Lost Inventory'." Knowing the M15 Data did not account for the Settlement Data, the Trustee could have introduced the Settlement Data into evidence at trial. He did not. As a consequence, this Court cannot conclude what sales proceeds were paid to Debtor by Amazon or what inventory units were transferred by Amazon to Debtor in lieu of cash payments.
DE 378, Bachand's testimony at p. 93:3-12; Trial Day February 19, 2021. Although the Trustee vehemently contends the Debtor and Trustee preserved the Seller Central Data, including the Settlement Data (see DE 396, p. 2, ll. 13-15), Williams' Rebuttal Report notes Morones' Report relied on Settlement Data in her damages analysis, yet her data had certain Data Gaps which Williams largely filled in with information retrieved from Amazon. (See Trial Ex. 5, ¶¶ 19-20 and 49-51. See also § VI(D)(2) above.) The Court need not determine whether Debtor or the Trustee did preserve all this Settlement Central Data as it was never admitted into evidence at trial.
See DE 396, pp. 24-25 at n.55 where the Trustee quotes Ashworth's deposition testimony where he said the M15 Data "was described to us - to us being the trustee and his legal team - as the end-all, be-all record for those transactions [i.e. inventory transactions] . . . and so we had enough data as it was and we struck to the [M15 Data]. The one area that we did not have any other format were the settlement records." (Emphasis supplied.) "Settlement records' are also known as Settlement Data."
See Trial Ex. 118 at Bates page No. 118.0036.
DE 332, Joint Pretrial Statement, p. 7 of 122, ll. 1-2.
The Trustee takes Amazon to task by suggesting "Amazon refuses to respond to discovery directed to differences between M15 Data and Seller Central Data." Trustee cites to three portions of Bachand's testimony in support of this proposition. A close review of Bachand's testimony, however, reveals that she was not refusing (on behalf of Amazon) to explain the differences between the M15 Data and Seller Central Data. Rather, she tried to explain to Trustee's counsel that these documents exist for different reasons and that, in any event, Jeff Moore played a bigger role than her in Amazon's September 20, 2016 Discovery Responses.
DE 396, p. 28 of 71, § 5 Heading.
Id. at nn.56, 57 and 58.
The Trustee points to the FJC Manual for Complex Litigation which "advises that where complex computerized data is at issue, the [identification of computerized data may lead to agreement on a single data base on which all expert and other witnesses will rely in their testimony." In this Court's view, this is exactly what Amazon did to the extent possible. The M15 Data was admittedly the best data set available "to account for Debtor's units" in Amazon's hands "to account for [Debtor's] 'Lost Inventory.'" But the M15 Data did not (and apparently could not) also account for all sales transactions. That was available to all parties via the Settlement Data. Amazon complied with the spirit of the FJC's Manual by agreeing the M15 Data, together with the Settlement Data, supplied the two data sets for the parties' experts to rely upon in their Excel reports and expert testimony.
DE 396, p. 25 of 71; see n.50 citing the Manual For Complex Litigation (Fourth) § 30.2 (2004).
The Trustee's Comments to the Court's Tentative Under Advisement Orderwarns that this Court would commit clear error if it were to find that the Debtor and Trustee failed to preserve the Seller Central Data. Trustee points to testimony from Ashworth and contends the Debtor or Trustee did preserve the Seller Central Data and then twice produced that information on two separate hard drives delivered to Amazon's counsel. It is, however, not at all clear to this Court that the Trustee, through Ashworth and/or Cathy Cameron, preserved all pertinent Settlement Data. Ashworth testified that he downloaded the Settlement Data in pieces and did so from 2013 to 2015. He accomplished these downloads directly from Amazon's Seller Central and not from the Settlement Data earlier downloaded by Debtor's employees. This is problematic because, as Bachand testified at trial:
DE 332, Joint Pretrial Statement at ¶ 4, p. 53 of 122, ¶18, p. 58 of 133, and ¶ 37, pp. 73-4 of 122. See also DE 396, pp. 8-21.
There was concern that one hard drive was corrupted so a second was sent to Amazon's lawyers. Ashworth Deposition at 26:15 - 28:9. See also DE 396, n.46.
Q. So has Amazon ever produced any data in this case that would be adequate to make those determinations?
A. No, I don't think it exists.
Q. It doesn't - you don't think it exists? Would that have been Seller Central data?
A. Both Seller Central and the underlying systems that Seller Central pulls from, there are certain reports that are only existing on a rolling basis. So, by the time the information was requested, they would have already not existed anymore and those are what I would use to do some of the cross-checking.
In other words, the sales data initially input by Amazon into Seller Central to reveal the Settlement Data would not necessarily be the information that remained in Amazon's computer system, even shortly after the initial input, much less years later when Ashworth downloaded the Settlement Data on his two hard drives. For this reason, it is not clear to this Court that the Trustee saved all the Settlement Data. More importantly, the Settlement Data the Trustee or Debtor did maintain and transmit to Amazon was not introduced into evidence at trial. This resulted in a failure of proof of the Trustee's damages because no evidence of sales data (Settlement Data) was introduced at trial.
DE 378, p. 93, ll. 3-12. February 19, 2021, Trial Testimony by Bachand.
b) Plaintiffs Proven Inventory Damages.
Morones' Damages Report contends that, as of January 31, 2014, Debtor's inventory should have contained 39,331 units for a Replacement Value of $641,521. Williams' Rebuttal Report notes that January 31, 2014, is not the correct end date, that 20,405 units were held by Amazon as of March 31, 2015 and, in any event, Morones misapplied inventory adjustment data so that either Amazon owed Debtor nothing or, at most, $137,516.
Morones demonstrates that Amazon's December 31, 2013, ending inventory cannot be correct because, when the January 2014 inventory transactions are factored in, it would leave Debtor with only 16,335 units in Amazon's hands which, of course, cannot be true since Amazon later returned 19,980 units to Debtor. Moreover, the March 31, 2015, M15 Data indicates Amazon held 20,405 units of Debtor's inventory at that date.
