CITATION: Nortel Networks Corporation (Re), 2015 ONSC 2987

COURT FILE NO.: 09-CL-7950

DATE: 20150512

SUPERIOR COURT OF JUSTICE – ONTARIO

COMMERCIAL LIST

IN THE MATTER OF THE COMPANIES' CREDITORS ARRANGEMENT ACT, R.S.C. 1985, c. c-36, AS AMENDED

AND IN THE MATTER OF A PLAN OF COMPROMISE OR ARRANGEMENT OF NORTEL NETWORKS CORPORATION, NORTEL NETWORKS LIMITED, NORTEL NETWORKS GLOBAL CORPORATION, NORTEL NETWORKS INTERNATIONAL CORPORATION and NORTEL NETWORKS TECHNOLOGY CORPORATION

APPLICATION UNDER THE COMPANIES' CREDITORS ARRANGEMENT ACT, R.S.C. 1985, c. C-36, AS AMENDED

BEFORE: Newbould J.

HEARD: May 12, 13, 14, 15, 20, 21, 22, 27, 28, 29, 30, June 2, 5, 6, 16, 17, 18, 19, 20, 23, 24, September 22, 23 and 24, 2014

COUNSEL:

Benjamin Zarnett Peter Ruby Jessica Kimmel Graham D. Smith Alan Mark Julie Rosenthal Joseph Pasquariello Jennifer Stam Ken Coleman Jacob Pultman Paul Keller Laura Hall

Counsel for the Monitor and Canadian Debtors

R. Paul Steep Elder C. Marques Byron Shaw Kenneth T. Rosenberg Lily Harmer Max Starnino Karen Jones Megan Shortreed Mark Zigler Ari Kaplan Jeff Van Bakel Barbara Walancik

Counsel for the Canadian Creditors’ Committee

Sheila Block

Andrew Gray

Scott A. Bomhof

Avi Luft

James Bromley

Lisa Schweitzer

Jeffrey Rosenthal

Howard Zelbo

Counsel for the U.S. Debtors

Richard Swan Tom Matz Jonathan Bell Gavin H. Finlayson Kevin J. Zych Andrew LeBlanc Nick Bassett

Counsel for the Ad Hoc Group of Bondholders

Matthew P. Gottlieb

Matthew Milne-Smith

James Doris

William Maguire

Neil Oxford

John Whiteoak

Tracy L. Wynne

Derek Adler

Counsel for the EMEA Debtors

Michael E. Barrack

D.J. Miller

John L. Finnigan

Andrea McEwan

Rebecca (Lewis) Kennedy

Michael Shakra

Brian O’Connor Eugene Chang

Counsel for the UKPC

R. Shayne Kukulowicz

Goff Shaw

Abid Qureshi

David H. Botter

Counsel for the U.S. Unsecured Creditors’ Committee

Kenneth David Kraft

Karen B. Dine

John Salmas

David Crichlow

Counsel for Wilmington Trust, National Association, Trustee

Brett Harrison

Counsel for The Bank of New York Mellon, Trustee

John D. Marshall

Counsel for Law Debenture Trust Company of New York, Trustee

r e a s o n s f o r j u d g m e n t

T A B L E O F C O N T E N T S

pAGE nO.

Prologue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Nortel history and its matrix structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Bankruptcy filings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Decision to liquidate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Interim Funding and Settlement Agreement (“IFSA”) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Asset sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Position of the parties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The MRDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (i) Governing law of the construction of the contract . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) Position of the parties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a) The meaning of the exclusive license . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b) The right to sue for infringement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iv) Surrounding circumstances or the factual matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . a) 1996 APA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b) Sub-licences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . c) Representations to tax authorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . d) Avoiding permanent establishment status in the U.S. . . . . . . . . . . . . . . . . e) Patent litigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . f) Conclusions of factual matrix evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . (v) Commercial reasonableness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (vi) Conclusions on the meaning of the MRDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17 18 23 24 24 37 38 40 42 44 48 49 49 49 52

Applicability of the MRDA to the allocation issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

EMEA position on ownership of the Nortel IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appropriate method to allocate the proceeds of sale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appropriate pro rata allocation method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Allocation on a basis other than pro rata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Epilogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

APPENDIX A. . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allocation of the proceeds of the line of business sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (i) Mr. Kinrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) Mr. Malackowski and Mr. Huffard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) Mr. Green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 4 9

APPENDIX B . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Residual IP proceeds allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (i) Mr. Kinrich’s license approach to value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) Mr. Malackowski’s contribution approach to value . . . . . . . . . . . . . . . . . . . . . (iii) Mr. Green’s approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 1 18 25

Prologue

[1] Until January 14, 2009, Nortel Networks Corporation (“NNC”) was a publicly-traded Canadian company and the direct or indirect parent of more than 130 subsidiaries located in more than 100 countries, collectively known as the “Nortel Group” or “Nortel”. It operated a global networking solutions and telecommunications business.

[2] On January 14, 2009 most of the Nortel entities filed for bankruptcy protection. In Canada, the Canadian incorporated entities (the “Canadian Debtors”) filed under the Companies’ Creditors Arrangement Act (“CCAA”). In the United States, most of the U.S. incorporated entities (the “U.S. Debtors”) filed under chapter 11 of the U.S. Bankruptcy Code. In England, most of the entities incorporated in Europe, the Middle East and Africa (the “EMEA[1] Debtors”) were granted administration orders under the UK Insolvency Act, 1986.

[3] The initial intent of Nortel was to downsize and carry on those portions of its telecommunications business that it thought could be profitable. However that plan quickly evaporated and in June, 2009 Nortel decided to liquidate its assets. It sold its business lines for approximately $3.285[2] billion of which approximately $2.85 billion is now available to be allocated. It then sold its residual intellectual property for $4.5 billion. These amounts totalling $7.3 billion are held in escrow (the “lockbox funds”). At issue in these proceedings is how to allocate the $7.3 billion among the Canadian Debtors, the U.S. Debtors and the EMEA Debtors.

[4] The trial in this case was unique. It was a joint trial of the Ontario Superior Court of Justice (Commercial List) and the U.S. Bankruptcy Court for the District of Delaware[3]. It arose from the arrangements made by the parties as part of the process of selling assets, and from a Cross-border Insolvency Protocol (the “Protocol”). In short:

(i) The parties agreed in an Interim Funding and Settlement Agreement before any of the Nortel assets were sold to put the proceeds of sale into escrow and then attempt to agree on a protocol for resolving how the proceeds were to be allocated. If no agreement was reached, the issues were to be tried by the Ontario and U.S. Courts pursuant to the Protocol.

(ii) The parties could not agree on the allocation, nor could they agree on a protocol process. By orders of the Ontario and U.S. Courts, the allocation was directed to be determined in a joint trial pursuant to the Protocol. The EMEA Debtors were held to have attorned to the jurisdiction of these courts in the escrow agreements made with respect to the proceeds of the several sales that had occurred.[4]

[5] The Protocol was approved early in the CCAA and chapter 15 proceedings by orders the Ontario and U.S. Courts.[^5] This type of protocol has become standard in the last number of years to govern the administration of cross-border insolvency proceedings. The Protocol included it its purposes:

Accordingly, this Protocol has been developed to promote the following mutually desirable goals and objectives in the Insolvency Proceedings:

(a) harmonize and coordinate activities in the Insolvency Proceedings before the Courts;

(b) promote the orderly and efficient administration of the Insolvency Proceedings to, among other things, maximize the efficiency of the Insolvency Proceedings, reduce the costs associated therewith and avoid duplication of effort;

(c) honor the independence and integrity of the Courts and other courts and tribunals of the United States and Canada, respectively;

(d) promote international cooperation and respect for comity among the Courts, the Debtors, the Creditors Committee, the Estate Representatives (which include the Chapter 11 Representatives and the Canadian Representatives as such terms are defined below) and other creditors and interested parties in the Insolvency Proceedings;

(e) facilitate the fair, open and efficient administration of the Insolvency Proceedings for the benefit of all of the Debtors’ creditors and other interested parties, wherever located; and

(f) implement a framework of general principles to address basic administrative issues arising out of the cross-border nature of the Insolvency Proceedings.

[6] The Protocol contained a number of provisions regarding the independence of the Canadian and U.S. Courts and the exclusive jurisdiction of each Court in the determination of matters arising in the Canadian and U.S. proceedings respectively. Included in the Protocol were the following provisions:

The approval and implementation of this Protocol shall not divest nor diminish the U.S. Court’s and the Canadian Court’s respective independent jurisdiction over the subject matter of the U.S. Proceedings and the Canadian Proceedings, respectively…
The U.S. Court shall have sole and exclusive jurisdiction and power over the conduct of the U.S. Proceedings and the hearing and determination of matters arising in the U.S. Proceedings. The Canadian Court shall have sole and exclusive jurisdiction and power over the conduct of the Canadian Proceedings and the hearing and determination of matters arising in the Canadian Proceedings.

[7] The Protocol provided in paragraph 12 for the harmonization and co-ordination of the administration of the two proceedings in Canada, including the holding of joint hearings of the two Courts and providing for discussions between the two judges. Included were the following:

To harmonize and coordinate the administration of the Insolvency Proceedings, the U.S. Court and the Canadian Court each may coordinate activities and consider whether it is appropriate to defer to the judgment of the other Court. In furtherance of the foregoing:

(a) The U.S. Court and the Canadian Court may communicate with one another, with or without counsel present, with respect to any procedural matter relating to the Insolvency Proceedings.

(d) The U.S. Court and the Canadian Court may conduct joint hearings (each a “Joint Hearing”) with respect to any cross-border matter or the interpretation or implementation of this Protocol where both the U.S. Court and the Canadian Court consider such a Joint Hearing to be necessary or advisable, or as otherwise provided herein, to, among other things, facilitate or coordinate proper and efficient conduct of the Insolvency Proceedings or the resolution of any particular issue in the Insolvency Proceedings. With respect to any Joint Hearing, unless otherwise ordered, the following procedures will be followed:

(vi) The Judge of the U.S. Court and the Justice of the Canadian Court, shall be entitled to communicate with each other during or after any joint hearing, with or without counsel present, for the purposes of (1) determining whether consistent rulings can be made by both Courts; (2) coordinating the terms upon of the Courts’ respective rulings; and (3) addressing any other procedural or administrative matters.

[8] A joint hearing was held for this allocation dispute. The court rooms in Toronto and Wilmington were set up electronically so that lawyers and witnesses could and did appear in either courtroom and communicate with a lawyer, witness or the judge in the other courtroom through state of the art telecommunications services.

[9] After the evidence was heard, written closing and reply briefs were filed by the parties and oral argument was made. It was agreed that at the conclusion of the case that each Court would release its decision at the same time. This judgment is being released at the same time as the opinion of Judge Gross in Wilmington.

[10] Judge Gross in Wilmington and I have communicated with each other in accordance with the Protocol with a view to determining whether consistent rulings can be made by both Courts. We have come to the conclusion that a consistent ruling can and should be made by both Courts. We have come to this conclusion in the exercise of our independent and exclusive jurisdiction in each of our jurisdictions. These insolvency proceedings have now lasted over six years at unimaginable expense and they should if at all possible come to a final resolution. It is in all of the parties’ interests for that to occur. Consistent decisions that we both agree with will facilitate such a resolution.

Nortel history and its matrix structure

[11] NNC was the successor to a long line of technology companies headquartered in Canada dating back to the founding of Bell Telephone Company of Canada in 1883. Prior to being named Nortel, it was known as Northern Telecom. NNC’s principal, direct operating subsidiary, also a Canadian company, was Nortel Networks Limited (“NNL”), which in turn was the direct or indirect parent of operating companies located around the world.[6]

[12] From the mid-1980s, Nortel expanded substantially through the continued development of ground-breaking technology. The Nortel Group moved from developing and manufacturing traditional landline phone technology and equipment into digital, wireless and photonic technologies. At the same time, the Nortel Group expanded into Europe, Asia, Africa, the Middle East and Latin America.

[13] At the time of its insolvency, Nortel had four main product groups (also known as Lines of Business):

• The “Carrier Networks” segment provided wireless networking solutions that enabled service providers and cable operators to supply mobile voice, data and multimedia communications to individuals and enterprises using mobile phone and other wireless devices. The Carrier Networks business also offered products providing local, toll, long distance and international gateway capabilities to telephone service providers as well as providing support to customers transitioning from one network to another.

• The “Enterprise Solutions” segment provided enterprise communications solutions addressing the headquarters, branch and home office needs of large and small businesses. The Enterprise Solutions segment’s offerings included, among other things, Unified Communications, Ethernet routing and multiservice switching, IP and digital telephony (including phones), wireless LANs, security, IP and SIP contact centers, self-service solutions, messaging, conferencing and SIP-based multimedia solutions.

• The Metro Ethernet Networks (“MEN”) segment provided carrier-grade Ethernet transport capabilities focused on meeting customers’ needs for higher performance and lower cost emerging video-intensive applications. MEN included optical networking, carrier Ethernet switching products and multi-service switching products.

• The “Global Services” segment provided a broad range of services and solutions including network implementation services, network support services, network managed services (which related to the monitoring and management of customer networks and hosted solutions) and network application services.

[14] The Nortel Group consists of more than 140 separate corporate entities located in 60 separate sovereign jurisdictions including Canada, the United States and the EMEA region, as well as the Caribbean and Latin America and Asia. NNC, the Nortel Group’s ultimate parent holding company, was publicly listed and traded on both the Toronto Stock Exchange and the New York Stock Exchange.

[15] One of NNC’s direct subsidiaries is NNL, which was the Canadian operating company of the Nortel Group. NNL in turn owns 100% of the equity of each of NNI, which was the Nortel Group’s operating company in the United States, NNUK, which was the Nortel Group’s operating company in the United Kingdom, NN Ireland, which was the Nortel Group’s operating company in Ireland, and 91.17% of the equity of NNSA, which was the Nortel Group’s operating company in France.

[16] The Nortel Group operated along business lines as a highly integrated multinational enterprise with a matrix structure that transcended geographic boundaries and legal entities organized around the world. Each entity, such as NNL, NNI, NNUK, NN Ireland and NNSA, was integrated into regional and product line management structures to share information and perform research and development (“R&D”), sales and other common functions across geographic boundaries and across legal entities. The matrix structure was designed to enable Nortel to function more efficiently, drawing on employees from different functional disciplines worldwide, allowing them to work together to develop products and attract and provide service to customers, fulfilling their demands globally.

[17] As a result of Nortel’s matrix structure, no single Nortel entity, either NNL or any of the other Canadian debtors in Canada, NNI or any of the other US debtors in the United States or NNUK or any of the other EMEA debtors, was able to provide the full line of Nortel products and services, including R&D capabilities, on a stand-alone basis. While Nortel ensured that all corporate entities complied with local laws regarding corporate governance, no corporate entity carried on business on its own.

[18] R&D was the primary driver of Nortel’s value and profit. Together with NNL, the principal companies that performed R&D were NNI, NNUK, NNSA and NN Ireland. These were known as Integrated Entities or, in transfer pricing terms, Residual Profit Entities (“RPEs”) due to their participation from 2001 in a residual profit pool in connection with Nortel’s transfer pricing arrangements[7]. Other operating companies performed sales and distribution functions and were known as Limited Risk Distributors or Entities (“LREs”).

[19] R&D was performed at labs around the world. The advanced technology primary research which was intended to develop novel, cutting edge intellectual property technologies was performed mostly in NNL laboratories in Ottawa, which also did R&D for various lines of business. From 2000 to 2009 NNL accounted on average for just under half of all R&D expenditures, more in the latter years than the earlier years. NNI accounted for 38 to 42% and EMEA accounted for 16 to 20% in the earlier years and 11.7 % from 2005 to 2009. The R&D was shared throughout the Nortel Group as needed by the lines of business and customer needs in the various regions and countries.

[20] Because R&D was the primary driver of Nortel’s value and profit, the residual profits of Nortel, after payment of fixed rates of return to all Nortel companies for sales and distribution functions, were paid to the RPEs under a Master Research and Development Agreement (“MRDA”) in accordance with a residual profit split method (“RPSM”) based on each RPE’s expenditure on R&D relative to the R&D expenditure of all RPEs.

[21] Under the MRDA, NNL was the legal owner of the Nortel intellectual property and each RPE other than NNL was granted an exclusive license by NNL to make and sell Nortel products in its territory using or embodying Nortel intellectual property developed by Nortel companies anywhere in the world and a non-exclusive license to do so in territories that were not exclusive to an RPE. What the ownership rights of NNL were and what the license rights were that were granted in the MRDA are highly contested. Also contested is the role that the MRDA should play in this allocation proceeding.

Bankruptcy filings

[22] Beginning around 2001, the burst of the dot-com bubble had a severe effect on the global economy and on the telecommunications industry in particular, including Nortel. Market forces led to a decline in Nortel’s revenues and market share, and a decline in customer demand for Nortel’s products. Subsequently, Nortel was faced with accounting issues which impacted Nortel’s credit rating and its cost of financing and required Nortel to restate its financial statements for the fiscal years 2000 to 2005. The rating downgrades affected Nortel’s access to capital markets and cost of financing for some years. The fortunes of Nortel improved for a few years but for various reasons, including the financial meltdown in the fall of 2008, Nortel saw its business decline in the two profitable lines of business that it was operating.