Most surprising of all, Exhibits 181 and 182 demonstrate that, after all Removals were effectuated by Amazon, as of March 31, 2015, Amazon's M15 Data still indicated it held 20,405 units of Debtor's product. Debtor had no access to the M15 Data until May 2015 so it could not have known to ask Amazon for Removal of those 20,405 units and, of course, it was by then long out of business so it could not monetize those inventory units.
No evidence was presented to indicate Amazon returned these retained units to Debtor or otherwise compensated Debtor for these units after March 31, 2015. Amazon contends that it indisputably reimbursed Debtor for over $300,000 and that these reimbursements should be applied against the 20,405 units held by Amazon as of March 31, 2015. None of Amazon's reimbursements to Debtor occurred after March 31, 2015. Moreover, Amazon does not contend that Debtor owed it money as of March 31, 2015.While this Court (and the Trustee) recognizes that Amazon reimbursed Debtor over $300,000 over the course of the parties' relationship, none of those reimbursements can be nor will they be applied to the damages sustained by Debtor due to Amazon retaining and not compensating Debtor for the 20,405 units held by Amazon as of March 31, 2015. Again, March 31, 2015, was nearly two years after this Adversary Proceeding commenced.
DE 397, pgs. 1-11.
Amazon's set off and recoupment counterclaims were withdrawn in the Joint Pretrial Statement. See DE 332 at pg. 9 and DE 397 at pg. 5.
The Court finds that Debtor's "Ending Inventory" was 20,405 units. The Court rejects the Ending Inventory calculations of Morones and Williams for the reasons more fully discussed below. The Court further finds Amazon breached the Contract with respect to these units. The Court finds Plaintiff has proven his damage in the amount of the Replacement Value of 20,405 units. Amazon is liable to Debtor for $332,806. By this ruling, the Court is not shifting the Amazon the burden of proving it reimbursed Debtor for units of inventory. The fact that Amazon held 20,405 units of Debtor's inventory as of March 31, 2015, and the fact that all of Amazon's reimbursements to Debtor occurred long before that date and the fact that Debtor owes no money to Amazon necessarily means Amazon owes Debtor for 20,405 units. The parties' Contract compels this result.
20,405 units x $16.31/unit.
What Morones does not convince this Court of in her Declaration, or her testimony is why the Debtor's damage amount should include the full $22/unit gross sales price. $5.69/unit is an amount which Debtor would not be entitled to if these units were fully paid for by a customer because Amazon would be entitled to receive the entire $5.69/unit. Under any scenario, at most, Debtor would be entitled to receive only its due, i.e., $16.31/unit. In effect, Morones is encouraging the Court to punish Amazon by forfeiting its claim to fees and costs totaling $5.69 per unit of loss, damage, etc. While the Court assumes the Trustee would gladly accept such punishment damages, the Court has already denied Trustee's claim for punitive damage. Assessing Amazon for $5.69/unit would not be a measure of Debtor's compensatory damages but, rather, a form of punitive damages. In any event, the Court was not supplied with evidence suggesting Amazon obtained any fees on units not sold. More specifically, the Court received no evidence indicating Amazon sold any of the 20,405 units or was paid a fee of 25% (or any amount) on any of these 20,405 units. The Court rejects Morones' suggestion that Debtor's inventory damages could or should be increased by $5.69/unit.
See DE 91. See also DE 390 at 16.
c) The Court's Analysis of Codes Used by Amazon and the Impact on Claimed Damages.
After discussing her Ending Inventory analysis, Morones' Damages Report addresses the other three components of her damage calculations and then reduces overall damage by $305,611, the amount she finds Amazon Reimbursed back to Debtor. The Court now reviews these components of Morones' damages calculations and discusses the flaws in her findings as well as flaws in Williams' Rebuttal Report.
Morones identifies three categories of Reimbursable Adjustments: Warehouse Damage Adjustments, Lost/Found Adjustments, and Mis-Received Adjustments. Units in the Warehouse Damage Adjustments category were placed there by Morones when she saw Amazon's application of Codes 5, 6, 7, D or E. This methodology is faulty because, for example, a unit damaged by Amazon at its fulfillment center will also likely be assigned another Code or two, like destroyed or paid to Debtor.
Morones reviews a number of Amazon's inventory codes but neither she nor Williams provided data indicating the number of units bearing a given assigned code to this Court.
See the Amazon Code descriptions at Exs. 2 and 147 which is also attached hereto as Attachment 3.
(1) Code M.
Units in Morones' Lost/Found category were placed there because she saw these units had an assigned Code of M (misplaced) or F (found). M and F Codes were offset against one another by Morones because M is a negative unit and F is a positive unit. Again, this methodology is defective because a unit marked M could later be found (F) or sold or destroyed (D), etc. A Code M unit would not necessarily be given a Code F when it is found because it might be a unit whose useful life had expired and would therefore be destroyed (Code D) or sent back to Debtor or a found unit (F) could go straight to a sale and not first logged in as a Code F.
Code M units cannot simply be offset by Code F units to determine Amazon's net liability to Debtor. Trustee contends "… a unit given a Code M is no longer in the virtual inventory, so that unit can't be given more codes." Trustee is incorrect. A Code M unit is still in the "virtual inventory" maintained by Amazon. It just is in the inventory roster as "missing." When that "M" unit is found, it could, contrary to Trustee's contention,be given a different code (e.g., F (found), D (destroyed), O (transferred to another owner), N (reimbursed by a transfer to Debtor's inventory from another owner, etc.). The tally for units coded M does not increase or decrease based on subsequent events. Once an M, always an M. But that M coded unit could be subject to a whole host of other coding events after the unit was first coded M. This is why Morones (and the Trustee) take too simplistic a view by saying one need only take Code M units (102,341 says Morones) and subtract Code F units (63,088) to get a damage suffered by Debtor in the amount of 39,253 units.