[23] In light of the impact of the deteriorating market conditions and weakening customer commitments on Nortel's financial outlook, Nortel made the decision to commence formal bankruptcy and insolvency proceedings in Canada, the U.S. and England (respecting various EMEA entities) on January 14, 2009.

[24] On January 14, 2009 NNC, NNL, the wholly owned subsidiary of NNC which was its operating subsidiary and a number of other Canadian corporations filed for protection under the CCAA. On the same date, Nortel Network Inc. (“NNI”), the principal US subsidiary of NNL, and a number of other US corporations filed for protection under chapter 11 of the US Bankruptcy Code and Nortel Networks UK Limited (“NNUK”), the principal UK subsidiary of NNL, and certain of their subsidiaries (the “EMEA Debtors”) save the French subsidiary Nortel Networks S.A. (“NNSA”) were granted administration orders under the UK Insolvency Act, 1986. On the following day, a liquidator of NNSA was appointed in France pursuant to Article 27 of the European Union’s Council Regulation (EC) No 1346/2000 on Insolvency Proceedings in the Republic of France.

[25] Subsequent to the filing date, certain other Nortel subsidiaries have filed for creditor protection or bankruptcy proceedings in the local jurisdiction in which they are located. Certain solvent indirect subsidiaries of NNUK are not in administration, but are represented in these proceedings by the Joint Administrators with respect to the allocation issues.

Decision to liquidate

[26] The initial intent on filing was to attempt to restructure the business and downsize it by focusing on Nortel’s legacy CDMA (Code Division Multiple Access) wireless business and a potential business based on LTE (Long-Term Evolution) wireless technology with all other Nortel business lines being sold. However, Nortel's major customers did not support this plan and advised they were not prepared to provide new contracts to Nortel for this purpose. As well, it became clear that it would not be possible for Nortel to obtain the funding that would have been required to restructure around a CDMA business.

[27] In June 2009, management and the Debtor Estates collectively determined that the best means to maximize value for its creditors was to sell Nortel's lines of business and other assets and to commence a liquidating insolvency. No party in these proceedings has suggested that it was a viable option to restructure along geographic lines or for a country-specific entity to independently continue in Nortel’s business.

Interim Funding and Settlement Agreement (“IFSA”)

[28] From the petition date of January 14, 2009, NNL incurred significant expenses to preserve the value of the business, including R&D expenses, and it was experiencing negative cash flow. It had not received any transfer pricing payments from its subsidiaries under the MRDA as a result of the insolvency proceedings.

[29] It was evident that there would be significant issues among the parties as to whom the proceeds of the sale of Nortel’s assets should be paid. The parties appreciated that if determining the allocation of proceeds from Nortel’s assets were a precondition to their sale, sales would be substantially delayed, and the value of the assets would depreciate, resulting in less money for all creditors. Avoiding a dispute during the sale processes about how to allocate the proceeds allowed the parties to obtain the highest monetary value for the assets being sold.

[30] On June 9, 2009, the US Debtors (excluding NN CALA, which had not yet filed for bankruptcy), the Canadian Debtors and the EMEA Debtors (excluding NNSA, which later acceded to the agreement) entered into the Interim Funding and Settlement Agreement (“IFSA”) to address both interim funding of NNL as well as principles under which collaborative sales of Nortel’s businesses and assets could take place.

[31] The IFSA provided for a payment by NNI to NNL of $157 million in full settlement of any transfer pricing and other claims NNL might have had against NNI for the period from the petition date through September 30, 2009. The parties also agreed:

(a) to cooperate in the anticipated sales of the Nortel Group’s assets;

(b) that their execution of sale documentation or the closing of a sale transaction would not be conditioned upon reaching agreement either on allocation of the sale proceeds or on a binding procedure for determining the allocation question;

(c) that the sale proceeds would be deposited into escrow, and that there would be no distribution out of escrow without either the agreement of all of the selling debtors or the determination of any dispute relating thereto by the relevant dispute resolver;

(d) that in order to facilitate the lines of business sales, the U.S. and EMEA Debtors would enter into appropriate license termination agreements which would provide for the termination of the license rights granted by NNL under the MRDA; termination or relinquishment of a license would be deemed a sale with the licensed participants each being deemed a seller; and

(e) that the agreement would not have any impact on the allocation of proceeds to any Debtor from any asset sale and would not prejudice a party’s rights to seek its entitlement to the proceeds from any sale.

[32] The US and Canadian Courts entered orders approving the IFSA following a joint hearing on June 29, 2009.

[33] On December 23, 2009 the Canadian and U.S. Debtors signed a Final Canadian Funding and Settlement Agreement (the “FCFSA”) under which NNI agreed to pay NNL $190.8 million in full and final settlement of all claims that NNL might have against NNI. Further, NNL granted NNI an allowed $2 billion unsecured claim in NNL’s CCAA proceedings ranking pari passu with other pre-petition unsecured claims against NNL, with such claim not being subject to offset or reduction. This claim had resulted from the tax authorities reviewing requests by the parties for approval of their transfer pricing arrangements. In 2009 NNL and NNI were advised that an agreement between the CRA and IRS sought a reallocation of income from NNL to NNI in the amount of U.S. $2 billion for the tax years ending 2001 to 2005. The tax authorities did not specify on what basis the $2 billion figure was calculated. The FCFSA, including the $2 billion admitted claim of NNI against NNL, was approved by the Canadian Court on January 21, 2010 and by the U.S. Court on the following day.

Asset sales

[34] With the IFSA framework in place, the Debtor Estates embarked on a process that resulted in a series of sales of the various business lines, which occurred from mid-2009 through late 2010, with the last transaction closing in March 2011. The total proceeds were approximately $3.285 billion. There remains approximately $2.85 billion of that amount now available to be allocated.

[35] In order to sell the lines of businesses separately, Nortel engaged in a “carve-out process” to identify the bundle of assets, rights and obligations that would have to be conveyed in each sale to enable the lines of business to function on a stand-alone basis.

[36] An important aspect of the carve-out process was the identification of which IP rights, principally patent rights, needed to be conveyed. Each prospective purchaser of a business line wished to obtain as many patents as possible as part of each sale transaction and, conversely, the Nortel sellers wanted to ensure that the only patents transferred were those incorporated exclusively or principally in the business line in question so as to retain value within Nortel and not to jeopardize the ability to sell the other business lines that might require rights to the same patents.

[37] Ultimately, those patents that were “predominantly used” in any given line of business were transferred to the purchaser of that line of business as part of the transaction. In the end, 2,700 patents were transferred as part of the business line sales.

[38] For all other patents that were used in each line of business but not predominantly used, a non-exclusive license was granted to the purchaser for use of those patents in the operations of the particular business line being purchased.

[39] By the time that all of the business sales were completed in March 2011, Nortel had no remaining operating businesses. What it did retain was a residual patent portfolio consisting of approximately 7000 patents and patent applications. These were principally patents and patent applications that were not used in any of the lines of business and therefore were not subject to licenses to the business sale purchasers. In addition, the residual IP portfolio included patents used by multiple lines of businesses and licensed to the purchasers of those lines of businesses.

[40] On April 4, 2011, after significant negotiations with two prospective purchasers, certain Nortel entities (including NNC, NNL, NNI and NNUK) entered into a stalking horse asset sale agreement with a wholly owned subsidiary of Google Inc. with a purchase price of $900 million.

[41] An auction was held at the end of June 2011, and the residual patent portfolio was ultimately sold to Rockstar Bidco, LP, a single purpose entity backed by a consortium of major technology companies (Apple, Microsoft, Ericsson, Blackberry, Sony and EMC), for $4.5 billion.

Position of the parties

[42] In this case the Monitor is acting under what is now referred to as a “super monitor” order of October 3, 2012 in which the Monitor was authorized to exercise any powers which may be exercised by a board of directors of any of the applicants, which includes NNC and NNL. This order occurred after NNC and NNL were left without any board of directors or management and it was necessary for the Monitor to be appointed to advance the interests of NNL and NNC in this CCAA proceeding. While I will refer to the Monitor, I do so in recognition that the Monitor is advancing the position of the Canadian Debtors in this litigation.

[43] The intellectual property of Nortel represented by far the largest portion of the assets sold. The Rockstar sale of the residual IP generated $4.5 billion. The lines of business generated $3.285 billion of which approximately $2.85 billion is now available. Intellectual property was a substantial part of the assets of the business lines that were sold, although the experts differed as to its value.

[44] The parties and their experts for the most part relied on their interpretation of the MRDA in support of their allocation positions for the proceeds from intellectual property for both the Rockstar sale and the lines of business sales. Two parties, the UKPC (the UK pension claimants, being the trustee of the UK pension plan, and the board of the UK Pension Protection Fund) and the Canadian Creditors Committee[8] contended that the MRDA should not govern the allocation and that a pro rata allocation based on a pari passu distribution to all creditors should be used to allocate the lockbox funds.

[45] It is necessary therefore to consider the MRDA and whether it should govern the allocation.

The MRDA

[46] The parties look to the rights of the various Nortel entities to intellectual property under the MRDA as a central issue in this proceeding. What these rights are is contested. Many of its terms have been excruciatingly parsed. I will first deal with the meaning of the MRDA as an operating agreement. I will then deal with the issue as to whether it applies, or was intended to apply, to the allocation of the Nortel assets after the world-wide insolvency of Nortel.

[47] The MRDA and its predecessor Cost Sharing Agreements[^9] (“CSA”) were developed for and driven by transfer pricing concepts. Transfer pricing is the act of assigning a monetary value, or price, to movements of resources or economic contributions that occur within a multinational enterprise across different taxing jurisdictions. Against the risk that companies attempt to use transfer pricing to increase operating income (and therefore taxable income) in jurisdictions with low income tax rates and correspondingly to decrease operating income in high-tax jurisdictions, tax authorities around the world have instituted regulations governing intercompany transfer pricing. These regulations centre on the arm’s length principle. The arm's length principle necessitates that intercompany transactions be priced in a manner consistent with the way in which similarly situated uncontrolled parties bargaining at arm’s length would price the transactions i.e., within an arm's length range.

[48] Dr. Eden, a transfer pricing expert who testified on behalf of the U.S. Debtors, well described in her report the way in which transfer pricing agreements are made in light of the fact that governments have developed a dense regulatory framework for transfer pricing due to worries about the potentially negative impacts that transfer pricing can have on government tax and customs duty revenues. The setting of transfer pricing policies for corporate income tax purposes of a multinational enterprise (MNE) is a highly regulated, data-driven and fact-intensive activity dominated by professionals. The establishment of an MNE’s transfer pricing policy typically involves not only MNE group in-house staff, but also accountants, economists, lawyers, tax experts and other consultants. Moreover, an MNE’s transfer pricing policy may involve the input of revenue authorities through an advance pricing agreement (APA) procedure.

[49] All of this applied to Nortel and much evidence was given by tax people as to the process by which the MRDA was made and changed. Evidence was also given by some of them as to their view of the meaning of the agreement, the admissibility of which is contested.

(i) Governing law of the construction of the contract

[50] The MRDA is by its terms to be construed in accordance with and governed by the law of Ontario. The same applied to the predecessor CSAs.

[51] A number of authorities have been cited. A brief consideration of them is required in light of the various arguments made about the MRDA, particularly as it involves the principles of interpreting commercial contracts, what can be looked at when considering the factual matrix of the agreement and the use of recitals in an agreement in the interpretive process.

[52] Winkler C.J.O. articulated the test for construing a commercial contract in Salah v. Timothy's Coffees of the World Inc. (2010), 2010 ONCA 673, 74 B.L.R. (4th) 161 as follows:

16 The basic principles of commercial contractual interpretation may be summarized as follows. When interpreting a contract, the court aims to determine the intentions of the parties in accordance with the language used in the written document and presumes that the parties have intended what they have said. The court construes the contract as a whole, in a manner that gives meaning to all of its terms, and avoids an interpretation that would render one or more of its terms ineffective. In interpreting the contract, the court must have regard to the objective evidence of the "factual matrix" or context underlying the negotiation of the contract, but not the subjective evidence of the intention of the parties. The court should interpret the contract so as to accord with sound commercial principles and good business sense, and avoid commercial absurdity. If the court finds that the contract is ambiguous, it may then resort to extrinsic evidence to clear up the ambiguity.

[53] In Kentucky Fried Chicken v. Scott's Food Services Inc. (1998), 1998 CanLII 4427 (ON CA), 41 B.L.R. (2d) 42 (Ont. C.A.) Goudge J.A. stated the following regarding the interpretation of a commercial agreement at para. 27

Where, as here, the document to be construed is a negotiated commercial document, the court should avoid an interpretation that would result in a commercial absurdity. [City of Toronto v. W.H. Hotel Ltd. (1966), 1966 CanLII 8 (SCC), 56 D.L.R. (2d) 539 at 548 (S.C.C.)]. Rather, the document should be construed in accordance with sound commercial principles and good business sense; [Scanlon v. Castlepoint Development Corporation et al. (1992), 1992 CanLII 7745 (ON CA), 11 O.R. (3d) 744 at 770 (Ont.C.A.)]. Care must be taken, however, to do this objectively rather than from the perspective of one contracting party or the other, since what might make good business sense to one party would not necessarily do so for the other.

[54] I take the principles in Kentucky Fried Chicken and in Salah, the latter adopted by Cronk J.A. in Downey v. Ecore International Inc. 2012 ONCA 480 and by Juriansz J.A. in Ariston Realty Corp. v. Elcarim Inc. 2014 ONCA 737, as the applicable principles governing this case. See also Unique Broadband Systems Inc. (Re) 2014 ONCA 538 at para. 88.[^10]

[55] The factual matrix of the contract is to be considered. What may be considered was expressed in Kentucky Fried Chicken as follows:

25 …While the task of interpretation must begin with the words of the document and their ordinary meaning, the general context that gave birth to the document or its "factual matrix" will also provide the court with useful assistance. In the famous passage in Reardon Smith Line Ltd. v. Yngvar Hansen-Tangen, [1976] 1 W.L.R. 989 at 995-96 (H.L.) Lord Wilberforce said this:

No contracts are made in a vacuum: there is always a setting in which they have to be placed. The nature of what is legitimate to have regard to is usually described as "the surrounding circumstances" but this phrase is imprecise: it can be illustrated but hardly defined. In a commercial contract it is certainly right that the court should know the commercial purpose of the contract and this in turn presupposes knowledge of the genesis of the transaction, the background, the context, the market in which the parties are operating.

26 The scope of the surrounding circumstances to be considered will vary from case to case but generally will encompass those factors which assist the court "... to search for an interpretation which, from the whole of the contract, would appear to promote or advance the true intent of the parties at the time of entry into the contract." Consolidated Bathurst Export Ltd. v. Mutual Boiler and Machinery Insurance Co., 1979 CanLII 10 (SCC), [1980] 1 S.C.R. 888 at 901.

[56] More recently, Rothstein J. in Sattva Capital Corp. v. Creston Moly Corp. 2014 SCC 53 referred to the use of surrounding circumstances and cautioned as to the extent they can be considered:

57 While the surrounding circumstances will be considered in interpreting the terms of a contract, they must never be allowed to overwhelm the words of that agreement (Hayes Forest Services, at para. 14; and Hall, at p. 30). The goal of examining such evidence is to deepen a decision-maker's understanding of the mutual and objective intentions of the parties as expressed in the words of the contract. The interpretation of a written contractual provision must always be grounded in the text and read in light of the entire contract (Hall, at pp. 15 and 30-32). While the surrounding circumstances are relied upon in the interpretive process, courts cannot use them to deviate from the text such that the court effectively creates a new agreement (Glaswegian Enterprises Inc. v. B.C. Tel Mobility Cellular Inc. (1997), 1997 CanLII 4085 (BC CA), 101 B.C.A.C. 62).

58 The nature of the evidence that can be relied upon under the rubric of "surrounding circumstances" will necessarily vary from case to case. It does, however, have its limits. It should consist only of objective evidence of the background facts at the time of the execution of the contract (King, at paras. 66 and 70), that is, knowledge that was or reasonably ought to have been within the knowledge of both parties at or before the date of contracting. Subject to these requirements and the parol evidence rule discussed below, this includes, in the words of Lord Hoffmann, "absolutely anything which would have affected the way in which the language of the document would have been understood by a reasonable man" (Investors Compensation Scheme, at p. 114). Whether something was or reasonably ought to have been within the common knowledge of the parties at the time of execution of the contract is a question of fact.

[57] It is clear that the factual matrix that can be considered may not include evidence of the subjective intent of a party or what a party believed a contract to mean. See Sattva, supra, at para. 59. It may also not include evidence of negotiations or create an ambiguity where none exists in an agreement. See also Primo Poloniato Grandchildren's Trust (Trustee of) v. Browne (2012), 2012 ONCA 862, 115 O.R. (3d) 287 in which Feldman J.A. stated:

71 While the scope of the factual matrix is broad, it excludes evidence of negotiations, except perhaps in the most general terms, and evidence of a contracting party's subjective intentions: Geoff R. Hall, Canadian Contractual Interpretation Law, 2d ed. (Markham: LexisNexis, 2012), at p. 27. As the cases above suggest, the factual matrix includes only objective facts known to the parties at or before the date of the agreement, and what is common to both parties: Hall, p. 30. Hall goes on to state that while the factual matrix can "be used to clarify the parties' intentions as expressed in a written agreement, it cannot be used to contradict that intention, create an ambiguity which otherwise does not exist in the written document, or have the effect of making a new agreement": p. 31 (footnotes omitted). Ultimately, the words of the agreement are paramount.

[58] The recitals in the MRDA are the subject of debate in this case. A clear statement of how recitals may be used in the interpretation of an agreement can be found in Elliott Estate (Re), 1962 O.J. No. 164 (C.A.); aff’d 1963 CanLII 11 (SCC), [1963] S.C.R. 305. In that case, Kelly J.A. stated that a recital could be used only if there is an ambiguity in the operative parts of the agreement and the recital is clear. He stated:

11 I turn therefore to consider to what extent the recital may be used to overcome the patent deficiencies of clauses 6 and 7 and in fact of the whole operative parts of the agreement. In the first instance it must be borne in mind that a recital is not a necessary part of a document and its use in the interpretation of the document as a whole is strictly limited.

"The reciting Part of a Deed is not at all a necessary Part either in Law or Equity. It may be made use of to explain a Doubt of the Intention and Meaning of the Parties but it hath no Effect or Operation. But when it comes to limit the estate, there the Deed is to have its Effect according to what Limitations are therein set forth."

Per Holt, C.J., Bath and Mountague's Case (1693) 3 Cas. in Ch. 55 at 101; 22 E.R. 963 at 991. An oft quoted statement of the extent to which reference may be had to recitals is contained in the judgment of Lord Esher, M.R. in Ex Parte Dawes. In Re Moon, (1886) 17 Q.B.D. 275 at p. 286:

"Now there are three rules applicable to the construction of such an instrument. If the recitals are clear and the operative part is ambiguous, the recitals govern the construction. If the recitals are ambiguous, and the operative part is clear, the operative part must prevail. If both the recitals and the operative part are clear, but they are inconsistent with each other, the operative part is to be preferred."

It is to be noted that the qualifying condition for the use of a recital in the interpretation of the operative parts is that there must be ambiguity in the operative parts; in such a case the preferred meaning to be given to the operative words should be that consistent with the intention expressed in the recital, provided that the words of the operative part are by themselves capable of such an interpretation. MacKenzie v. Duke of Devonshire, (1896) App. Cas. 400; Ex Parte Dawes. In Re Moon, Supra; In re Sugden's Truts, Sugden v. Walker, (1917) 2 Ch. 92. It is essential, however, that the construction to be placed upon the operative part in the light of the recital be a construction which the words themselves of the operative part are capable of bearing. Where, however, the operative parts of a document, due to the lack of appropriate words, are incapable of a construction which will fulfil the intention expressed in recitals, the recital may not be used for the purpose of reading into the operative clause a meaning which it is incapable of conveying when considered by itself.

[59] It was held in PUC v. Distribution Inc. v. Brascan Energy Marketing Inc., 2008 ONCA 176, that an elevation of a recital to a mutual promise or operative provision was an error.

[60] Eli Lilly & Co. v. Novopharm Ltd., 1998 CanLII 791 (SCC), [1998] 2 S.C.R. 129 at para. 57, in which Iacobucci J. in discussing the meaning of an agreement referred to the recitals, was referred to in argument. Iacobucci J. did not discuss the principles to be used in considering recitals. Sistem v. Kyrgyz Republic, 2012 ONSC 4983, has also been referred to in argument, a decision in which I did not refer to the principles to be used in considering recitals in interpreting contracts. I consider the decision in Sistem to be consistent with the principles enunciated by Kelly J.A. in Elliott Estate. I do not see either Eli Lilly or Sistem establishing any different criteria for the use of recitals from Elliott Estate.

[61] I turn now to the interpretation of the MRDA and the rights accorded in it keeping these interpretive principles in mind.

(ii) Position of the parties

[62] The essential differences in allocation positions advanced by the parties flow from the different manner in which each characterizes the terms of the MRDA, the interests held by the parties in Nortel’s IP, and the applicability of terms of the MRDA to the value ascribed to various assets.

[63] The Monitor, supported by the CCC, contends that under the MRDA, NNL owned the IP and the interests of NNI and the other participants to the MRDA were restricted to certain exclusive and non-exclusive license rights granted to them by NNL pursuant to the terms of the MRDA. The Monitor says that the license rights were not unlimited, as they did not cover all rights in the IP in question, but rather covered only a subset (albeit a substantial subset) of the IP rights, on certain terms, all of which have valuation implications. In particular, the Monitor says that the license rights granted to NNI and the other licensed participants were not all rights to the IP but were subject to “field of use” restrictions that gave the licensees the right to use the IP to make, use or sell “Products” as defined in the MRDA, which meant products, software or services that were made or sold by, or for, any of the licensees. This meant that the Products must have been created or marketed by or for the Nortel Group. No product that was part of a third party’s business rather than the business of Nortel could fall within the definition of Products. While the license gave the licensees the right to sublicense, this could not permit the licensees to sublicense what they did not have.

[64] The Monitor’s position, supported by the CCC, is that what was sold in the Rockstar sale of IP was the ownership of residual patents and patent applications owned by NNL. The purchasers would not have bought the residual IP to make Nortel products, and that as the license rights held by NNI and the other licensees would not have permitted them to sublicense to the Rockstar consortium the right to use the IP for the Rockstar consortium’s own purposes, the proceeds of the Rockstar sale belong to NNL.

[65] The position of NNI, supported by the other U.S. interests, asserts that each of NNI and the other licensees held all of the rights and all of the value in the IP in their respective exclusive territories as defined in the MRDA. The U.S. Debtors assert that the license rights NNI held were not subject to any field of use or scope restriction or limitation, resulting in an assertion that all of the economic value in the IP in the exclusive territory belonged to the licensee. They contend that the legal title held in the IP under the MRDA was a purely “bare” legal title with no monetary value. They also rely on a right to sue for damages in the U.S. for infringement of NN Technology by others.

[66] The position of the EMEA debtors is that each of the parties to the MRDA jointly owned all of the IP in proportion to their financial contributions to research and development, and that all share in the sale proceeds attributable to IP in those same proportions. The joint ownership is said to arise independent of, but recognized in, the MRDA.

(iii) Analysis

(a) The meaning of the exclusive license

[67] The agreement is headed MASTER R&D AGREEMENT. It was entered into on December 22, 2004 with an effective date of January 1, 2001 and states that it confirms and formalizes the operating arrangements of the participants as and from that date. It provided that NNL was the legal owner of the NN Technology (the IP), and it contained grants of licenses from NNL to the other participants, referred to as the Licensed Participants. Each Licensed Participant was given an exclusive license for its territory and a non-exclusive license for those parts of the world other than Canada and where the Licensed Participants had their exclusive territory. The exclusive territory for NNI was the U.S. and Puerto Rico, for NNUK was the United Kingdom, for NNSA was France and for Nortel Ireland was the Republic of Ireland.

[68] At its core, so far as the ownership and licensing of the IP is concerned are articles 4(a) and 5(a) and (b). The original language remained in substance but was amended from time to time. These articles as amended are as follows:

Article 4 – Legal Title to NN Technology

(a) Except as otherwise specifically agreed, legal title to any and all NN Technology whether now in existence or hereafter acquired, or developed pursuant to the terms of this Agreement, shall be vested in NNL. In consideration therefor, NNL agrees to enter into an Exclusive License and a Non-Exclusive License with each of the Licensed Participants as set forth in Article 5.

Article 5 – Grant of Exclusive Licenses by NNL

(a) To the extent of its legal right to do so, and subject to the rights of relevant third parties, NNL hereby:

(i) continues to grant to each Licensed Participant an exclusive, royalty-free license, including the right to sublicense, which except as hereinafter provided shall be in perpetuity, rights to make, have made, use, lease, license, offer to sell, and sell Products using or embodying NN Technology in and for the Exclusive Territory designated for that Licensed Participant, and all rights to patents, industrial designs (or equivalent) and copyrights, and applications therefor, and technical know-how, as necessary or appropriate in connection therewith (“Exclusive License”); and

(ii) grants to each Licensed Participant, as of January 1, 2009 (the “Non-Exclusive License Effective Date”), a non-exclusive, royalty-free license, including the right to sublicense, which except as hereinafter provided shall be in perpetuity, rights to make, have made, use, lease, license, offer to sell, and sell Products using or embodying NN Technology in and for the Non-Exclusive Territory, and all rights to patents, industrial designs (or equivalent) and copyrights, and applications therefor, and technical know-how, as necessary or appropriate in connection therewith (“Non-Exclusive License”).

[69] To support their differing interpretations of these provisions, the parties augment to some extent their arguments by reference to other provisions in the MRDA. It will be necessary to deal with these. As can be seen from article 5(i), NNL “continues to grant”, a reflection of the fact that prior to the MRDA, the parties were governed by Cost Sharing Agreements (CSAs)[^11]. Recitals to the MRDA make this clear:

WHEREAS legal title to all NN Technology is held in the name of NNL;

WHEREAS each Licensed Participant held and enjoyed equitable and beneficial ownership of certain exclusive rights under NT Technology for a Specified Territory pursuant to the Amended Research and Development Cost Sharing Agreement entered into on January 1, 1992, and it is the intent of NNL and the Licensed Participants that the Licensed Participants continue, as of the effective date of this Agreement, to hold and enjoy such rights;

[70] In considering the various interpretations of the MRDA put forward by the parties, it is helpful to compare those provisions with the earlier CSA provisions. Under the CSA, the parties split the costs of R&D by a certain formula. That agreement did not purport to split profits in any way. However, the tax authorities made it clear that they no longer would permit a cost sharing arrangement at Nortel and instead wanted an arrangement whereby profits would be shared among the participants by a residual profit split method (RPSM) that allocated profits according to the amount each participant spent on R&D. Relevant recitals in the MRDA that were not contained in the previous CSA are:

WHEREAS each Participant bears the full entrepreneurial risks and benefits for the Nortel Networks business;

WHEREAS each Participant has performed, in the past, and intends to continue to perform R&D Activity with respect to the Nortel Products;

WHEREAS each Participant desires to avoid the duplication of R&D Activity;

WHEREAS each Participant believes that it is appropriate that each Participant should benefit from its contribution to R&D activity commensurate with the value of its contribution to that R&D activity in the context of the manner in which the Nortel Networks business is conducted and that the residual profit split methodology (RPSM) is the best arm’s length measure, in the circumstances of NNL and the Participants, of such contributions with reference to such benefits;

WHEREAS this Agreement reflects the Participants’ intent and agreement since January 1, 2001 to enter a license arrangement with the Licensed Participants, and the Participants have operated from January 1, 2001 in accordance with the terms set forth herein;

WHEREAS Participants acknowledge that as a result of a collective review by the Canadian Customs and Revenue Agency, the US Internal Revenue Service, and the UK Inland Revenue regarding the application of the RPSM, the calculation of the RPSM as set forth in Amended Schedule A may be amended which amendments would require the consent of the Participants;

[71] These recitals and the RPSM method contained in the MRDA were driven by transfer pricing considerations. The language, for example, that each Participant (NNL and the Licensed Participants) bears the full entrepreneurial risks and benefits for the Nortel Networks business was not in the prior CSA and was part of the rationalization adopted to support a RPSM.

[72] The MRDA provided in article 2 that each Participant would perform R&D at a level consistent with past practices and share the results of its R&D with the other participants. Article 3 provided payment for the R&D as follows:

Article 3 – R&D Activity Payments

(a) For and as a consequence of the performance of R&D Activity, each Participant shall be entitled to receive a payment in an amount equal to the allocation determined under the RPSM (the “R&D Allocation”) as the measure of the benefit to which it is entitled commensurate with its performance of, and contribution to, R&D Activity.

(b) Each Participant hereby accepts and agrees to make the payment determined under the RPSM in Amended Schedule A[^12] as representing such Participant’s share of the R&D Allocation.

(c) The R&D Allocation will be computed pursuant Amended Schedule A which sets forth the basis of the RPSM as originally proposed to the Revenue Authorities. The Participants understand that the RPSM is the subject of review, discussions and negotiations with the Revenue Authorities. The Participants agree to amend this Agreement and to adjust the RPSM to the extent necessary to reflect any negotiated determination with the Revenue Authorities as to the final R&D Allocation.

[73] The U.S. Debtors and EMEA take the position that the legal title that is vested in NNL under article 4 of the MRDA is bare legal title given to NNL for administrative convenience to enable it to administer all NN Technology and that the licensed participants own the equitable and beneficial interest in the NN Technology. It draws on the recital that provides:

WHEREAS each Licensed Participant held and enjoyed equitable and beneficial ownership of certain exclusive rights under NT Technology[^13] for a Specified Territory pursuant to the Amended Research and Development Cost Sharing Agreement entered into on January 1, 1992, and it is the intent of NNL and the Licensed Participants that the Licensed Participants continue, as of the effective date of this Agreement, to hold and enjoy such rights;

[74] I do not see this recital as clearly stating that a Licensed Participant has equitable and beneficial ownership of the NT Technology. It states that a Licensed Participant held equitable and beneficial ownership of “certain exclusive rights under NT Technology” and would continue to have such rights. The recital does not say what the “certain exclusive rights” were and it is just as consistent with those rights being license rights rather than ownership rights in the technology. As well, having equitable and beneficial ownership of certain exclusive rights “under NT Technology” would seem to be something different from having equitable and beneficial ownership of certain exclusive rights “of” or “in” the NT Technology.

[75] In the CSA referred to in the recital, the language used is as follows:

Article 4 legal title to nt technology

The Parties hereto acknowledge that, except as otherwise specifically agreed, legal title to all NT Technology whether now in existence or developed pursuant to the terms of this Cost Sharing Agreement, except patents owned by Participant [Northern Telecom Inc., now NNI] on January 1, 1980, shall be vested in Northern Telecom [now NNL]. With respect to patentable inventions and copyrightable property encompassed by NT Technology, Northern Telecom shall have the exclusive right but not the obligation to file and prosecute applications in its name for patent or copyright protection in every country of the world. Participant shall execute or cause to be executed such documents reasonably requested by Northern Telecom as may be necessary or desirable to give effect to the foregoing. (Underlining added).

[76] The exception in this provision for patents owned by Northern Telecom Inc., now NNI, suggests that the legal title vested in Northern Telecom (now NNL) was ownership rather than bare legal title. Otherwise there would have been no purpose in excluding the patents owned by Northern Telecom Inc. It would not have been necessary.

[77] In article 6 of the CSA, dealing with confidential information, it is stated:

Participant acknowledges that Northern Telecom is the legal owner of the NT Technology developed pursuant to this Cost Sharing Agreement and that the NT Technology is proprietary and constitutes a trade secret. Participant shall hold the NT Technology in confidence and only make use of or disclose it as permitted by this Cost Sharing Agreement.

[78] This provision refers to Northern Telecom being the “legal owner”. This is consistent with the language of article 4 of the CSA. If, as stated in the recital to the MRDA, it was the intent of NNL and the Licensed Participants that the Licensed Participants would continue under the MRDA to hold and enjoy such rights as they held under the CSA, those rights would not include legal ownership of the NN Technology.

[79] NNI also relies on language in Schedule A of the MRDA to assert its beneficial ownership of the NN Technology. It provides in part:

Calculation of Arm’s Length R&D Allocation to each Participant

The purpose of this section is to provide a brief summary of Nortel’s transfer pricing policy and to provide clarity as to how each Participant is to be compensated under this Agreement.

The current transfer pricing methodology is the residual profit split method (“RPSM”) which was adopted by the Participants at the request of the tax authorities as the most appropriate method for determining the arm’s length compensation to each of the Participants for the R&D Activity to be provided pursuant to the Master R&D Agreement. The RPSM acknowledges the fact that the key profit driver in the Nortel business is the development and maintenance of rapidly depreciating intellectual property (“IP”).

Accordingly, the R&D Allocation provided to Participants under the RPSM reflects the fact that the Participants bear the full entrepreneurial risk of the Nortel business such as the risks attendant with the substantial and continuous development and ownership of the NN Technology. Mathematically, the RPSM accords the Participants all the upside risk in the Nortel business as well as the downside risk. (Underlining added).

[80] Schedule A is part of the MRDA. I do not, however, read it as granting rights. The rights are granted in the operative provisions of the MRDA. Schedule A states at the outset that its purpose is to give a brief summary of Nortel’s transfer pricing policy and to provide clarity as to how each participant is to be compensated. Schedule A provides in some detail how the residual profit is to be calculated and split amongst the Participants. Stating that Participants bear risks such as risks attendant with the development and ownership of the NN Technology does not state that ownership of the technology is being granted. What the Licensed Participants were granted in the MRDA were license rights.

[81] Various dictionary definitions were resorted to in arguing what the meaning of “legal title” to the NN Technology was that was vested in NNL under article 4 of the MRDA. In the end, I do not think it necessary to get into that debate. NNL had ownership of NN Technology to the extent that NN Technology was not licensed to the Licensed Participants. Rights in inventions were assigned by the inventors to NNL and NNL applied for the patents and was named as owner of them. It was NNL who granted licenses to the Licensed Participants. NNUK, for example, did not provide a license to NNI for IP developed by NNUK. It was NNL that did so. Although NNL had the exclusive right to the NN Technology in Canada under the MRDA, the MRDA did not grant any license to NNL. That was recognition that it was NNL that owned the NN Technology.