DE 396, p. 41, ll. 16-17.
Id., p. 40, ll. 15-17.
Trustee takes aim at Amazon for identifying theoretically possible double-counting scenarios without admitting into evidence any actual units which were double counted. However, the Court has been presented with evidence of units bearing many of the possible double-counting codes (Q, D, N, etc.). Trustee has not and cannot show all the codes ever attributed to a single unit because any given unit is not assigned a series of codes through the life of that unit. Rather, a Code M is given to a unit which Amazon realizes is missing but many codes could later be assigned to that particular unit as events give occasion for another code assignment. That is why each code has a gross number of units assigned that particular code. One cannot, as Trustee suggests, just say M minus F = the number of units for which Amazon is liable to Debtor and that we can ignore all other codes. We cannot ignore other codes because those net M minus F units (38,371) may well be assigned other codes after they receive an "M" coding.
Amazon does not bear the burden of proving which "M" units received a subsequent code or codes. Trustee has the burden of proving its damages. Amazon does not bear the burden of proving Debtor was not damaged. Amazon is not, as Trustee maintains, "strictly liable" for the difference between units coded M and F. Amazon has provided proof of a reasonable likelihood that a material number of "M" coded units could have subsequently been coded with one or more of a number of codes.
DE 396, p. 40 of 71, l. 1.
The Court finds Amazon has demonstrated that it cannot be liable for the entirety of the missing units (M) which were later found (F) and therefore is not "strictly liable" to reimburse Debtor for these 38,371 units. These multiple code designations cast a big shadow upon Morones' Damages Report calculations.
Trustee contends code N units are a "red herring" because the description of Code N was not applied to lost units and in any event were only 171 that would need to be netted against Code O units. Amazon hotly contests Trustee's contention that "during the entire existence of Amazon's relationship with [Debtor] Amazonly only transferred 171 inventory units to [Debtor] as reimbursement for warehouse damaged and lost inventory." (See Trustee's Brief at DE 396, pp. 46 of 71, ll. 11-13). Amazon notes that the "reason those 171 units appear in the adjustments file (Code N) as well as the reimbursement file (Reimbursement Inventory Quantity) is because, as Bachand testified, Amazon in 2013 began transitioning its tracking process for unit reimbursements and Amazon accurately tracked these transactions in both data sets." DE 398, p. 15 of 21, ll. 7-10. Even if the Trustee is right as to application of N, the netting of O against N and the de minimus number of units implied there are many other codes at play, not the least of which are units denoted with Codes D and Q.
(2) Code D.
As to Morones' analysis of Mis-Received Adjustments, the Court notes a given unit can have several coding events while in Amazon's possession. That unit could be booked as a Receipt, but later coded as mis-placed (Code M), then found (Code F), then damaged in Amazon's fulfillment center (Code E), and finally Removed (i.e. sent back to Debtor or destroyed (Code D)). If the total amount of mis-placed inventory is added to the total number of units destroyed there will be some overlap. The extent of that overlap cannot be ascertained through the M15 Data.
Some coded data does not necessarily reveal who is responsible for the changed status of a given unit. A Code D (destroyed unit) could be Debtor's loss (if, for example, it was destroyed because the unit was defective or expired and not removed but destroyed) or could be a loss for which Amazon is responsible (i.e., a unit damaged at Amazon's warehouse). While units coded D may be "the end of the line," that may not be the beginning of the line for that unit. For example, that D unit could also bear an earlier Code M (missing), F (found), H (damage via customer return), K (damaged as a result of item defect), U (damaged by merchant) or 6 (damaged by inbound carrier). Trustee has not carried his burden of proving the 17,924 units coded D must all (or even mostly) be paid for by Amazon.
See DE 310, Williams' December 21, 2020, Declaration at page 6, ¶ 16e which further discusses this scenario.
DE 396, p. 46, l. 5.
See DE 396 pp. 45-48
Morones did not account for Code N units. The M15 Data indicates there were 16,585 Code N units. The Court finds the flaws in Morones' Damages Report calls into question the validity of her Reimbursable inventory adjustments and Unpaid Refund Reimbursements.
Williams' Declaration at DE 310 also discussed Code 6 adjustments (damage by inbound carrier). Note that a Code 6 damage could be an Amazon responsibility (if Amazon's carrier was used) or a Debtor responsibility (if a non-Amazon carrier was used). This further highlights the problem with using a given code to universally lay the damage responsibility on one party or the other.
Based on these problems with Morones' analysis of these three "Reimbursable Adjustments," the Court finds her damage calculations unreliable and unpersuasive. The Court finds Plaintiff has not sustained its burden of proving "Reimbursement Adjustment" damages by a preponderance of the evidence.
(3) Code Q.
Morones' Damages Report does not mention Code Q related damages and her Declaration does not explicitly opine as to whether Code Q units, sellable or otherwise, are properly counted as damages to Debtor's inventory. Although Trustee's expert did not point to Debtor being entitled to damages based on units coded "Q," her trial testimony did discuss Code Q damages and now the Trustee argues he is entitled to recover damages for unsellable 84,000 Code Q units. A total of 150,092 to 166,279 units were given a Code Q designation. From testimony by Bachand, Morones and Williams, the Court concludes Amazon heavily used Code Q as a dumping ground when an Amazon employee could not find anything else to do within a given unit. Over 40% of all units given a Code Q were assigned that code within one month, July 2013.
DE 396, pp. 48-58.
DE 396, p. 49. See also DE 396 at p. 59, ll. 4-5.
DE 396, p. 59, l. 5. 67,920 ÷ 166,279.