[82] A licensee does not enjoy property rights. Its rights are contractual. A licence is merely a permission to do that which would otherwise amount to trespass. See Euro-Excellence Inc. v. Kraft Canada Inc., [2007] S.C.R. 20 at para. 27. A licensee’s rights are not necessarily equivalent to those of the patentee; rather, they are limited to, and qualified by, the express terms of the license. See Eli Lilly & Co. v. Novopharm Ltd., 1998 CanLII 791 (SCC), [1998] 2 S.C.R. 129 at para. 49. It is the determination of what those license rights were that were granted to the Licensed Participants in the MRDA that is important because it is those license rights that were given up by Licensed Participants to permit the business line sales and the sale of the residual IP to Rockstar.

[83] The grant of the exclusive license in the MRDA in article 5(i) is:

…NNL hereby:

[84] The license is not a license of NN Technology, but rather a license “to make… and sell Products using or embodying NN Technology”. Thus the MRDA definition of “Products” is of central importance and the Monitor says that “Products” is defined to mean products, software or services that were made or sold by, or for, NNL and the Licensed Participants. The Monitor contends that products not made for NNL or the Licensed Participants, such as products that would be made by the Rockstar consortium members or their licensees are not covered by the license.

[85] The definition of “Products” at Article 1(g) of the MRDA is:

“Products” shall mean all products, software and services designed, developed, manufactured or marketed, or proposed to be designed, developed, manufactured or marketed, at any time by, or for, any of the Participants, and all components, parts, sub-assemblies, features, software associated with or incorporated in any of the foregoing, and all improvements, upgrades, updates, enhancements or other derivatives associated with or incorporated in any of the foregoing. (Underlining added).

[86] The U.S. Debtors parse the language of the license grant and contend that the Licensed Participants obtained all of the rights to the NN Technology. They break down the grant of the exclusive license into four clauses as follows:

NNL hereby:

continues to grant to each Licensed Participant an exclusive, royalty-free license, including

the right to sublicense, which except as hereinafter provided shall be in perpetuity,

rights to make, have made, use, lease, license, offer to sell, and sell Products using or embodying NN Technology in and for the Exclusive Territory designated for that Licensed Participant, and

all rights to patents, industrial designs (or equivalent) and copyrights, and applications therefor, and technical know-how, as necessary or appropriate in connection therewith (“Exclusive License”).

[87] The U.S. Debtors stated in their opening brief that the opening grant of an exclusive, royalty-free license in the first clause is not limited by the word “including”. They say the word “including” does not create a limitation, that the word “including” follows the words “exclusive, royalty-free license” and thus the words that follow cannot, and do not purport to, limit the broad exclusive licenses granted to the licensed participants under the MRDA. In effect they argue that the opening words before the word “including” created a complete grant of a license without reserve.

[88] I cannot accept that argument. The words “continues to grant an exclusive, royalty-free license”, on their own, do not say what the license is, or what it is for, or for how long. Given that a licensee’s rights are limited to, and qualified by, the express terms of the license (Eli Lilly & Co at para. 49), a license grant of uncertain scope, such as proposed by the U.S. Debtors, would have no meaning. Moreover, the words “in and for the Exclusive Territory designated for that Licensed Participant” appear after “including”. On the U.S. Debtors’ reading of the license, the territorial limitation would only apply to the license to make Products, and would not apply to a broad exclusive license that they say is already created before one gets to the word “including”. It would also mean that the words “in perpetuity” which follow the reference to a sublicense would not apply to the broad exclusive license, which is inconsistent with what the U.S. Debtors say is the case.

[89] There would be no commercial purpose in the MRDA granting a broad unrestrictive license and then providing more specific grants in the license that are restricted. For example, the third clause restricts the licensee to selling Products, which contains terms of limitation.

[90] The U.S. Debtors also contend that the third clause permits NNI or any other Licensed Participant to make or have made for it Products using NN Technology and that the sublicense rights contained in the second clause are not so limited to Products using NN Technology. I cannot accept that contention. A sublicense could not sublicense more than the licensee had under its license and the second clause could not purport to do so. This argument of the U.S. Debtors relies on its argument that the first clause was a broad unrestrictive grant of a license, which argument I cannot accept.

[91] The U.S. Debtors contend that the fourth clause is a free-standing or “catch-all” license grant of all rights to patents etc. unconnected to the license to make, use or sell Products. The language of this provision is:

[92] The U.S. Debtors say that the concluding words “in connection therewith” refer to the preceding words “technical know-how”. The contention of the U.S. Debtors is that this last clause is like the first clause, being a separate grant not limited by the right to make, use or sell Products. The Monitor says that these all of these words in the clause relate to the license to make, use or sell Products and that the words “in connection therewith” do not relate only to the reference to technical know-how.

[93] I must say that I find it difficult to accept that the concluding words “in connection therewith” modify only the words “technical know-how”. There would be no need for a comma after the words technical know-how”. Those words, even if only applicable to the last clause, could apply equally to “industrial designs (or equivalent)” and “applications therefor”.

[94] I do not find persuasive at all the attempt of the U.S. Debtors to parse the language of the grant of license as they have done. On their reading, there are several different grants of license. Yet at the end of the paragraph are the words “Exclusive License” in parenthesis. There is only one license and the words should be read together harmoniously.

[95] The U.S. Debtors make the point that what they refer to as the last clause in the license grant would be superfluous on the reading of the Monitor. That is because the definition of Products and NN Technology includes patents and the other things contained in that last clause. The U.S. Debtors say that because in interpreting a contract one should strive to give meaning to all of its terms, the last clause should be read as providing rights different from the rights to make, use or sell Products. While this argument on its face has a certain attractiveness, I do not think it right in this case.

[96] The grant of license rights in article 5 is one grant. It does not in the paragraph expressly spell out the definition of Product or NN Technology. The draftsman may have thought it prudent to include the final clause. The words “in connection therewith” must be given some meaning and I do not accept the meaning given to them by the U.S. Debtors. I read the words as relating to the grant of a license to make, use and sell Products employing NN Technology, which in my view was the intent of the entire license granted in clause 5(i).

[97] The Monitor refers to a statement of Lord Hoffman, no stranger to contract interpretation and a legal giant of his day, in Beaufort Developments (NI) Ltd. v. Gilbert-Ash NI Ltd, [1999] A.C. 266 at 274 (H.L.) that arguments of redundancy should be treated with caution. He stated:

I think, my Lords, that the argument from redundancy is seldom an entirely secure one. The fact is that even in legal documents (or, some might say, especially in legal documents) people often use superfluous words. Sometimes the draftsmanship is clumsy; more often the cause is a lawyer’s desire to be certain that every conceivable point has been covered. One has only to read the covenants in a traditional lease to realise that draftsmen lack inhibition about using too many words.

[98] In Long v Delta Catalytic Industrial Services Inc., 1998 CanLII 18145 (AB KB), [1998] 6 W.W.R. 792, Fruman J. (as she then was) said much the same thing:

Some might argue that this interpretation makes the provision redundant…That may well be the case, but it won’t be the first time that a repetitive provision has been inserted into an agreement.

[99] Redundancy could also be laid at the feet of the U.S. Debtors in their interpretation of the license grant. If their reading is correct, all of the second, third and fourth clauses would be redundant as the first clause was an unrestricted grant of a license. I think in this case redundancy arguments are just that, arguments that do not deal with the commercial purpose of the agreement.

[100] An addendum to the MRDA dated December 14, 2007 with effect from January 1, 2006 was made to adopt changes to the terms of the MRDA that had been reflected in the financial statements of the Participants. The first two recitals of this addendum stated:

Whereas each Participant holds and enjoys equitable and beneficial ownership of NN Technology as defined in the Prior Agreement,

Whereas this Addendum continues each Participant’s rights and obligations in the NN Technology,

[101] The reason for this addendum was stated in the third recital

Whereas given changes in the Nortel business, NNL and certain other Participants are seeking governmental approval of modifications to the RPSM.

[102] This was the first of two addenda that changed the way of calculating the residual profit split each year from an amortized 30% spend of each Participant each year on R&D to a five year rolling average spend by each Participant on R&D. The operative parts of this addendum did not change the operative terms of the prior MRDA relating to the licence rights granted to the participants. I do not read the first two recitals that “each Participant holds and enjoys equitable and beneficial ownership of NN Technology as defined in the Prior Agreement” and the addendum “continues each Participant’s rights … in the NN Technology” as changing anything with respect to those rights in the prior MRDA. It is how the prior MRDA defines the rights of the participants that is important.

[103] Confidentiality provisions are contained in the MRDA. The Monitor contends that because under article 6(a) the licensed participants owe a duty of confidentiality to NNL regarding the NN Technology but NNL does not owe such a duty to the Licensed Participants is an indication of the ownership by NNL of the NN Technology. The U.S. Debtors contend that because exceptions to the duty of confidentiality in article 6(d) give the right to the Licensed Participants to communicate to suppliers, customers and third persons licensing rights to use the NN Technology that they must have been given the authority to license to such third parties.

[104] I think too much is made by each side of these confidentiality provisions. There is something perhaps in each side’s argument, but I would not read article 6 as expanding on or limiting the ownership or license rights of the NN Technology. That was not its purpose. Regarding article 6(d)(iii), it begs the question as to whom the rights were given to license to third parties, and in light of the evidence of sub-licensing prior to the MRDA, to which I will refer in dealing with surrounding circumstances or the factual matrix, it is clear that NNL was a party to all such sub-licensing and NNI alone never sub-licensed.

[105] The U.S. debtors contend that what was intended by IPCo comfortably falls within the definition of a Product and that therefore what was sold to Rockstar embodied rights that NNI had. They contend:

IPCo was a licensing service business that the Participants proposed to be developed and indeed were actively developing, and which indisputably embodied the entirety of the Patent Portfolio sold to Rockstar, fits comfortably within the plain meaning of a “service” and thus the definition of “Products”.

[106] I do not agree. IPCo was considered for a time after the insolvency filings in January 2009. It could not be considered to have been part of the operating arrangements of Nortel while it carried on its business or intended to be governed by the MRDA. IPCo was not intended to be a “licensing service” business. The evidence of Sharon Hamilton, which I accept, is that the proposed business of IPCo was to use threatened or actual litigation against technology companies making their own products which arguably used or embodied NN Technology, in an attempt to encourage them to take and pay for a license to NN Technology. That was not a business contemplated in any meaningful way at any time that the MRDA or its predecessor was negotiated or signed.

[107] The economic analysis prepared by Horst Frisch in 2002 as part of its work in devising the RPSM for the MRDA referred to Nortel customers choosing Nortel products and services because Nortel is committed to using its R&D resources in providing full pro-active service and support to its customers. A functional analysis for the years 2000 to 2004 sent by Nortel to the tax authorities in 2004 said the same thing. It also stated:

“Nortel’s networking solutions generally bring together diverse networking products from its various product families, and related services, to create either a customized or “off the shelf” solution for customers. Nortel’s business consists of the design, development, manufacture, assembly, marketing, sale, licensing, servicing and support of these networking solutions”.

[108] The definition of Products in the MRDA is:

[109] Taken this definition, the license to NNI and the other participants was to “make, use…, license…sell” Products using or embodying NN Technology by, or for, the Participants. The Monitor contends that giving someone else (i.e. not any of the Participants) the right to use or embody NN Technology in their own products are not “services” within the Products definition in the MRDA. The Monitor contends that on the U.S. Debtors’ reading of the word “services” in the MRDA, NNI could have provided a “service” to competitors of Nortel by permitting them to use in the U.S. the entirety of Nortel’s patent pool to make their own products to compete with Nortel. The plain reading of the MRDA and common sense are contrary to this interpretation.

[110] I agree with the Monitor’s interpretation of the MRDA. At the time the MRDA was being considered, Nortel was not in a business of licensing its services to others for the business of others. It was providing a service to its customers to support the technology being acquired by its customers. The MRDA must be read in that context. What was contemplated for a relatively short period of time after the world wide insolvency of the Nortel Group was simply not in the cards prior to that time.

(b) The right to sue for infringement

[111] The U.S. Debtors contend that the right to sue is central to their rights as exclusive licensee in the U.S. The right to sue is contained under Article 4 which is headed Legal Title to NN Technology. The right is not contained in the exclusive or non-exclusive licenses under article 5. I cannot read this right to sue as being part of the licenses granted to the licensed participants in article 5. Articles 4 (a) and (e) are relevant, and provide:

(e) Licensed Participants have the right to assert actions and recover damages or other remedies in their respective Territories for infringement or misappropriation of NN Technology by others.

[112] This right was not contained in the prior CSA. It first appeared in the MRDA.

[113] This right in sub-article 4 (e) does not state that the Licensed Participants have the exclusive right to bring action in their territories. The exclusive rights which the Licensed Participants have are contained in the exclusive license rights in article 5. There is no provision in the MRDA that precluded NNL from suing for patent infringement in a territory in which Licensed Participants had exclusive license rights. Indeed, the limited practice in the U.S. before the MRDA was signed was that both NNL and NNI were named as plaintiffs in infringement actions. To the extent those actions can be considered to be part of the factual matrix, it explains why the right to sue granted to NNI was not an exclusive right.

[114] The right to sue for damages given to the Licensed Participants in their exclusive territories would obviously require a Licensed Participant to establish that it had been damaged. If the suit involved a breach of rights which the Licensed Participant had under its license, damages could presumably be proven. However, if the suit involved a breach of rights which the Licensed Participant did not have under its license, damages could not be proven.

[115] If a Licensed Participant were the only plaintiff, which does not appear to have ever been the case, presumably it would be open to a defendant to contend that the Licensed Participant had not suffered any damages as what was being done by the defendant was not something that the Licensed Participants could have done under its license. That defence would not likely be run if both NNL and the Licensed Participant such as NNI were plaintiffs.

[116] The Licensed Participants were not given any right to sue for damages for patent infringement in non-exclusive territories. This right was held by NNL.

(iv) Surrounding circumstances or the factual matrix

[117] What may be looked at in constructing an agreement is objective evidence of the background facts at the time of the execution of the contract. It may not include evidence of the subjective intent of a party or what a party believed a contract to mean. Whether something was or reasonably ought to have been within the common knowledge of the parties at the time of execution of the contract is a question of fact.

[118] There is an issue regarding the timing of the evidence that may be looked at. The exclusive licence to the Licensed Participants was contained in the 1985 CSA between Northern Telecom Limited [now NNL] and Northern Telecom Inc. [now NNI] signed in December 1984 and continued with no substantive changes in the 1992 CSA and in the MRDA and its later addenda. I would not, however, limit the time of the surrounding circumstances to the time that the CSAs were signed. The MRDA was made on December 22, 2004 effective January 1, 2001. Thereafter, while there were a number of changes to the MRDA in various addenda, no changes of substance were made to the operative provisions regarding the rights of the participants in NN Technology. I think the surrounding circumstances to the time of the signing of the MRDA in December 2004 can be looked at. Although there were some modifications to the MRDA after that, none involved any substantive change to the rights of NNL or to the exclusive licenses given to the Licensed Participants.

[119] There was a great deal of evidence led by the U.S. and EMEA interests as to the subjective views of the witnesses, mostly tax personnel, regarding the rights of the parties under the CSA or MRDA or what the witnesses understood the language to mean, or in one case as to the witness’s understanding of what others understood the documents to mean. Apart from the latter being inadmissible hearsay, all of this evidence was not admissible as it amounted to subjective views as to the meaning of an agreement. Nor was it admissible under the factual matrix rule permitting objective surrounding circumstances at the time of the execution of the agreement to be considered, and I do not consider it[^14]. For example, what Mr. Henderson thought about the rights under the CSA license, that he copied from an earlier version of the CSA, or what others thought the MRDA meant or what they thought the intent of it was is not to be taken into account. See Sattva, supra, at para. 59.

[120] I think it right to point out that not all of the evidence was one way. For example, the evidence of Angela De Wilton, the director of Intellectual Property in the Nortel IP law group and the director of IP strategy, was that Nortel was the owner of the patents and not just for administrative reasons. This evidence, elicited on cross-examination, also suffers from it being her subjective view of the rights of the parties under the MRDA. There were other witnesses who said much the same thing, such as Mr. Binning, the Executive Vice-President and CFO of NNC and NNL from November, 2007 to March, 2010 who said on his cross-examination that he understood that NNL owned the IP. This evidence suffers from the same problem of being a subjective view of the rights of the parties. The point is that that not all witnesses agreed with the subjective views of other witnesses.

[121] There was also some evidence led of a prior draft of the MRDA and the views of an outside tax lawyer at Oslers who acted for NNL as to the particular draft language. This evidence is also inadmissible as being a prior draft and as constituting that particular lawyer’s subjective views as to what the MRDA should contain.

[122] A great deal of evidence, including evidence of statements made to tax authorities, had to do with economic theories of transfer pricing. As the transfer pricing principles changed from a cost sharing approach to a residual profit sharing approach, the economic theories and statements to tax authorities changed. One thing that did not change from the CSA approach to the RPSM approach was the language of the license grant from NNL to the other participants. It is that language that must be considered.

(a) 1996 APA

[123] The 1992 CSA between Northern Telecom Limited (now NNL) and Northern Telecom Inc. (now NNI) was made with effect from January 1, 1992 but was not drafted until 1996 after the negotiations with the CCRA in Canada and the IRS is the U.S. in the advance pricing agreement process, so as to reflect the terms of the APA made with each of those tax authorities.