Cross-examination of Morones demonstrated that Code Q units could also be designated as Code D (destroyed) or Code E (damaged at Amazon fulfillment center) and, therefore, would be valueless. This, therefore, calls into serious question whether the 147,968 Code Q units referred to as "sellable" by Morones are indeed sellable. This Court rejects Trustee's blanket demand that Amazon be held liable for any sellable units which have been given a Code Q designation. First, units given a Code Q will also be given other codes and are, in such cases, duplicative codes for the same unit. Bachand testified that all Code Q units will also be coded as D (destroyed), M (missing), 5 (stolen/theft) or would be removed to the seller.
Bachand Trial testimony (DE 364), February 18, 2021, p. 113, 1. 3- p. 114, l. 14.
Bachand acknowledges that during Debtor's eight-year relationship with Amazon, Code Q was poorly defined. In the Inventory Adjustments section of Amazon's Help section (part of Amazon's "Program Policies" incorporated into the parties' Contract, Code Q is identified as "damaged-miscellaneous" and is defined as "[a] decrease of your sellable inventory when damages cannot be attributed to a source." Trustee reads this to mean if there is a sellable unit that is given a Code Q, Amazon owes the Debtor for that unit because the Contract notes (and even Bachand and Williams agree) that where Amazon cannot put their finger on who or what caused harm to Debtor's units, Amazon will be responsible for payment to Debtor. However, this Court is persuaded by Bachand's testimony to the effect that, at all times, Amazon consistently managed Code Q units in a manner that would always result in at least one of four other codes being applied to a Code Q unit. Because all Code Q units ultimately had a different disposition (D, M, 5 or Removal), Code Q cannot be a reliable measure of damage to a unit for which Amazon is contractually liable to Debtor. Application of one or more of these four codes to a given unit may give rise to a damage claim but coding a unit with a "Q" cannot be relied upon as an accurate measure of Debtor's damage, even if the unit remained sellable. This fact will affect the damage count. More importantly, Code Q is a code for which Amazon is not necessarily liable because all Code Q units also bear another code designation. For example, Amazon contends 86 to 91% of the units coded Q were products with expiration dates suggesting that expired units may have initially been coded Q and later destroyed (D) or removed to the Debtor. To grant the Trustee damages for all Code Q units, even just as to sellable units (assuming the number of "sellable units" could be accurately ascertained), would be contrary to the terms of the parties' Contract. The Court finds Morones' measure of Code Q damages at $2,413,470 is unpersuasive. The Court finds Plaintiff has not carried his burden of proof on this measure of Plaintiffs inventory damages. The Court rejects Debtor's claim that all sellable units bearing a Code Q should be paid to Plaintiff at the rate of $16.31/unit.
Id. at 116, l. 6
Trial Exs. 2 and 147.
Bachand Trial testimony (DE 364), February 18, 2021; Williams' Rebuttal Report, Ex. 5 at Bates page No. 5.0020, n.43.
See Bachand testimony.
DE 398, p. 16 of 21, ll. 8-9. In this, Amazon cites to Trial Ex. 131.
147 974 x $16.31/unit.
Bachand contends Code Q signals a unit designation for which Amazon cannot be held responsible. The Court is not finding this to be so because Ice testified Code Q signified that the seller [Debtor] or Amazon or someone else could have caused damage to a unit assigned a Code Q. See Ice Deposition at page 129, lines 1-18. In any event, since Q units will later bear another code designation, Amazon may or may not ultimately be liable for damages pertinent to a Code Q unit.
d) Unpaid Refunds.
Morones next discusses damages identified as "Unpaid Refund Reimbursements." When discussing her measure of "Unpaid Refund Reimbursements," Morones' cross- examination revealed that some (maybe many) purchase refunds were paid to customers because the products they received were defective, damaged, or expired. A customer would not necessarily return such units to Amazon and, if they did, the unit would be valueless to Debtor as it could not be resold. Therefore, Debtor would not be damaged by receiving no refund Reimbursement from Amazon in this instance. This calls into question the entire $186,247 of damages Morones tallied up as "Unpaid Refund Reimbursements." The Court finds Plaintiff has not carried its burden of proof on the "Unpaid Refund Reimbursements" portion of his claimed inventory damages.
In any event, this $186,247 amount referenced in Morones' Declaration corrects her Damages Report which said this amount of damages totaled $197,721 per ¶ 9 of Ex. 172 or $237,706 per ¶ 11 of Ex. 172.
e) Sales Proceeds Not Remitted to Debtor.
Morones' final damage category is sales proceeds not remitted to Debtor. Sales proceeds not remitted are acknowledge by Morones as not totaling $1,156,495 as first reported by her if the Court finds the Data Gaps have been closed by data authenticated by Amazon. The Court finds Amazon did authenticate no less than 87% of the Data Gaps. Given this finding, Morones conceded her damage calculations attributable to sales proceeds not remitted to Debtor by Amazon total $172,851. As to remaining $172,851 identified by Morones as Sales Proceeds not remitted to Debtor, Williams contends reliance upon the M15 Data is misplaced because it does not address the financial transactions of the parties and that Morones has not provided reliable evidence as to how many units were lost or damaged or not otherwise accounted for in Ending Inventory or Code N adjustments. The Court agrees with Williams and finds Morones' Damages Report does not reliably or sufficiently substantiate Plaintiffs damages for sales proceeds unpaid to Debtor. Failing in his proof on this issue, the Court finds Plaintiffs demand for damages of $172,851 for Sales Proceeds Not Remitted to the Debtor are not proven by a preponderance of the evidence.
Ex. 172, table at page 6.
f) Trustee Did Not Carry Its Burden of Proof on the Basis of Amazon's Motivation for Termination of the Contract.