[124] The U.S. Debtors say that the APA makes clear that NNI was entitled to all of the benefits of the NN Technology in the U.S., including all sub-licensing rights. I think they draw too long a bow. The APA between Northern Telecom and the CCRA was an agreement which by its terms was to “establish a cost sharing methodology which will result in the allocation of expenses to NNI by [NNL] for R&D done by [NNL] and its subsidiaries…which will constitute reasonable amounts in the circumstances for the purposes of section 69 of the Income Tax Act”. The concern of the tax authorities was that the costs of R&D be properly allocated between NNL and NNI. The purpose of the APA was not to agree how the income of NNL and NNI was to be shared or allocated, but how to apply R&D expenditures to whatever the income was for each of NNL and NNI.

[125] Article 1.1 of the APA stated that the allocation of R&D expenses was to be determined in accordance with the cost sharing methodology described in appendix A. Appendix A is headed Cost Sharing Methodology. It contains detailed formulae to determine how R&D is to be allocated. At the outset, it has a section headed Understandings. The first understanding is that all benefit derived from R&D expenses is recognized either in the selling of a finished product to an unrelated customer or from the licensing of the technology resulting from the R&D expense (the “Benefit”) within a defined geographical market by a Cost Sharing Participant (“CSP”). It goes on to state that “[NNL], as a CSP, “is entitled to all Benefits in all geographical markets except for the part(s) thereof granted to another CSP” and “NNI is also a CSP and its geographical market is the united States of America and the Commonwealth of Puerto Rico”.

[126] This statement that NNL is entitled to all Benefits in all geographical markets except for the part(s) thereof granted to another CSP is somewhat unclear. It could refer to all Benefits except for parts thereof granted to another CSP, or to all geographical markets except for those parts granted to another CSP. The former would seem to make sense because there would be no purpose in stating that NNL was entitled to all benefits in all geographical markets except those granted to another CSP as the following sentence states that the geographical market of NNI is the U.S. and Puerto Rico. If as I read it the understanding was that NNL was entitled to all benefits except for those granted to a CSP, the document begs the question as to what benefits were granted to NNI, which is the issue in this case.

[127] It is understandable, as Mr. Henderson testified, that the parties needed to wait until the APA was settled with the tax authorities before the 1992 CSA was settled as the APA stated that it was to apply to the taxation years 1992 to 1999. That does not mean, however, that the parties needed to know how revenue was to be allocated by the APA. The purpose of the APA was to obtain an agreement from the revenue authorities how to allocate R&D costs, not revenues. This is borne out by Mr. Henderson’s admissions on cross-examination that the 1992 CSA just adopted the license language of the 1985 CSA and that all operative provisions were the same.

(b) Sub-licences

[128] Both sides refer to the evidence of sub-licensing as refuting the case of the other. In this I think they are incorrect. Not a great deal is clear from this evidence.

[129] The reply closing argument of the U.S. Debtors contains a list of “Sublicenses Involving NNI”. None were made only by NNI. Many were made by NNL and NNI and others were made by NNL on behalf of itself and its subsidiaries.

[130] The sublicenses made by NNL and NNI recited that NNL has granted to NNI “certain rights to license said patents” in the U.S. In the body of the agreement it provides that NNL and NNI “to the extent of their legal right to do so” grants a license to the licensee for the countries and jurisdictions in which Nortel now or in the future holds the Nortel patent. What rights had been licensed to NNI is not stated in the sublicense. Some sublicenses provided that the royalties were payable to NNI, with NNI having the right to direct some or all of the payments to NNL. Others, being a majority of them, provided for the royalties to be payable to NNL with NNL having the right to direct some or all of the payments to NNI. In the case of cross-license agreements, the royalties were payable to NNL and there was no provision for NNL to direct some or all of the payments to NNI.

[131] In agreements made by Nortel, defined as NNL on behalf of itself and its subsidiaries, regarding U.S. patents, it was stated that Nortel was the owner of the patents and that Nortel granted world-wide license rights for the patents. Other agreements involving other patents made by Nortel, also defined as NNL on behalf of itself and its subsidiaries, recited that Nortel was the owner of the patents. The effect of the language in this form of agreement is that the patents are owned by Nortel on behalf itself and its subsidiaries, which supports the position of EMEA that all patents were jointly owned by the MREs.

[132] The U.S. Debtors point to some evidence of certain Nortel tax personnel to explain the forms of sublicensing agreements used by Nortel. One is an e-mail exchange in 2002 involving two different views from two different tax persons, in which subjective views as to what the license in the CSAs from NNL to the other participants meant and what the theory of the sublicensing agreements was. The other is an e-mail in 2000 from someone professing not to be an expert in tax and passing on his understanding of what the tax people’s view was. Apart from the latter being hearsay and inadmissible, this e-mail evidence contains subjective views of the extent of the license in the CSAs from Nortel to the other participants in the CSAs and is inadmissible.

[133] Nortel’s IP team prepared a presentation after the sale of the business lines in connection with the stalking horse bid process for the residual intellectual property. The presentation reviewed the history of Nortel’s portfolio including its past licensing activities. It stated that Nortel previously had a small licensing group which was not a core focus of the company. There were virtually no assertions against major players, customers or partners and they focused on smaller companies with limited ability to fight back. They had earned approximately $37 million per year in royalty income. The licensing operations ceased in 2007 for budgetary reasons but in 2008 Nortel made a decision to restart the licensing organization. Mr. Binning, the Executive Vice-President and CFO of NNC and NNL from November 2007 to March 10, 2010 said that during that time Nortel was not in the license business. I take it from this evidence that for a business as large as the Nortel business, it would appear that sublicensing was an insignificant business for Nortel prior to its bankruptcy.

[134] If one follows the money from the sublicenses, the evidence is that the royalties were split on the basis of the MRDA participant’s contributions to R&D. The royalties were incorporated into the RPSM calculations even although they were not mentioned in Schedule A to the MRDA. Why this was done was not made clear by Mr. Stephens who gave the evidence of this happening. With one exception, we were not pointed to any evidence as to what was done with any royalties received prior to 2006, which is perhaps a more germane period as being prior to the signing of the MRDA. In 2004 a settlement of $35 million with Foundary Networks, Inc. was split on the basis of the RPSM.

[135] In sum, there were no sublicenses when the license was granted by NNL to NNI at the time of the 1985 CSA. There were a number after that which do not indicate any clear pattern of what sublicense rights either NNL or the other participants were recognized to have. Sublicensing was a very insignificant part of the Nortel business prior to its insolvency.

(c) Representations to tax authorities

[136] I have already discussed the 1996 APA process.

[137] In connection with the switch from a CSA approach to a RPSM approach, Horst Frisch, a leading firm of transfer pricing economists, was retained to advise Nortel. Horst Frisch prepared a report dated March 14, 2002 titled Economic Analysis of Nortel Networks' Intercompany Transactions and this report was given to the tax authorities.

[138] The U.S. interests point to a statement at page 10 of the report that stated from an economic standpoint, each participant could be considered to “own” the NT Technology. The paragraph in question made clear that what was being discussed was the situation under the CSA that existed up to the end of 1999. It stated:

Prior to 2000 Nortel shared it global R&D expenses pursuant to its R&D cost sharing arrangement ("R&D CSA"), which dates back to the mid-1970's (with several amendments). Under the arrangement, each cost sharing participant ("CSP") had the right to use the intangible property developed pursuant to the R&D cost sharing arrangement (i.e., the NT Technology") in its respective market. From an economic standpoint, each R&D cost sharing participant could be considered to "own" the NT technology as it related to its specific region.

[139] What is meant by “from an economic standpoint” each participant could be considered to “own” the NT technology as it related to its specific region is not clear. The OECD Guidelines and transfer pricing regulations in the U.S. and Canada all define intangible property to include licenses or rights to use assets. The statement of Horst Frisch that each participant had the right to use the IP and from an economic standpoint could be considered to “own” the NT technology could well have referred to owning the license rights held by each participant rather than referring to the underlying NT technology. The U.S. Debtors in their opening brief acknowledged case law to the effect that the rights an exclusive licensee holds are referred to as beneficial ownership.

[140] Horst Frisch was clearly not talking about legal rights, nor were they discussing particular language in the CSA. Even had they purported to give their views as to what legal rights the parties had under the CSA or the MRDA, which they were not, those views would not have been admissible. Horst Frisch was discussing economic theory.

[141] A few pages further in the report, when Horst Frisch were discussing what would occur under the RPSM method of allocating profits under the MRDA, they stated that the economic theory underlying the CSA was not applicable to the RPSM of allocating profits. The report stated:

As noted above, under the prior R&D CSA the CSP which ultimately made the sale to a third party in its exclusive territory was deemed to have economic ownership of the NT Technology since the third party sale attracted an R&D allocation under the CSA.

In the absence of the R&D CSA, with the two exceptions noted above, each old CSP will incur R&D expense which should entitle it to share in Nortel's global profits or losses. We have not attempted to attach these R&D expenses to the manufacturing or the distribution operation of the old CSPs since there will no longer be a formula by which global R&D expenses are shared (i.e., a third party sale will not attract an R&D allocation so it is not reasonable to assume that only the selling entity will continue to own the valuable intangibles). The amount of R&D performed is not necessarily correlated with third party sales or manufacturing activity. Rather, each entity will perform and pay for its own R&D expenses, and has the ability to sell Nortel products worldwide and share in global profits or losses.

[142] What Horst Frisch were saying was that the economic theory of the participant in a third party sale in its territory “owning” the intangibles would not apply to such a sale under the MRDA. Perhaps implicitly they were saying that the economic “ownership” of the intangibles would be owned by all of the participants in accordance with their R&D spend under the MRDA, although this is not expressly stated. If so, from an economic point of view, it would be more consistent with the position of EMEA rather than the U.S. or Canadian interests.[^15]

[143] After the APAs were applied for in 2002, the tax authorities visited various Nortel sites. They then posed a number of questions. In September 2003 Nortel send a 45 page response to these questions. One of the questions was to update the authorities on any developments since the APA submission was made and whether any changes to the proposed transfer pricing methodology were anticipated. One of the responses of Nortel had to do with restructuring charges. The U.S. interests point to a statement that the residual entities are the owners of the intangible property. The context is important to see what was being said. Included was the following:

In 2000 and later years, the telecommunications industry experienced a decline in demand unlike any other substantial industry in modem history. For that reason, there does not appear to be any precedent for analyzing the above issues. Accordingly, a reliance on basic economic principles was deemed necessary.

In an arm's length situation, it was determined that the residual entities would agree to reimburse the distribution entities for a portion of their restructuring charges rather than have those entities become insolvent and forced into receivership. Generally, the underlying economic rationale for this argument is this: the residual entities, as the owners of the intangible property, as well as the manufacturers of the tangible goods, would recognize that its distribution network is critically necessary for their long-term survival. Should members of the distribution network become insolvent/cease operations, the residual entities' ability to offer their products for sale may be severely impacted. Therefore, it is in their best economic interests to ensure that a strong global distribution network exists.

[144] This is a discussion of economic theory. It cannot be construed as a discussion of legal principles or the meaning of the MRDA. The reference to being owners of the intangible property could well be a reference to license rights rather than the underlying intangibles, as license rights are intangible property.

[145] Dr. Timothy Reichert, a transfer pricing expert called by the Monitor, made the following statement about economic ownership as considered in transfer pricing, a statement not contradicted by any witness:

A central concept in transfer pricing is that of “economic ownership” (referred to, alternatively, as “beneficial ownership,” and simply “ownership”). Economic ownership is not a reference to ownership in a legal sense, but rather refers to a party’s right to benefit from an income stream attributable to a defined undertaking or activity.

[146] This statement cannot be disregarded. An economic right to an income stream attributable to a defined undertaking or activity requires one to know what is the defined undertaking or activity. The economic statements made to the taxing authorities did not purport to define the precise limits to the license granted by NNL to the participants under either the CSA or the MRDA. As seen, the statements made to the taxing authorities did not at all make clear what rights were being referred to and in particular, whether the “economic or beneficial ownership” was in the underlying NN Technology or in the license rights to that technology. They cannot be taken as statements that under the MRDA the licensees legally owned the NN Technology.

[147] To the contrary, Mr. Weisz, the Leader of International Tax at NNI who was involved in Nortel’s transfer pricing policies, and who was asked by Mr. Dolittle, the VP of Tax for Nortel to become involved in the APA process that had stalled, told the IRS during that process that it was NNL that was the legal owner of the IP.

[148] It must also be recognized that the APA process with the tax authorities was to arrive at an agreement with them regarding the Nortel operating business. It was an operating business with profits and losses that the tax authorities were interested in. This was the case both for the APA processes for both the CSA and MRDA. There were no discussions with the tax authorities as to what would happen on the insolvency of the entire business of Nortel. The discussion of the economic theory of economic or beneficial ownership must be considered in that light. They were not discussing such a theory in so far as it might apply to the situation of a cessation of business as occurred with Nortel.

[149] The issue in this case has to do with the breadth of the licenses granted to the Licensed Participants and whether that included the right to sublicense the residual patent portfolio that was eventually sold to Rockstar. However at the time of the APA processes, sublicensing was a miniscule part of the business of Nortel and not surprisingly I have been pointed to no presentation to the tax authorities that discussed this issue. It was not on the Nortel radar.

[150] The MRDA was provided to the tax authorities. They also had the prior CSA. The U.S. interests do not say that NNI rights were obtained other than in the CSA and then the MRDA. The tax authorities had these agreements and were able to read them, and were in as good a position as anyone to form their own view of what the agreements did or did not do. It cannot be suggested that the tax authorities did not understand transfer pricing. It was their business to know it.

[151] There is evidence relied on by the EMEA debtors to support their position that the proceeds of the sale of the residual IP should be allocated in accordance with the relative expenditure on R&D by NNL and the licensed participants. After the APS application was filed by Nortel with the tax authorities in 2002, planning for a June 19, 2002 joint meeting with the three tax authorities took place in earnest, including the preparation of answers to questions Nortel anticipated might be raised by the tax authorities. In preparation for the meeting, Nortel engaged advisors from Deloitte & Touche LLP, KPMG LLP, Horst Frisch, and Sutherland Asbill & Brennan LLP to assist. Mr. O’Connor of Deloitte’s prepared the following answer to an anticipated question regarding the sale of any IPCo, an answer that was circulated and agreed before the meeting:

[Q:] How does Nortel propose to account for any future sale of intellectual property developed prior to or during the term of the APA? Which entities are considered the legal owner of IP and which are considered the economic owners?

[A:] Proceeds from the sale of IP will be allocated to residual profit split participants on the basis of their economic ownership of the IP – that is, on the basis of their share of total R&D capital stock in the year of sale.

[152] The document was taken to the joint meeting. There is no evidence one way or the other as to whether the question was asked. While the question does not deal with what would happen if all of the Nortel IP was sold on a world-wide insolvency of Nortel, it is evidence that at the time of the APA process, Nortel was prepared to say that any sale of IP would be split in accordance with the RPSM in the year of the sale.

(d) Avoiding permanent establishment status in the U.S.

[153] The U.S. interests contend that it was important to NNI not to be considered a U.S. resident for U.S. tax purposes and thus not have “permanent establishment” in the U.S. It is contended that it was this concern that drove separate legal entities to be set up for each country and for the license to be exclusive to the Participants for their territory so that the Participants would be the only ones dealing with customers in that territory.

[154] Taken that as the situation, I do not see that it gets very far. There is no question that the exclusive license gave NNI the exclusive right to sell Nortel products in the U.S. The important issue in this case is what rights NNI had to sub-license and whether it was restricted to Nortel “Products” as defined in the MRDA. Sub-licensing was a miniscule part of the business of Nortel and there is no evidence at all that sublicensing issues drove the setting up of companies in different jurisdictions.

(e) Patent litigation

[155] I have previously referred to the patent litigation that had taken place prior to the MRDA. Both NNL and NNI were plaintiffs in the actions in the U.S. Why that was so does not really matter. It explains why the right given to licensed participants to sue in their territories for damages for patent infringement was not an exclusive right.

[156] As to what was done with the proceeds of patent litigation, there was a settlement in October, 2004 of an action commenced by NNL and NNI against Foundary Networks, Inc. $35 million was paid for past infringement and for future royalties under a license agreement. The entire payment was treated as royalty income and allocated to NNL and the Licensed Participants in accordance with the RPSM in the MRDA.

(f) Conclusion of factual matrix evidence

[157] I do not consider the surrounding circumstance or factual matrix evidence to provide much clear assistance in construing the meaning of the terms in the MRDA. Even if it did, I would be required to be guided by the dictates of Sattva that while the surrounding circumstances will be considered in interpreting a contract and the goal of examining such evidence is to deepen a decision-maker's understanding of the mutual and objective intentions of the parties as expressed in the words of the contract, they must never be allowed to overwhelm the words of the contract and the interpretation of a contract must always be grounded in the text. While the surrounding circumstances are relied upon in the interpretive process, courts cannot use them to deviate from the text such that the court effectively creates a new agreement.

(v) Commercial reasonableness

[158] The U.S. interests and EMEA say that the Monitor’s interpretation of the MRDA is commercially unreasonable. They contend that no party at arm’s length would agree to spend large amounts to develop patents but only one party would be entitled to all of the proceeds from the sale of those patents.