Through (1) testimony by Azzarelli to the effect that "lots" of Debtor's inventory was unaccounted for by Amazon and (2) testimony by Shaffer to the effect that Debtor brought this discrepancy to Amazon's attention and then found its access to the amazon.com platform was terminated, the Trustee suggests Amazon terminated the Contract because it was retaliating against Debtor. While the Trustee may well be justified in his suspicions, this Court finds Plaintiff did not prove by a preponderance of the evidence that Amazon's termination was linked to the Debtor challenging Amazon's inventory data or methods. This Court lifted the Bankruptcy Stay to enable Amazon's termination because Amazon satisfied the Court that Debtor was in breach of the Contract in 2013. This finding was not contradicted by evidence at trial. The Court will not now find Amazon breached the Contract by basing its platform access termination upon Debtor's challenge to Amazon's accountings or inventory control methods.
g) Unclean Hands.
In its Answer and through some of the evidence presented at trial, Amazon suggests Debtor has unclean hands and the Trustee should be denied recovery in this case due to Debtor's allegedly unclean hands. For example, Amazon made much of the fact that the Trustee in December 2012 asked Bellino to make a $1 million demand on Amazon for claimed inventory shortages, but Bellino instead made a demand for $10.5 million. This Court's decision does not stand on the veracity or character of Bellino so this Court makes no findings as to whether his testimony is credible or not. Neither does this Court find Amazon sustained its burden of proving its unclean hands defense.
h) Money Claimed Owing by Debtor to Amazon.
To the extent Williams' Rebuttal Report suggests the M15 Data and/or Settlement Data indicate Debtor owed Amazon anything as of March 31, 2015, this Court finds Amazon withdrew its counterclaims including counterclaims for setoffs or recoupment. The data is confusing, yes, but it is also not as conclusive as Williams would have this Court find. Both the M15 Data and the Settlement Data are incomplete data sets. Even if Amazon did assert claims against Debtor, such claims could not be fairly ascertained by this Court from the data provided.
5. Conclusion on Breach of Contract Claims
The Contract calls for Amazon to pay Debtor for lost, destroyed, and unaccounted for units. Amazon breached the Contract by failing to fully live up to this duty with respect to the 20,405 units of Debtor's inventory in Amazon's hands as of March 31, 2015. Amazon is liable for this Contract breach for damages in the amount of $332,806 plus interest at 12% from April 1, 2015, until paid. All other inventory related damages sought by Plaintiff are hereby denied as Plaintiff failed to sustain his burden of proof with respect to such additional claimed damages. To the extent Morones recognized and Williams confirms that Amazon reimbursed Debtor through the course of their relationship ($305,611 says Morones), this Court finds those Reimbursements were fully accounted for long before March 31, 2015, through the morass of coded data supplied in the Seller Central Data (including the Settlement Data) and are offset by all but the remaining 20,405 units. In other words, Amazon's Reimbursements to Debtor cannot now be offset against or recouped from the $332,806 damages awarded in this Order.
D. Prejudgment Interest on Damages
The Trustee seeks $2,261,347 in prejudgment interest. Under Washington law, if a Plaintiff is entitled to pre-judgment interest, that interest accrues at the rate of 12% per annum simple (not compounded) interest. Morones was instructed by the Trustee to apply prejudgment interest to the inventory damages at 12% simple interest. Prejudgment interest was calculated from January 11, 2011, the mid-point of the Debtor's involvement in the FBA Program, through May 10, 2019, the date of Morones Damages Report.
Morones Expert Report, Schedule 1.
Morones Expert Report, ¶ 54.
Morones assumed a beginning date of January 1, 2008, and an end date of January 31, 2014. Morones Expert Report, ¶ 54.
Morones Expert Report, ¶ 54.
a) Legal Analysis
When addressing the question of Plaintiffs claim to pre-judgment interest this Court must look to Washington law. Washington courts generally favor prejudgment interest based on the premise that a party that retains money it should have paid to another should be charged interest. Awarding pre-judgment interest compels a party that wrongfully holds money to disgorge the benefit. It may be safely said that the tendency has been in favor of allowing interest rather than against it, and that the degree of certainty or ease with which the approximate amount can be ascertained has grown less and less stringent.
Rekhter, 323 P.3d at 1050 (quoting Pierce County v. State, 185 P.3d 594 (Wash.Ct.App. 2008)).
Rekhter, 323 P.3d at 1050 (quoting Mahler v. Szucs, 957 P.2d 632 (Wash. 1998)).
Rekhter, 323 P.3d at 1050 (quoting Prier v. Refrigeration Eng'g Co., 442 P.2d 621 (Wash. 1968)).
Prejudgment interest can be awarded if the claim upon which recovery is based is "liquidated." A claim is "liquidated" where the evidence furnishes data which, if believed, makes it possible to compute the amount with exactness, without reliance on opinion or discretion. The rationale for this rule is that it would be unfair to hold a defendant accountable for interest on an amount that is unquantifiable and unforeseeable prior to a jury verdict.
Car Wash Enters., Inc. v. Kampanos, 874 P.2d 868, 875 (Wash.Ct.App. 1994) (citing Hansen v. Rothaus, 730 P.2d 662 (Wash. 1986)). See also OTR Wheel Engineering, Inc. v. West Worldwide Services, Inc., 743 Fed.Appx. 771 (2018) (9th Cir. Memorandum Decision).
Rekhter v. State, Dep't of Soc. & Health Servs., 323 P.3d 1036, 1047 (Wash. 2014).
A claim is unliquidated if the factfinder must exercise discretion to determine the measure of damages. The fact that an amount is disputed does not render the amount unliquidated. A claim may be liquidated even if a dispute exists over all or part of the claim. It is the character of the original claim, rather than the court's ultimate method for awarding damages, that determines whether prejudgment interest is allowable.
Aker Verdal A/S v. Lampson, Inc., 828 P.2d 610 (Wash.Ct.App. 1992).
25 David K. DeWolf et al., Washington Practice Series, Contract Law And Practice § 14:14 (3d ed. 2020) (citing Spradlin Rock Products, Inc. v. Public Utility Dist. No. 1 of Grays Harbor County, 266 P.3d 229 (Wash.Ct.App. 2011) (sufficient evidence supported contractor's claim for lost profits resulting from public utility district's alleged breach of contract); Forbes v. American Bldg. Maintenance Co. West, 198 P.3d 1042 (Wash.Ct.App. 2009), judgment affd in part, rev'd in part on other grounds, 240 P.3d 790 (Wash. 2010) (trial court properly awarded prejudgment interest on attorney's fees claim even though parties disputed method of calculation); Polygon Northwest Co. v. American Nat. Fire Ins. Co., 189 P.3d 777 (Wash.Ct.App. 2008) (trial court properly applied prejudgment interest to equitable indemnity claim for settlement even though formula for allocation was disputed); Bostain v. Food Exp., Inc., 153 P.3d 846 (Wash. 2007) (trial court properly awarded prejudgment interest on claim for overtime wages wrongly withheld); Scoccolo Const, Inc. ex rel. Curb One, Inc. v. City of Renton, 145 P.3d 371 (Wash. 2006) (claim was liquidated even though defendant successfully challenged portions of plaintiffs damage claim); Aker Verdal A/S v. Neil F. Lampson, Inc., 828 P.2d 610 (Wash.Ct.App. 1992); Pederson's Fryer Farms, Inc. v. Transamerica Ins. Co., 922 P.2d 126 (Wash.Ct.App. 1996) (since the character of the underlying claim-in this case the cost of pollution cleanup-was liquidated, prejudgment interest was appropriate)).
Spradlin Rock Products, Inc. v. Public Utility Dist. No. 1 of Grays Harbor County, 266 P.3d 229 (Wash.Ct.App. 2011) (citing Prier v. Refrigeration Eng'g Co., 442 P.2d 621 (Wash. 1968), rev. den 'd, 863 P.2d 1352 (Wash. 1993)).
The 20,405 units held by Amazon on March 31, 2015, appear to be a matter of fact not in dispute between the parties yet Morones did not focus on this amount as she was of the belief that Plaintiffs damages were much larger and could be ascertained from a different view of the data. Williams, on the other hand, mentions these 20,405 units more to demonstrate Morones' assumptions and methodology were faulty as opposed to conceding Amazon must pay Plaintiff for these units. The complexity of the data, the shortcomings in the data and the hot contest over what the data could and could not prove are all givens in this case. That said, the M15 Data demonstrated that 20,405 units of Debtor's inventory were held by Amazon as of March 31, 2015. Plaintiffs inventory damages of $332,806 were liquidated damages upon which Plaintiff shall be awarded prejudgment interest. Plaintiffs prejudgment interest on its $332,806 damage award shall run from April 1, 2015, at the rate of 12% (Washington statutory rate), until paid.
b) Conclusion on Prejudgment Interest
Plaintiffs inventory damages awarded by this Court were liquidated as of March 31, 2015. The Trustee previously stipulated to his damages being approximate and that Plaintiffs evidence is admittedly not precise. Nevertheless, this Court has fixed Plaintiffs inventory damages at $332,806 based on the unaccounted-for inventory of 20,405 units. This amount was quantified and foreseeable before trial. The Court finds Plaintiffs inventory damages of $332,806 were liquidated so Plaintiff is entitled to prejudgment interest on such amounts from April 1, 2015, at 12% simple interest per annum.
DE 370, Trustee's Brief on Burden of Proof, pg. 7, lines 21-22, "The trustee met his burden of proof with admissible evidence showing at least the approximate amount of missing or damaged inventory." (Emphasis added). See also pg. 6, lines 11-13, "The Court should thus place the burden on Amazon to prove the precise amount of inventory that was lost, damaged or destroyed, and if Amazon can't do so, then the Court should accept the trustee's evidence of damages, even if it's only approximate."
DE 370, pg. 3, line 27, pg. 4, lines 1-4, "Although [Debtor] has presented admissible evidence from which damages can be quantified, the evidence is admittedly neither perfect nor precise. Sometimes evidence seems in conflict, the effect of some facts remains unclear, and some facts simply don't 'add up,' both literally and metaphorically."
VIII. SUMMARY OF CONCLUSIONS
This Adversary Proceeding concerns the Trustee's claims against Amazon (1) for breach of the Contract and turnover of the damages owed by Amazon and (2) for stay violation damages. Amazon's stay violation itself was ascertained by the Court long before trial so the trial was simply a matter of trying Plaintiffs claimed damages. The Court rejects Amazon's efforts to contain the stay violation damages to the 15 days when Amazon was violating the Bankruptcy Stay. Amazon's harm to Debtor reverberated throughout Debtor's business right until the day its business was effectively obliterated by Amazon's ultimately lawful termination of Debtor's access to amazon.com. While the Court agrees with Morones view that the stay violation damages continued through October 22, 2013, the Court disagrees that Debtor's correct financial baseline was established between November 2012 and March 2013 or that Debtor's sales could be fairly projected to grow after April 2013. The Court awards Plaintiff stay violation damages of $668,484 plus interest from the entry of this Court's judgment, until paid.
Plaintiffs inventory damage claims were handicapped by inadequacies in both the M15 Data and available Seller Central Data or Settlement Data. Plaintiff blames Amazon for not providing all the data from which damages could be fully and clearly established. However, the Trustee did not introduce at trial all the Seller Central Data which Debtor had been supplied by Amazon over the course of this eight-year relationship or even the Settlement Data which reflected sales transactions concerning Debtor's inventory. Plaintiff contends all this data was preserved by Plaintiff, yet it was not introduced into evidence. Moreover, Plaintiff has not shown this Court that Amazon had a duty to preserve what unquestionably would be a colossal data set, especially where that data was not questioned by Plaintiff until years into their business relationship. The burden of proof on Amazon's Contract breach and the burden of proving Plaintiffs damages will not be shifted to Amazon. Plaintiff was required to prove by a preponderance of the evidence both Amazon's breach and Plaintiffs resulting damages.