[159] It must be remembered that the MRDA and its predecessor CSA were drafted to come to terms with the tax authorities. The parties to the negotiations were Nortel on the one hand and the tax authorities on the other. The resulting CSA and then MRDA were operating agreements premised on cost sharing under the CSA and profit or loss sharing under the MRDA. The tax authorities were interested in the tax that each Nortel entity would pay each year. The tax authorities dealt in only limited periods of time. The 1992 CSA was settled with the tax authorities only in 1996, yet in 1999 they made it clear they wanted Nortel to abandon the CSA agreement and instead change to a RPS method of transfer pricing.

[160] Nortel and the tax authorities were not negotiating on what would happen if Nortel stopped operating or in the event of a world-wide insolvency of Nortel. More particularly, they were not negotiating on how the proceeds of the sale of the entire Nortel world-wide patent portfolio would be allocated amongst the various Nortel companies in the event of the insolvency of Nortel. That was not discussed. This is not surprising because, as Dr. Eden testified, transfer pricing rules were developed only in connection with ongoing entities for purposes of determining their corporate tax.

[161] The issue of commercial reasonableness must be considered in the context of who was involved in the preparation of the MRDA. It was not the technology people. Mr. Brian McFadden, the Chief Technology Officer of Nortel at the time the MRDA was drafted and signed, was not consulted about its terms and never heard of the MRDA while at Nortel. Ms. Angela de Wilton, the Nortel Director of Intellectual Property had no recollection of ever seeing the MRDA. In order to make the argument that it would have been commercially unreasonable for a Nortel company to agree to do R&D leading to patents and not be paid for the patents on the sale of the business, one would think that the people responsible for the R&D would have at least known of the agreement and its terms. The fact that they did not indicates that a trade-off of R&D for future receipts on a sale of the business was not on the radar screen at all so far as the operating people were concerned. The language of the MRDA was all tax driven.

[162] So far as what was on the radar of the tax people at Nortel at the time the terms of the MRDA were settled, in so far as the quid pro quo for doing R&D was concerned, the MRDA expressly provided in Article 2(c) that any compensation for R&D was to be based solely on the RPSM allocation. It stated:

(c) All costs incurred directly or indirectly by each Participant for R&D Activity shall be borne exclusively by it. Any reimbursement for costs including any other compensation shall be provided to such Participant for its R&D Activity solely as provided in Article 3 below.

[163] Article 3 that the annual sharing of profits or losses under the residual profit split method was what was to be received for each participants R&D that year. It stated:

Article 3 – R&D Activity Payments

[164] In the context of what the parties were dealing with in the MRDA, I do not see how it can be said that it was commercially unreasonable for them to agree that in return for doing R&D each year, they would share only in the profits or losses in accordance with the RPSM allocation. That is all they had in mind. While Nortel had suffered losses by 2004 when the MRDA was signed, there is no evidence that Nortel expected to have only losses in the future. To the contrary, the operating people at Nortel expected that Nortel would return to profitability.

[165] Mr. Henderson testified that a bankruptcy or insolvency of Nortel was not in their minds at the time the RPSM in the MRDA was created. The fact that they did not consider or provide what was to happen to the proceeds if all of the IP was sold after a world-wide insolvency does not make the agreement commercially unreasonable. The time for considering whether an agreement properly interpreted is commercially reasonable or unreasonable is surely the time when it was agreed, not in hindsight.

[166] The U.S. Debtors called Dr. Catherine Tucker, a transfer pricing expert, whose evidence was to the effect that under the Monitor’s interpretation of the MRDA, the Licensed Participants would lack appropriate incentives to undertake expensive and speculative R&D for the next potential generation of products. I do not think her evidence helpful. It is really an inadmissible subjective view as to how the MRDA license should be interpreted.

[167] There is no basis for Professor Tucker’s assumption that the MRDA was intended to create incentives for the Licensed Participants to make forward-looking innovations. The fact that Mr. McFadden, the Chief Technology Officer of Nortel at the time of the MRDA, was not consulted about the MRDA and knew nothing about it belies any such assumption. Professor Tucker’s assumption also ignores the way in which R&D was carried out at Nortel.

[168] The majority of Nortel R&D was directed by the various Business Lines, which had to prepare annual R&D plans for approval. The remaining R&D, the advanced technology research (the “leap from one S-curve to the next” that Professor Tucker describes), was coordinated by the Chief Technology Officer. The evidence of Mr. McFadden was that all advanced technology programs were based in Ottawa and were operated by NNL. While product development R&D groups within each of Nortel's lines of business reported directly to the heads of their business units, the advanced technology programs personnel within each line of business reported directly to the CTO's office at NNL. The R&D was not the bailiwick of any Licensed Participant.

(vi) Conclusions on the meaning of the MRDA

[169] I interpret the MRDA, and find, that under it, and while Nortel operated as a going concern business, NNL had all ownership interests of the NN Technology subject to grants, being (i) the grant to each Licensed Participant of a non-exclusive right to assert actions and recover damages in their territory under article 4(e) and (ii) the grant of exclusive and non-exclusive licenses to the Licensed Participants under article 5(a).

[170] The licenses under article 5(a) were not licenses of all rights to the NN Technology but were subject to field of use restrictions that gave the Licensed Participants the right to use the NN Technology to make, use or sell Products as defined in the MRDA, which meant products, software or services that were made or sold by, or for, any of the Licensed Participants. The Products must have been created or marketed by or for the Nortel Group. No product that was part of a third party’s business rather than the business of Nortel fell within the definition of Products. The business considered by IPCo was not covered by the licenses. The Licensed Participants’ rights to sublicense were subject to these restrictions.

Applicability of the MRDA to the allocation issues

[171] Nortel Networks UK Pension Trust Limited and the Board of the UK Pension Protection Fund (the “UKPC”) contend that the MRDA was never intended to provide an answer to the question of how to allocate among the bankrupt estates the proceeds of the sale of the Nortel Group’s assets following the world-wide insolvency of Nortel.

[172] I agree. The MRDA was an operating agreement and was not intended to, nor did it, deal with the disposal of all of Nortel’s assets in a situation in which no revenue was being earned and no profit or losses were occurring. The MRDA provided in its opening line that it was an agreement “confirming and formalizing the operating arrangements” of the parties.

[173] There is a provision in schedule A to the MRDA added in the third addendum effective January 1, 2006 but signed by the parties late in late December 2008 or early January 2009 that indicates that sales of property were not intended to be dealt with under the MRDA. That schedule A provided that in dealing with the calculations of the Nortel earnings/losses to be used in the RPSM calculation, there was to be deducted “gain/loss on the sale of business”. A gain or loss would normally be taken into account on the particular company’s statement of profit and loss and the Participants decided they did not want any such gain or loss to influence the calculation of profits or losses for the purposes of calculating the allocation of profits or losses in the RPSM calculation. Mr. Orlando testified that the sale of a business was seen to be a non-operating activity. This provision is an indication that the MRDA was not intended to deal with the sale of any assets, let alone the world-wide assets of the Nortel Group.

[174] The MRDA and its predecessor Cost Sharing Agreements (“CSAs”) were developed for and driven by transfer pricing concepts. They were drafted to come to terms with the tax authorities. The MRDA expressly provided in a recital that the calculation of the RPSM might have to be adjusted as a result of its review by the tax authorities. The MRDA was drafted by tax lawyers and tax advisors. The primary external counsel involved, and lead drafter of the MRDA, Giovanna Sparagna, testified that the MRDA is “primarily focused on transfer pricing,” which is “part of tax law,” and it is “primarily [a] tax law document[].” The MRDA was signed on behalf of NNL by John Doolittle, Nortel’s Vice-President of Tax. All parties acknowledge that the MRDA was a tax-driven document designed to implement Nortel’s transfer pricing policies.

[175] Following the insolvency proceedings on January 14, 2009, no transfer pricing payments were made under the MRDA. The two special cash payments made by NNI to NNL were made under different agreements, being the IFSA dated June 9, 2009 and the FCFSA dated December 23, 2009.

[176] Dr. Eden, who testified on behalf of the U.S. Debtors, testified that transfer pricing was only for ongoing businesses. She also testified that she saw the MRDA as a transfer pricing document and that the RPSM method contained in it was only used for corporate income tax purposes. She and other transfer pricing experts such as Dr. Richard Cooper, Dr. Steven Felgran and Dr. Timothy Reichert testified to the effect that transfer pricing does not address entitlement to the proceeds of the sale of assets on insolvency.

[177] I accept that the MRDA was a transfer pricing document created for tax purposes. The licenses were a part of it. The licenses granted under it were never dealt with separately from the MRDA. Their only purpose was to support the intended tax treatment resulting from the MRDA.

[178] It can perhaps be argued that under article 9 of the MRDA the rights of NNL under article 4 and of the Licensed Participants under article 5 continued on the expiry or termination of the agreement, indicating a purpose other than tax that survives the insolvency of the Nortel enterprise. I would not construe those provisions that way.

[179] The relevant provisions of article 9 of the MRDA provide:

Article 9 — Duration and Continuing Rights and Obligations

(a) This Agreement shall be effective from January 1, 2001 until December 31, 2004, provided however that this Agreement will automatically renew for additional and unlimited one-year terms until terminated by the mutual written consent of all Participants.

(b) Upon the expiry or termination of this Agreement as provided herein, each Licensed Participant shall be deemed to have acquired a fully paid up license permitting it to continue to exercise the rights granted to it herein, and, in particular, the rights granted to it in Article 5 as though this Agreement had continued.

(c) The provisions of Article 4 (Legal Title to NN Technology) with respect to NN Technology acquired or developed pursuant to this Agreement from the Effective Date of this Agreement up to and including its expiry or termination date, Article 6 (relating to confidentiality) and Article 7 (relating to liability) shall survive notwithstanding the expiry of this Agreement, or any termination of this Agreement for any cause whatsoever.

[180] Under article 9, the MRDA automatically renewed after 2004 unless terminated by mutual consent of all parties to it. If terminated, the Licensed Participants were to be deemed to have acquired a fully paid up license “permitting it to continue to exercise the rights granted to it herein… as though this Agreement had continued”. This provision by its terms contemplated the business of the Licensed Participants continuing to operate. It did not contemplate a situation in which all of the Licensed Participants liquidated their assets and went out of business.

[181] The CSAs contained a similar provision regarding the rights of the Licensed Participants on termination to a fully paid up license. At the end of 1999, the tax authorities did not want to renew the APAs and they encouraged Nortel to adopt a RPSM. In December, 2001, Nortel’s CSAs for R&D were terminated effective January 1, 2001. In spite of the termination of these agreements, Nortel continued to operate and it was only on December 22, 2004 after negotiations with the tax authorities that the MRDA was executed with an effective date of January 1, 2001. The MRDA stated that it confirmed and formalized the operating arrangements of the Participants as and from that date. That is, the license rights under the CSAs continued to be used in accordance with the terms of the CSAs, and Nortel’s tax advisors told that to the tax authorities on April 26, 2004. This was contemporaneous with the MRDA being settled.

[182] One can see from this that the purpose of continuing rights under article 9 of the MRDA after a termination was to permit the Participants to continue operating, during which a new agreement would have to be negotiated. Nortel was a multi-national enterprise that had to live with tax authorities where it operated and could not live without a transfer pricing agreement of some kind. As previously discussed, there was no thought at the time of the MRDA being settled that Nortel was not going to return to profitability.

[183] Article 11(a) of the MRDA provided that any Participant that was not a party to an APA with the tax authorities could elect to withdraw from the MRDA. Article 11(c)(iv) of the MRDA provided that it was a defaulting event if one of the Participants became insolvent, in which case the Participant automatically was terminated from participation in the agreement. A fourth addendum was made to the MRDA effective December 31, 2008 and signed in early January 2009. It was headed Standstill Provision and provided that in the event of an occurrence of an event described at Section (sic) 11(c)(iv), i.e. if a Participant became insolvent, (i) no Participant effected by such insolvency shall be automatically terminated from participating in the agreement, (ii) no Participant shall elect to withdraw from the agreement under Article 11(a), and (iii) NNL would have the right in its sole discretion to terminate participation in the MRDA of any Participant affected by such event, i.e. of a defaulting Participant by reason of its insolvency.

[184] This provision did not contain provisions expressed to apply in the event of a world-wide insolvency of all of the Nortel companies. It contained provisions of a standstill nature dealing with a situation in which one Participant became insolvent. It was obviously designed to prevent a Participant from declaring insolvency and then trying to take positions contrary to other Participants and to prevent the other Participants in such an event trying to take positions contrary to other Participants. I do not read this provision as an indication that in the event of a world-wide insolvency of all of the Nortel companies with no operations, the agreement was to continue to govern the affairs of a non-operating enterprise. Had that been the parties’ intention, they could have said so in the addendum. They did not.

[185] I conclude that the circumstances surrounding the creation of the MRDA lead to no other result but that the construct of legal title to the NN Technology being in NNL in return for NNL granting exclusive licenses to the Licensed Participants was only for the purpose of supporting the proposed method to split profits or losses on a tax efficient basis while Nortel operated as a going concern business. The agreement in its application was intended to apply only to Nortel while it operated and not to deal with rights after Nortel and its subsidiaries stopped operating its businesses.

EMEA position on ownership of the Nortel IP

[186] The EMEA Debtors’ position is different from the position of the Canadian and U.S. Debtors. They say that the Participants, or RPEs, have joint ownership of all Nortel IP under common law principles by reason of the IP belonging to the RPEs that employed the inventors. They say that the MRDA recognizes that joint ownership and that the joint ownership should be the basis for allocating the proceeds of the sale of the Nortel residual IP.

[187] The basic premise of the EMEA Debtors’ argument is that the Participants or RPEs were joint owners of all Nortel IP by reason of law. They argue that under Canadian law, the inventor is the first owner of an invention and is the legal title holder entitled to apply for any related patent. However, where the inventor is employed to invent, as the Nortel Group’s researchers were, then the employer by operation of law beneficially owns any resulting IP. While the employee inventor who is listed on the patent application holds legal title, the employer is the beneficial owner. Given the integrated nature of Nortel and its R&D created in multiple jurisdictions, the EMEA Debtors argue that each participant beneficially owned not the IP created in its jurisdiction, but rather a share of the indivisible pool of the Nortel Group’s IP.

[188] Canadian and U.K. law appears to support the principle that where an employee creates an invention as part of his or her employment, the employer is the beneficial owner of the patent.[16] U.S. law is otherwise. Under U.S. law, unless there is agreement to the contrary, it is the inventor and not the employer who is the owner of his or her invention until he or she assigns it[17].

[189] All Nortel employees, whether employed by NNL or a subsidiary, were required to assign directly or indirectly to NNL any intellectual property which they generated during the course of their employment. At least 98% of the patents and patent applications sold to Rockstar had been assigned by the inventors to NNL.

[190] The EMEA Debtors say that while the inventors assigned their rights to NNL, the subsidiaries that employed the inventors did not. Thus they say that NNUK, the employer of inventors in the U.K., continued to have beneficial ownership of the patents for inventions created by its employees. I do not accept that. NNUK was required under article 4 (b) of the MRDA to execute such documents as NNL reasonably required to give effect to article 4 (a), which provided for legal title to NN Technology to be vested in NNL. In light of that obligation, NNUK is in no position now to say that the assignment of the IP from its employees to NNL was ineffective.

[191] This argument of the EMEA Debtors would not apply to NNI in the U.S. as under U.S. law NNI did not have any such common law rights to IP developed by its employees who assigned their rights to NNL. Nor would it apply to patents invented by employees of NNL who assigned their rights to NNL. The EMEA Debtors' argument, if accepted, would mean that NNUK would only have rights to the IP developed by its employees and would have no joint ownership interest in the IP developed by employees of NNL, NNI, NNSA or Nortel .

[192] I cannot accept the joint ownership theory of the Nortel IP or use that theory as a basis for allocating the proceeds of sale of the Nortel IP assets.

Appropriate method to allocate the proceeds of sale

[193] While the Monitor on behalf of the Canadian Debtors and the U.S. Debtors take diametrically different views as to their rights under the MRDA, they each look to the MRDA and the rights they say were given to them as the basis of their allocation positions. I have not accepted their position that they obtained rights under the MRDA that determine their right to the proceeds of the sale of Nortel’s assets.

[194] Nor have I accepted the position of the EMEA Debtors that the RPEs have joint ownership of all Nortel IP under common law principles recognized in the MRDA and that such joint ownership should be the basis for allocating the proceeds of the sale of the Nortel residual IP.

[195] Without the MRDA to govern the allocation, and without the joint ownership theory of the EMEA Debtors, the issue becomes one of deciding what metric should be used to allocate the proceeds of sale.

[196] In so far as the IP is concerned, while the patents were registered in the name of NNL, I would not for that reason hold that NNL is entitled to the proceeds of the IP sales. The patents and application rights to apply for patents were held in the name of IP for administrative purposes. It was best practices in a multi-national enterprise to have all patents assigned to one company, in this case to NNL, as explained by Ms. Anderson and Ms. De Walton, and made management of the portfolio much easier. While these witnesses expressed subjective views that it was NNL who owned the patents, these views are not determinative, as acknowledged in the Monitor’s reply brief at paras 65-66.

[197] This was not one corporation and one set of employees inventing IP that led to patents. Nortel was a highly integrated multi-national enterprise with all RPEs doing R&D that led to patents being granted. It was R&D that drove Nortel’s business. R&D and the intellectual property created from it was the primary driver of Nortel’s value and profits. All parties agree on that. It would unjustly enrich NNL to deprive all of the other RPEs of the work that they did in creating the IP just because the patents were registered in NNL’s name.

[198] Canadian law permits recovery for unjust enrichment whenever a plaintiff can establish three elements: an enrichment of or benefit to the defendant, a corresponding deprivation of the plaintiff, and the absence of a juristic reason for the enrichment: Kerr v. Baranow, 2011 SCC 10 at para. 32.

[199] U.S. law provides that unjust enrichment occurs where a party obtains a benefit which, under the circumstances and in light of the relationship between the parties, it would be inequitable to retain: Counihan v. Allstate Ins. Co., 194 F.3d 357, 361 (2d Cir. 1999). It requires the retention of a benefit to the loss of another or the retention of money or property of another against the fundamental principles of justice or equity and good conscience. It is not available if there is a contract that governs the relationship between the parties that gives rise to the claim: Kuroda v. SPJS Holdings, L.L.C., 971 A.2d 872, 891(Del. Ch. 2009).

[200] On either test, I find that NNL would be unjustly enriched by being entitled to all of the proceeds of the sale of Nortel IP at the expense of the other RPEs who contributed to the creation of that IP just because the patents were registered in NNL’s name. It would be inequitable. There would be no juristic reason for the enrichment as the MRDA as I have interpreted it does not deal with the allocation rights of the parties in this world-wide insolvency of Nortel.

[201] It would also unjustly enrich NNI if it were to be allocated the amount from the IP sales that it claims based principally on its revenues, which is the basis of the claim by the U.S. Debtors. NNI was able to sell Nortel products based on the R&D and resulting IP performed by other RPEs.

[202] This is an unprecedented case involving insolvencies of many corporations and bankrupt estates in different jurisdictions. The intangible assets that were sold, being by far the largest type of asset sold, were not separately located in any one jurisdiction or owned separately in different jurisdictions. They were created by all of the RPEs located in different jurisdictions. Nortel was organized along global product lines and global R&D projects pursuant to a horizontally integrated matrix structure and no one entity or region was able to provide the full line of Nortel products and services. R&D took place in various labs around the world in a collaborative fashion. R&D was organized around a particular project, not particular geographical locations or legal entities, and was managed on a global basis. The fact that Nortel ensured that legal entities were properly created and advised in the various countries in which it operated in order to meet local legal requirements does not mean that Nortel operated a separate business in each country. It did not.

[203] Nortel’s matrix structure also allowed Nortel to draw on employees from different functional disciplines worldwide (e.g. sales, R&D, operations, finance, general and administrative, etc.), regardless of region or country according to need. Individuals could be part of a team with horizontal responsibility without removing them from their respective position vertically (or departmentally) within the Nortel group.[18]

[204] In these circumstances, what principles should be applied to determine the allocation of the proceeds of the asset sales? In my view, doing what is just in the unique circumstances of this case should govern the allocation.

[205] A court has wide powers in a CCAA proceeding to do what is just in the circumstances. Section 11(1) provides that a court may make any order it considers appropriate in the circumstances. Although this section was provided by an amendment that came into force after Nortel filed under the CCAA, and therefore by the amendment the new section does not apply to Nortel, it has been held that the provision merely reflects past jurisdiction. In Century Services, Deschamps J. stated:

65 I agree with Justice Georgina R. Jackson and Professor Janis Sarra that the most appropriate approach is a hierarchical one in which courts rely first on an interpretation of the provisions of the CCAA text before turning to inherent or equitable jurisdiction to anchor measures taken in a CCAA proceeding (see G. R. Jackson and J. Sarra, "Selecting the Judicial Tool to get the Job Done: An Examination of Statutory Interpretation, Discretionary Power and Inherent Jurisdiction in Insolvency Matters", in J. P. Sarra, ed., Annual Review of Insolvency Law 2007 (2008), 41, at p. 42). The authors conclude that when given an appropriately purposive and liberal interpretation, the CCAA will be sufficient in most instances to ground measures necessary to achieve its objectives (p. 94).

67 The initial grant of authority under the CCAA empowered a court "where an application is made under this Act in respect of a company ... on the application of any person interested in the matter ..., subject to this Act, [to] make an order under this section" (CCAA, s. 11(1)). The plain language of the statute was very broad.

68 In this regard, though not strictly applicable to the case at bar, I note that Parliament has in recent amendments changed the wording contained in s. 11(1), making explicit the discretionary authority of the court under the CCAA. Thus in s. 11 of the CCAA as currently enacted, a court may, "subject to the restrictions set out in this Act, ... make any order that it considers appropriate in the circumstances" (S.C. 2005, c. 47, s. 128). Parliament appears to have endorsed the broad reading of CCAA authority developed by the jurisprudence. (underlining added)

[206] This Court has a broad inherent jurisdiction to make orders as required to fill in gaps or lacunae not covered by specific provisions in the CCAA. As a superior court of general jurisdiction, the Superior Court of Justice has all of the powers that are necessary to do justice between the parties. Except where provided specifically to the contrary, the Court’s jurisdiction is unlimited and unrestricted in substantive law in civil matters. See 80 Wellesley St. East Ltd. v. Fundy Bay Builders Ltd. et al., 1972 CanLII 535 (ON CA), [1972] 2 O.R.280 (C.A.) at para. 9. See also TCR Holding Corp. v. Ontario, 2010 ONCA 233 at para. 26, Beach v. Moffatt (2005), 2005 CanLII 14309 (ON CA), 75 O.R. (3d) 383 (C.A.) at para. 8, J.M. v. W.B. (2004), 2004 CanLII 8541 (ON CA), 71 O.R. (3d) 171 (C.A.) at para. 43 and McVan General Contracting Ltd. v. Arthur (2002), 2002 CanLII 45035 (ON CA), 61 O.R. (3d) 240 (C.A.) at para. 56.

[207] In Century Services Inc. v. Canada (Attorney General), 2010 SCC 60 at paras. 57-61, it was recognized by the Supreme Court and stated by Justice Deschamps that the CCAA is skeletal in nature and does not contain a comprehensive code that lays out all that is permitted, that the incremental exercise of judicial discretion with respect to the CCAA has been adapted and has evolved to meet contemporary business and social needs and that when large companies encounter difficulty and reorganizations become increasingly complex, CCAA courts have been called upon to innovate accordingly.

[208] In this case, insolvency practitioners, academics, international bodies, and others have watched as Nortel’s early success in maximizing the value of its global assets through cooperation has disintegrated into value-erosive adversarial and territorial litigation described by many as scorched earth litigation.[^19] The costs have well exceeded $1 billion. A global solution in this unprecedented situation is required and perforce, as this situation has not been faced before, it will by its nature involve innovation. Our courts have such jurisdiction.

[209] It is a fundamental tenet of insolvency law that all debts shall be paid pari passu and all unsecured creditors receive equal treatment. See Shoppers Trust Co. (Liquidator of) v. Shoppers Trust Co. (2005), 2005 CanLII 7878 (ON CA), 74 O.R. (3d) 652 (C.A.) at para. 25, per Blair J.A., Indalex Ltd. (Re) (2009), 55 C.B.R. (5th) 64 (Ont. S.C.), at para. 16 per Morawetz J. and my comments in Nortel Networks Corp. (Re) (2014), 2014 ONSC 4777, 121 O.R. (3d) 228 at para.12. A pro rata allocation in this case goes partway towards such a result.

[210] According to the various protocols, the task in this proceeding is to determine the amount that is to be allocated to each of the Canadian, U.S. and EMEA Debtors' Estates. I do not read the protocols or the IFSA as precluding a pro rata allocation. While payment to the Selling Debtors is to be made from the $7.3 billion in the lockbox funds, neither the protocols nor the IFSA determine how the allocation is to be made.

[211] Directing a pro rata allocation will constitute an allocation as required. Once the lockbox funds have been allocated, it will be up to each Nortel Estate acting under the supervision of its presiding court to administer claims in accordance with its applicable law. A pro rata allocation can be achieved by directing an allocation of the lockbox funds to each Debtor Estate based on the percentage that the claims against that Estate bear to the total claims against all of the Debtor Estates.

[212] It is argued that a pro rata allocation would constitute an impermissible substantive consolidation of the Estates, or as put by the U.S. Debtors, an impermissible “global substantive consolidation”. I do not agree. A pro rata allocation in this case would not constitute a substantive consolidation and, even if it did, it would in my view be permissible within established case law.

[213] In a liquidation or reorganization of a corporate group, the doctrine of substantive consolidation has emerged in order to provide a mechanism whereby the court may treat the separate legal entities belonging to the corporate group as one. In particular, substantive consolidation allows for the combination of the assets and liabilities of two or more members of the group, extinguishes inter-company debt and creates a single fund from which all claims against the consolidated debtors are satisfied. In effect, under substantive consolidation, claims of creditors against separate debtors instantly become claims against a single entity.

[214] A pro rata allocation in this case would not constitute a substantive consolidation, either actual or deemed, for a number of reasons. First, and most importantly, the lockbox funds are largely due to the sale of IP and no one Debtor Estate has any right to these funds. It cannot be said that these funds in whole or in part belonged to any one Estate or that they constituted separate assets of two or more Estates that would be combined. Put another way, there would be no “wealth transfer” as advocated by the bondholders. The IFSA, made on behalf of 38 Nortel debtor entities in Canada, the U.S. and EMEA, recognized that the funds would be put into a single fund undifferentiated as to the Debtor Estates and then allocated to them on some basis to be agreed or determined in this litigation. Second, the various entities in the various Estates are not being treated as one entity and the creditors of each entity will not become creditors of a single entity. Each entity remains separate and with its own creditors and its own cash on hand and will be administered separately. The inter-company claims are not eliminated.

[215] Even if it could be said that a pro rata allocation involved substantive consolidation, which it cannot, I do not see case law precluding it in the unique circumstances of this case international case. Even in domestic cases, CCAA plans involving substantive consolidation are not unknown.

[216] In Canada, neither the CCAA nor the BIA contains express provisions authorizing substantive consolidation. Similarly, the U.S. Bankruptcy Code does not explicitly permit substantive consolidation. However, courts in both jurisdictions have rendered consolidating orders on the basis of their equitable jurisdiction. See See M. MacNaughton and M. Azoumanidis, Substantive Consolidation in the Insolvency of Corporate Groups: A Comparative Analysis, Annual Review of Insolvency Law, 2007, J. Sarra, ed. (Carswell: 2008).[^20]

[217] In Rescue! The Companies' Creditors Arrangement Act, by Dr. Janis Sarra, Carswell 2007, the grounds for permitting substantive consolidation were described as follows at page 242:

The court will allow a consolidated plan of arrangement or compromise to be filed for two or more related companies in appropriate circumstances. For example, in PSINet Ltd. the Court allowed consolidation of proceedings for four companies that were intertwined and essentially operated as one business. The Court found the filing of a consolidated plan avoided complex issues regarding the allocation of the proceeds realized from the sale of the assets, and that although consolidation by its nature would benefit some creditors and prejudice others, the prejudice had been ameliorated by concessions made by the parent corporation, which was also the major creditor. Other cases of consolidated proceedings such as Philip Services Canadian Airlines, Air Canada and Stelco, all proceeded without issues in respect of consolidation.

Generally, the courts will determine whether to consolidate proceedings by assessing whether the benefits will outweigh the prejudice to particular creditors if the proceedings are consolidated. In particular, the court will examine whether the assets and liabilities are so intertwined that it is difficult to separate them for purposes of dealing with different entities. The court will also consider whether consolidation is fair and reasonable in the circumstances of the case.

[218] In Re Lehndorff General Partner Ltd. (1993), 17 C.B.R. 3d 24, Justice Farley held that a consolidated plan was appropriate, noting that there was significant intertwining of the debtor companies, including multiple instances of inter-company debt, cross-default provisions and guarantees and the existence and operation of a centralized cash-management system. All of these features were present in Nortel.

[219] In Re PSINet Ltd. (2002), 2002 CanLII 49587 (ON SC), 33 C.B.R. 4th 284, Justice Farley noted that a plan of arrangement based on substantive consolidation avoided the “complex and likely litigious issues” that could result from the allocation of the proceeds of the sale of substantially all of the debtor companies’ assets. He also noted that the consolidated plan reflected the intertwined nature of the debtors and their operation. In that case, Farley J. stated that the overall effect of a consolidation was required:

In the circumstances of this case, the filing of a consolidated plan is appropriate given the intertwining elements discussed above. See Northland Properties Ltd., Re, 1988 CanLII 2924 (BC SC), 69 C.B.R. (N.S.) 266 (B.C.S.C.), affirmed (B.C.C.A.), supra, at p. 202; Lehndorff General Partner Ltd., Re, 17 C.B.R. (3d) 24 (Ont. Gen. Div. [Commercial List]) at p. 31. While consolidation by its very nature will benefit some creditors and prejudice others, it is appropriate to look at the overall general effect.

[220] In Northland Properties, a case involving a proposed plan for several companies that operated as a single entity, Justice Trainor considered the tests for permitting a substantive consolidation. He looked to U.S. law for guidance and began his analysis by adopting the balancing test articulated in Re Baker and Getty Fin. Services Inc., U.S. Bankruptcy Court, N.D. Ohio (1987) 78 B.R. 139:

The propriety of ordering substantive consolidation is determined by a balancing of interests. The relevant enquiry asks whether "the creditors will suffer greater prejudice in the absence of consolidation than the debtors (and any objecting creditors) will suffer from its imposition".

[221] Trainor J. then went on to list seven factors which had been developed to assist in the balancing of interests. Those factors were:

difficulty in segregating assets;
presence of consolidated financial statements;
profitability of consolidation at a single location;
commingling of assets and business functions;
unity of interests in ownership;
existence of intercorporate loan guarantees; and
transfer of assets without observance of corporate formalities.

[222] In considering these factors, it is clear beyond peradventure that Nortel has had significant difficulty in determining the ownership of its principle assets, namely the $7.3 billion representing the proceeds of the sales of the lines of business and the residual patent portfolio. This amount constitutes over 80% of the total assets of all of the Nortel entities[^21]. This issue has taken several years of litigation and untoward costs in the parties attempting to establish an entitlement to it. As the MRDA does not govern how the sales proceeds are to be allocated, there is no one right way to separate them. It cannot be said that there is no question which entity is entitled to the sale proceeds or in what amount. It is clear that these assets are in the language of Dr. Janis Serra “so intertwined that it is difficult to separate them for purposes of dealing with different entities”.

[223] Moreover, the evidence in this case is clear and uncontested that Nortel (a) had fully integrated and interdependent operations; (b) had intercompany guarantees for its primary indebtedness; (c) operated a consolidated treasury system in which generated cash was used throughout the Nortel Group as required; (d) disseminated consolidated financial information throughout its entire history, save for the year before its bankruptcy; and (e) created IP through integrated R&D activates that were global in scope.

[224] When consolidation occurs, some creditors may be prejudiced if they would have had a greater recovery of so many cents on the dollar against their debtor if there had been no consolidation. Conversely, other creditors may be benefitted by consolidation if they would have had a lesser recovery against their debtor if there had been no consolidation. In this case, even if a pro rata allocation amounted to a consolidation, the issue would be moot because it cannot be said that without consolidation one class of creditors, including the bondholders, would necessarily have had a greater recovery than with consolidation. The reason for this is that there has been no recognized measurable right in any one of the selling Debtor Estates to all or a fixed portion of the proceeds of sale.

[225] The bondholders who hold bonds with covenants of both NNL and NNI contend that they would be unduly prejudiced by a pro rata allocation of the lockbox funds as they are entitled to look to both NNL and NNI for payment of their claims and if one of these companies did not have sufficient funds to pay the bonds in full, they could look to the other. I agree that they are entitled to claim against both companies and this will be recognized in the pro rata allocation that will be ordered.

[226] The bondholders have the legal right to be paid in full on their bonds. But so do all of the other creditors. Like the pensioners and other creditors, the bondholders are not secured. Because of a shortfall in funds, all of these creditors cannot be paid in full. The issue is how the pain is to be shared.

[227] The total cash on hand in the U.S. Debtors’ and Canadian Debtors’ Estates as of June 2014 was a little over 25% of the face amount of the outstanding bonds. Without an allocation from the lockbox funds of a sufficient amount to enable NNL and NNI to pay the bonds in full, the bondholders could not be paid in full. The bondholders, however, have no covenants in their bonds requiring the lockbox funds to be allocated in any manner, and specifically, no right to have lockbox funds allocated to NNL or NNI. Nor do NNL or NNI have any such rights. The lockbox funds are not the property of any one of NNL or NNI or any other RPE.

[228] The bondholders are like other creditors in this regard. The other creditors of the Canadian Debtors could likewise argue that they will be prejudiced if the argument of the Monitor that all of the IP proceeds should be paid to NNL as the owner of the IP is not accepted. But the prejudice to be considered is not this kind of prejudice, but prejudice to legal rights. Neither the bondholders nor the other creditors of the Canadian Debtors have any legal right to have the lockbox funds allocated in a way that will benefit them.

[229] The bondholders with covenants of NNL and NNI contend that their expectations will be disregarded by a pro rata allocation and that it will harm the bond markets if they are not somehow paid in full. I think this argument is overblown in this case and in any event not supported by any evidence of their expectations.