At the end of the day this Court finds that, despite the game efforts of his damages expert Morones, Plaintiff failed to prove by a preponderance of the evidence that Amazon breached the parties' Contract beyond Amazon's failure to compensate Plaintiff for the 20,405 units held by Amazon as of March 31, 2015. Since Plaintiffs breach of Contract claims and damages were liquidated as of March 31, 2015, Plaintiff is entitled to prejudgment interest on Plaintiffs inventory damages of $332,806 at 12% per annum from April 1, 2015, until paid.
Plaintiffs attorneys are directed to lodge a form of judgment consistent with this Order.
The following terms either have been defined by the parties in this litigation or are described in the manner in which they are used in the Court's Under Advisement Order:
ABSA: Amazon Services Business Solutions Agreement. The ABSA is also referenced as the FBA Agreement.
Admin. DE: Docket entries in the Debtor's bankruptcy filed with the District of Arizona at Case NO. 2:11-bk-28944-DPC.
Adversary Proceeding: Case No. 2:13-ap-00799-DPC filed in the chapter 11 bankruptcy of the Debtor.
Adjustments: Inventory record changes by Amazon when an event impacted Debtor's inventory in Amazon's hands.
Adjustment Codes: Codes that describe inventory events impacting a seller's inventory account, such as damaged, missing, found, and transferred. See Trial Ex. 2.
Affirmative Expert Report: Serena Morones' Affirmative Expert Report dated May 10, 2019.
AFTT: Amazon Fulfillment Technology Team.
Aged Inventory: Inventory that has not been sold within 90 days. Amazon: Amazon Services, LLC
Amazon's September 20, 2016 Discovery Response: See Trial Ex. 118, in particular at Bates No. 118.0036.
Answer: The Amended Answer filed by Amazon at DE 5 in response to Plaintiffs Complaint.
Ashworth: Stephen Ashworth, a consulting expert witness employed by the Trustee.
ASIN: Amazon Standard Identification Number.
ASIN Merge: The combination of two or more ASIN's.
Azzarelli: Thomas Azzarelli, Debtor's one-time Chief Financial Officer.
Bachand: Tasha Bachand, an Amazon employee.
Bankruptcy Case or Bankruptcy Proceeding: Debtor's administrative bankruptcy proceeding, 2:11-bk-28944-DPC.
Bankruptcy Code: U.S. Bankruptcy Code, 11 U.S.C. §§ 101-1532.
Bellino: Daniel A. Bellino, a founder of Debtor.
Code 1: Indicates software correction of inventory discrepancies.
Code 2: Indicates software correction of inventory discrepancies.
Code 3: Indicates product redefinition and transfer in from original inventory item.
Code 4: Indicates product redefinition and transfer out to new inventory item.
Code 5: Indicates unrecoverable inventory.
Code 6: Indicates unit damaged by inbound carrier.
Code D: Indicates unit is destroyed.
Code E: Indicates unit is damaged at Amazon fulfillment center.
Code F: Indicates unit is found.
Code H: Indicates unit is damaged - customer return.
Code J: Indicates software correction of inventory discrepancies.
Code K: Indicates unit is damaged as result of item defect.
Code M: Indicates unit is misplaced.
Code N: Indicates receipt of unit from another owner.
Code O: Indicates transfer of unit to another owner.
Code P: Indicates unsellable inventory.
Code Q: Signifies damage - miscellaneous.
Code U: Indicates unit damaged by merchant.
Code X: Indicates correction for inbound shipment receiving discrepancies.
Complaint: Plaintiffs Complaint filed in the Adversary Proceeding on July 9, 2013.
Cone: Jeffrey Cone, a consulting expert witness employed by the Trustee.
Contract: Ex. 2. The Merchants@Amazon.com Program Agreement and the Amazon Services Business Solutions Agreement collectively referred to as the Contract.
Creditors Committee: Official Committee of Unsecured Creditors appointed by the United States Trustee on November 9, 2011.
Cutoff Date: January 31, 2014-the date after which Morones disregarded M15 Data at the direction of the Trustee's counsel.
CSV: Comma Separated Values.
DAB: An alternative name of the Debtor. The initials of David A. Bellino.
Data Gaps: Time periods within the Settlement Reports produced by the Trustee where no Debtor activity was included. The identified gaps are at least six 48-hour periods that occurred in January and March 2010.
DE: Docket entry in the Adversary Proceeding.
Debtor: Potential Dynamix, LLC
Deep Dive: See testimony of Justin Ice at page 61 of the Order.
Defendant: Amazon Services, LLC
Dewberry: Dustin Dewberry, an Amazon employee.
Disclosure Statement: The Disclosure Statement filed at Admin. DE 219.
Dot Missing SKU: Debtor incorrectly formatted SKUs and there were "dot missing SKUs." A "Dot Missing SKU" happened when Amazon would just choose a SKU for inventory it received if it did not know which SKU it belonged to.
Ex.: Trial Exhibits.
Expected Ending Inventory: The Debtor's inventory expected to be remaining in Amazon's possession at the end of the Debtor's relationship with Amazon.
FBA: Fulfillment By Amazon, which is an Amazon service in which Debtor paid Amazon to receive its inventory and pick, pack, ship, and provide customer service for that inventory.
FBA Agreement: Also known as the ABSA.
FBA Program: Amazon's Fulfillment by Amazon Program
First Motion in Limine: Trustee's Motion in Limine to Preclude the Admission of "Deep Dive" Documents into Evidence.
FJC: The Federal Judicial Center.
Flores: Diana Flores, an Amazon employee.
FNSKU: Fulfillment Network Stock Keeping Unit.
Ice: Justin Ice, a former employee of Amazon.
Joint Pretrial Statement: Joint Pretrial Statement: DE 332.
Lawcock: Sean Lawcock, Debtor's former inventory manager.
Limitations Clause: ¶ 8, page 5 of 28, of the Contract (Ex. 1).
Lost/Found Adjustment: As used in Morones' Damages Report this term represents inventory identified in the M15 Data as misplaced (Code M) or found (Code F).
M15 Data: Inventory transaction data that Amazon produced in May 2015. Amazon described the M15 Data as follows: "Amazon has made reasonable efforts to identify and produce in the May 2015 Production transaction-level data for the Debtor's Units in the Fulfillment By Amazon Program but ... the files may not be complete or accurate with respect to every Debtor Unit. The May 2015 Production consists of multiple data streams compiled across multiple Amazon teams and also likely includes some erroneous entries as a result of employee transcription error. During the period in which the Debtor was selling its products through Amazon's website, not all inventory in the currently used data warehouse. There may be incomplete information in the May 2015 Production from transcribing older inventory tracking resources to the currently used system. Nonetheless, the May 2015 Production is the most complete information known to be available to account for the Debtor's Units in the Fulfillment By Amazon Program."
May 2015 Data Production: Also known as M15 Data.
Merchant Program: Amazon Merchant Fulfilled Network
Moore: Jeff Moore, an Amazon employee.
Morones: Serena Morones, Plaintiffs damages expert.
Morones' Damages Report: Ex. 7, Serena Morones' Affirmative Expert Report dated May 10, 2019.
Morones Declaration: Ex. 172.
Net Average Price per Unit: The unit price Morones developed and applied in her report, defined as follows: "I calculated the average gross unit sales price for all PD products sold, and then subtracted Amazon's fees, costs, and other credits and charges, to arrive at the net average sales price of $16.31."
Net Sales: Customer Sales, less Customer Returns.
OMX: An Amazon inventory tracking system. See Reilly's deposition testimony.
Order: The Under Advisement Order of this Court entered at DE__.
Payment Report: See Settlement Data, Settlement Report and Settlement Record.
Peeples: Josh Peeples, an employee of Debtor.
Petition Date: October 13, 2011.
Plaintiff: Timothy H. Shaffer, Chapter 11 Trustee for Debtor.
Plan: Joint Plan of Reorganization filed at Admin. DE 217.
Reason Codes: A letter or number assigned to identify Adjustments. For example, Code "M" represents a missing inventory unit, and Code "F" represents a found inventory unit.
Receipt: Inventory units shipped from Debtor to Amazon and received into Amazon's fulfillment centers. A Receipt increases Debtor's inventory balance.
Refund Reimbursements: Payments or inventory transfers to compensate Debtor for certain customer Returns. If a customer initiates a Return but does not return the full item or if Amazon is responsible for the reason generating the return, Amazon issues a refund reimbursement.
Reilly: Thomas Reilly, Debtor's former Chief Operating Officer.
Reimbursements: Compensation provided by Amazon to Debtor, either in the form of payment or inventory replacement from Amazon.
Removed or Removal: At Debtor's request, the act of Amazon either sending inventory units back to Debtor or destroying inventory units (typically so that Debtor does not have to pay shipping or long-term storage fees). Debtor may choose to have certain Removals automated and they may manually request Removals. Removals decrease Debtor's inventory balance.
Return: The reversal of a customer sale, upon the request of a refund by a customer. Returns increase Debtor's inventory balance.
Sales: Inventory units shipped by Amazon to a customer after that customer purchases the inventory from Debtor. Sales reduce Debtor's inventory balance.
Sales Proceeds Not Remitted: Net proceeds from a sale by the Debtor that Morones asserted had not been remitted to the Debtor.
Schmidt: Matthew Schmidt, a co-founder of Debtor.
Second Motion in Limine: Amazon's Motion in Limine to Exclude Unpleaded and Untimely Claims and Damages Theories
Seller Central: A web-based interface displaying for sellers inventory reports and information about their participation in the FBA program.
Seller Central Data: The information disseminated on Seller Central is referred to as the Seller Central Data. The Trustee notes this is a "very broad term that encompasses many reports Amazon made available to FBA Sellers through the Seller Central portal to track and management [sic] their inventory." Settlement Data is but one of the reports contained in the Seller Central Data.
DE 396, p. 5, n.14.
Settlement Data: The information disseminated on Seller Central reflecting "payments to FBA Sellers." The Trustee appears to use the term "Settlement Data" interchangeably with "Settlement Report," Settlement Record" and "Payment Report."
DE 396, p. 5, n.14.
See DE 396, p. 17: 1-3 and ns.38 and 39.
Settlement Records: See Settlement Data.
Settlement Reports: See Settlement Data.
Shacklock: Susan Shacklock, Debtor's accounting manager.
SKU: Stock-Keeping Units.
Soder: Eric Soder, a former Amazon employee.
State Court: Arizona Superior Court, Maricopa County
Termination Date: October 22, 20
13. Third Motion in Limine: Amazon's Motion in Limine to Exclude the Testimony of Expert Serena Morones
Trustee: Timothy H. Shaffer, Chapter 11 Trustee of Debtor.
UST: United States Trustee
Unaccounted Inventory: Debtor's inventory which Amazon has not accounted to Debtor.
Unpaid Refund Reimbursements: Referenced reimbursements due Debtor but which have not been paid by Amazon or for which Amazon has not transferred product to Debtor.
VHD: Virtual hard drives.
Warehouse Damage Adjustments: As used in the Morones' Damages Report, this term represents inventory identified in the M15 Data as damaged pursuant to the following reason codes: 5, 6, 7, D or E.
Williams: E. Weiant Williams, Amazon's rebuttal expert.
Williams' Rebuttal Report: Trial Ex. 5, dated September 14, 2020.