[230] The evidence of Peter Currie, the CFO of NNC and NNL from 2005-2007, which is not contested, was that until the early to mid-2000s, Nortel's public debt was issued by NNL without guarantee from any other Nortel entity. In 2006, while Nortel's credit rating was still adversely affected by various factors, NNL issued notes having an aggregate principal amount of US$2 billion, which notes were conditionally guaranteed by NNI. NNI was a conditional guarantor in large part because at that time it carried certain hard assets on its balance sheet and because Nortel could obtain slightly better debt terms given that NNI was domiciled in the same place as the ultimate lenders, that is, the United States.

[231] Thus it is quite clear from the evidence that when Nortel went to the bond market in 2006 and 2007 to raise funds, Nortel believed that it required the covenant of NNI in order to get the financing on terms and at a cost that Nortel wanted. However, prior to the Nortel insolvency in January, 2009, the market place did not differentiate in any material way the bonds that were guaranteed by NNI and the bonds not carrying a NNI guarantee.

[232] From June, 2006 to December, 2008, Moody’s and DBRS issued nine credit ratings for Nortel that did not distinguish between Nortel bonds guaranteed by NNI and those that were not. The UCC’s expert witness Robert Kilimnik[^22] agreed on his deposition that if a guarantee is a risk differentiator from DBRS's point of view, and there were a series of bonds with a guarantee and a series of bonds without a guarantee, he would expect them to be rated differently. This is an indication that the market did not differentiate between the NNC bonds guaranteed by NNI from those that were not guaranteed.

[233] Another indication is the evidence of the Nortel bond spreads compared to U.S. government bonds contained in Ex. 58. The chart demonstrates that that Nortel bonds that carried an NNI guarantee traded at higher or equal spreads to Nortel bonds that did not carry an NNI guarantee. Mr. Kilimnik, an experienced bond trader, said on his deposition testimony was that bonds with a lower spread are considered less risky in the marketplace and that if guarantees were recognized by creditors as reducing the risk of issuances by the same company, he would have expected to see that expectation reflected in spread comparisons.

[234] Mr. Paviter Binning, the Executive Vice-President and CFO of NNC from 2007 to March 2010 and an impressive witness to be sure, agreed with that conclusion of Mr. Kilimnik and testified that the data implied that the market was giving no value to the guarantees. He also testified that in his experience, investors generally looked to the overall quality of the company and that the guarantees were neither here nor there. He agreed that part of the reason why the guarantees may have had no meaning for the market was that the bonds were sub-investment grade in the first place. His evidence, which I accept, means that after the bonds were issued, the guarantees by NNI did not have a material effect on the marketplace.

[235] John McConnell, a professor of business (finance) at Purdue University, delivered a report and testified on behalf the unsecured creditor’s committee of NNI in response to a report of Leif M. Clark and Jay L. Westbrook, the latter of whom did not testify at the trial. Professor McConnell’s report contained data from the date that the Nortel Group filed for protection on January 14, 2009 to January 2014 which indicated that the bonds not guaranteed by NNI traded at prices below the bonds guaranteed by NNI.

[236] I do not see this data as relevant. Counsel for the bondholders in his opening asserted that the expectations of bondholders that are relevant are the expectations pre-petition and not post-petition.

[237] If the expectations of those who purchased bonds post-petition were relevant, there was no evidence at all from such purchasers. Professor McConnell spoke to no bondholder and on cross-examination admitted that he had no way of knowing what factors went into the purchase and/or sale of any of the Nortel bonds by any of the current bondholders in the market post-filing. No bondholder testified or gave any evidence of expectations in acquiring bonds.

[238] The evidence of Professor McConnell is based entirely on the fact that after the insolvency filings, bonds without a NNI guarantee traded at a lower price than those with a NNI guarantee. There are two points that can be made. The first is that his conclusion is an inference drawn from the trading price of the bonds after the insolvency as to what motivated those purchasers of the bonds after the insolvency. Second, there was no analysis of Professor McConnell that would lead to the conclusion that his inference of bondholder purchaser expectations could apply to purchasers of bonds prior to the onset of insolvency. He said he could not do such an analysis because before insolvency the bonds had different attributes which would not permit him to draw inferences as to the effect of guarantees. Be that as it may, I would not accept the inference drawn by Professor McConnell regarding the effect of the guarantees on a purchaser of bonds. I prefer the evidence of Mr. Binning to which I have referred.[^23]

[239] Moreover, the evidence is clear that bonds trade on a much different basis after insolvency. Mr. Binning testified that prior to the threat of insolvency, the bonds traded on a yield to maturity basis, meaning that bondholders take all of the payments that would be expected to be made if the bond is held to maturity, and then calculate a percentage yield based upon the price paid for the bonds. Once insolvency or financial distress is anticipated, Mr. Binning testified that bonds trade in the hands of distressed investors who trade not on a yield to maturity basis but in a classic arbitrage market based upon price and expectations of future price and what they think they can make on the bonds during insolvency. He advised the board of Nortel on September 30, 2008, three and a half months before the Nortel filing, that RBC had advised that approximately 50% of the bonds had traded into the hands of distressed investors.

[240] Professor McConnell also testified that as new information came into the marketplace about the likely recoveries, that would be reflected in the price of the bonds. That is another way of saying that distressed investors have bet on the future outcome of this case. This is reflected in the volumes and trading prices of the bonds at various times between January 14, 2009 and June, 2014, including (i) in the immediate aftermath of the Filing Date when the bonds were trading at very low prices, (ii) during the prolonged three-day auction resulting in the residual IP sale to Rockstar at the beginning of July 2011 as purchasers placed bets on bond price increases and recoveries following the completion of that sale; and (iii) in reaction to Delaware Bankruptcy Court Judge Walrath’s September 2011 decision in In re Washington Mutual holding that post-petition interest must be awarded at the federal judgment rate and not at the rates in the various bonds.

[241] The bondholders group that at the time of the trial held a majority of the unsecured guaranteed bonds purchased the vast majority of their holdings after the filing date of January 14, 2009 and at a significant discount to par. Certain members purchased when the bonds were trading at as low as 30 cents on the dollar and others received smaller, but still substantial, discounts. This can be seen in exhibits 59 and 60. The vast majority of their collective holdings were acquired in the period between July 31, 2009, at or around the time when Nortel began to liquidate its assets, and July 18, 2011, at or around the time of the Residual IP Sale.

[242] The creditor expectations of the current bondholders, who acquired their bonds post-petition, even if known or supported by evidence, is not something I would take into account in this case. I infer from the evidence that any such expectations would have been based on their views as to litigation outcomes and should not be the basis of any decision by the courts.

[243] In considering potential prejudice to the bondholders in the event of a pro rata allocation, consideration must be given to what the bondholders would gain. The bonds provide access to the assets of the issuer and guarantor. They do not provide any right to assets of any other entity in the Nortel Group. The 2006, 2007 and 2008 offering memoranda for the guaranteed bonds set out risks associated with the bonds, including the following notice regarding the lack of access to other Nortel entities:

The Issuers' subsidiaries are separate and distinct legal entities and any subsidiary that is not a Guarantor will have no obligation, contingent or otherwise, to pay amounts due under the Notes or the Guarantees or to make any funds available to pay those amounts, whether by dividend, distribution, loan or other payment.

[244] The offering memoranda also contained the following risk that investors would face in the event of insolvency of Nortel entities and the lack of access to the assets of those entities:

In the event of a bankruptcy, liquidation or reorganization of any direct or indirect non-guarantor subsidiary of NNC, all of the creditors of that subsidiary (including trade creditors and creditors holding secured or unsecured indebtedness or guarantees issued by that subsidiary) and third parties having the benefit of liens (including statutory liens) against that subsidiary's assets will generally be entitled to payment of their claims from the assets of such non-guarantor subsidiary before any of those assets are made available for distribution to any Issuer that is a shareholder of such subsidiary. As a result, the Notes and the Guarantees are effectively junior to the obligations of non-guarantor subsidiaries.

[245] Whereas the investors who acquired their bonds pursuant to the offering memoranda were specifically made aware that in the situation in which Nortel now finds itself, they would not have access to assets of other Nortel entities that had not guaranteed the bonds, the effect of a pro rata allocation is to provide the current bondholders with such access. The lockbox funds represent the proceeds of sale of all of the assets of all of the 38 entities under the IFSA. Creditors holding guarantees have access under a pro rata allocation to not only the assets of the principal obligor and guarantor corporations, but the proceeds of sale of all the assets of the selling debtors. This is access to more pools of asserts than that for which the holder of a guarantee bargained.

[246] While the current bondholders may have thought, or bet on an outcome, that NNL or NNI would likely achieve a win in this litigation that would provide those two companies with sufficient assets to pay the bonds in full and with post-filing interest, there was no guarantee at all that this would be achieved. The bonds contained no covenants that required the assets of NNL or NNI to be maintained at a certain level and no covenants that required the lockbox funds to be allocated in any manner. Mr. Binning had advised the Nortel board in September, 2008 that there were no maintenance covenants in the bonds, meaning that Nortel did not have to live up to any debt servicing ration . He testified that what the guarantees under these bonds essentially gave the bondholders was access to the assets in Canada and in the US without a great degree of comfort as to what those assets would be from time to time. I accept that evidence.

[247] As to the effect of a pro rata allocation on the ability of issuers to issue bonds in the future, Professor McConnell on his cross-examination said that he had no opinion on that subject and that he had not tried to quantify what effect a pro rata allocation would have on the capital markets. Thus there is no evidence that a pro rata allocation in this case will detrimentally affect the capital markets and the ability of issuers to issue bonds in the future. Professor McConnell’s statement that he had no opinion on the subject is perhaps not too surprising taken the highly unusual facts surrounding the Nortel insolvency and the difficulty of determining ownership of the IP that was sold.

[248] It is said that the $2 billion claim of NNI against NNL that was approved by both courts is an impediment to a pro rata allocation. I do not think that is the case. The $2 billion claim will be treated as one of the unsecured liabilities of NNL.

[249] The same principles that apply to the US$ 2 billion claim by NNI against NNL will apply to the admitted claim of NNUK and Nortel Networks SpA against NNL pursuant to the Agreement Settling EMEA Canadian Claims and Related Claims dated July 9, 2014, and to the claim of the UKPC for £339.75 million recognized in my judgment of December 9, 2014.

Appropriate pro rata allocation method

[250] The allocation each Debtor Estate will be entitled to receive from the lockbox funds is the percentage that all accepted claims against that Estate bear to the total claims against all Debtor Estates.

[251] In determining what the claims against a Debtor Estates are, a claim that can be made against more than one Debtor Estate can only be calculated and recognized once. The one that is known is the bondholder claim for $4 billion, referred to as the claim on cross-over bonds. All but one of such bond issues was issued by NNC or NNL and guaranteed by NNI. One bond issue for $150 million was issued by NNCC, a subsidiary of NNI, and guaranteed by NNL[24]. The claims on the bonds in determining the claims are to be made on the Debtor Estate of the issuer. If a claim on a guaranteed bond is not paid in full by the issuer Debtor Estate, a claim for the shortfall can be recognized by the Debtor Estate that guaranteed the bond, but that shortfall claim will not be taken into account in determining the claims against the Debtor Estates.

[252] One of the known claims is the claim of the UKPC for the approximately £2.2 billion deficit in the NNUK pension plan. If the UKPC makes a claim for this amount against NNUK and also against other EMEA Debtors, those claims against the other EMEA Debtors will not be taken into account in determining the claims against the Debtor Estates. The claim may be taken into account only once in the pro rata allocation.

[253] I understand that for the Canadian Debtors and the U.S. Debtors, the claims for the most part are generally known although there are some claims still unresolved, such as the SNMPRI claim. The U.K. Administrator has not yet instituted a claims procedure, apparently awaiting a determination of this allocation proceeding. In my view, the process should be undertaken now and I expect this will happen.

[254] Interim distributions have been proposed. In my view, this would be especially important for the predominantly elderly pensioner population and disabled employees who have endured hardship as a result of the loss of their benefits but also for other creditors who have waited more than five years for a distribution on their claims. An interim distribution should be made if possible.

[255] Briefs should now be filed by those parties supporting an interim distribution with full details of what is requested. Opposing briefs would of course be required. The procedures and timing could be discussed at a 9:30 am appointment.

Allocation on a basis other than pro rata

[256] The evidence on this subject was complex and varied dramatically from party to party. To wit:

(a) The Monitor on behalf of the Canadian Debtors contended for an allocation of $6.034 billion to the Canadian Debtors, $1.001 billion to the U.S. Debtors and $300.7 million to the EMEA Debtors.[25]

(b) The U.S. Debtors contended for an allocation of $0.77 billion to the Canadian Debtors, $5.3 billion to the U.S. Debtors and $1.23 billion to the EMEA Debtors.

(c) The EMEA Debtors contended for an allocation of $2.32 billion to the Canadian Debtors, $3.636 billion to the U.S. Debtors and $1.325 billion to the EMEA Debtors.

[257] I have given consideration to the valuation issues. To a great extent, they are dependent on the various interpretations of the MRDA asserted by the parties. For that reason I would not use any of the valuations for the purpose of the pro rata allocation as I have found that the MRDA does not govern the allocation. However, my views and findings on the valuations are set out in Appendix A for the business line sales and Appendix B for the residual IP sale to Rockstar.

Conclusion

[258] A judgment is to go that the lockbox funds are to be allocated on a pro rata allocation basis with the following principles to govern:

(1) Each Debtor Estate is to be allocated that percentage of the lockbox funds that the total allowed claims against that Estate bear to the total allowed claims against all Debtor Estates.

(2) In determining what the claims are against the Debtor Estates, a claim that can be made against more than one Debtor Estate can only be calculated and recognized once in accordance with these reasons for judgment. Claims on bonds are to be made on the Debtor Estate of the issuer. A claim for any shortfall can be recognized by the Debtor Estate that guaranteed the bond, but that shortfall claim will not be taken into account in determining the claims against the Debtor Estates. If the UKPC makes a claim against more than one Debtor Estate, such additional claims will not be taken into account in determining the claims against the Debtor Estates.

(3) Intercompany claims against a Debtor Estate are to be included in the determination of the claims against that Estate.

(4) Cash on hand in any Debtor Estate will not be taken into account in the pro rata allocation. Each Debtor Estate with cash on hand will continue to hold that cash and deal with it in accordance with its administration.

(5) An interim distribution may be allowed upon further submissions. Briefs in favour of and opposed to an interim distribution are to be filed on a time-line to be considered at a 9:30 am appointment.

(6) Proposed schedules for expediting any remaining claims procedures are to be provided without delay.

Epilogue

[259] I cannot leave these reasons without commenting on the persons who made this unique case possible.

[260] First, to the technical staff who provided the facilities to permit this trial to be conducted in two different countries at the same time, I say it was a job more than well done. It was outstanding and we are indebted to you all. Judge Gross and I have no idea how it was all set up and operated, but I know he is as grateful for the facilities as I am. Thank you.

[261] Second, to the reporters and their staff, it was also a job more than well done. Apart from the instantaneous real time reporting that permitted all parties to see the evidence as it was being given, we were blessed with draft transcripts being electronically sent to us shortly after the evidence concluded each day and final transcripts later that evening.

[262] Third, to the lawyers. We were blessed with outstanding counsel on both sides of the border. In a case such as this with the amount at stake, one can understand the pressures on counsel and how those pressures could get in the way of a smooth preparation and presentation of the case. From what I could see, all acted in a professional manner that does them credit. Without that, the case could not have proceeded as well as it did. Their staff should also be congratulated for the smooth way in which the case was electronically presented. It was a marvel.

[263] Finally, I want to thank Judge Gross for his courtesies and good humour. It has been a pleasure to work with him. Without such a good relationship and the trust that we developed for each other, this trial and its conclusion would not have been possible.

“FJC Newbould J.”

Newbould J.

Date: May 12, 2015

APPENDIX A

Allocation of the proceeds of the line of business sales

[1] The experts for the various parties differ on the way that the proceeds of the sales of the LOB should be allocated amongst the Canadian estate, the U.S. estate and the EMEA estate.

[2] Mr. Kinrich, the valuer called by the U.S. Debtors, did not value the various assets sold and attempt to allocate them by any particular method. Rather he allocated the entire sale proceeds by taking the revenues of each company whose businesses were sold and allocating to each company (and the group they were in) the percentage of its revenues to the total revenues of all companies whose businesses were sold. His resulting allocation was 11.9% or $340 million to the Canadian Debtors, 18% or $510 million to the EMEA debtors and 70% or $1.99 billion to the U.S. Debtors.

[3] Mr. Malackowski called by the EMEA debtors valued the IP rights sold by using a revenue or license valuation method. His valuation of the IP sold in the business sales was $765.2 million.

[4] Mr. Huffard called by the EMEA debtors then allocated the various kinds of assets sold. He valued the tangible assets that were sold at $118 million and allocated this amount to the companies that sold them. He allocated the IP that was sold and valued by Mr. Malackowski at $765.2 million by a contribution approach which allocated the IP according to the amount of R&D expenditures of each of the RPEs. He attributed the balance of the LOB sale proceeds as “customer related assets and goodwill” and allocated them on the basis of the percentage of revenues generated by each entity in 2008. Mr. Huffard did not give a separate total figure in his report for the allocation of the LOB sale proceeds.

[5] Mr. Green called by the Canadian Debtors dealt with each kind of the various assets sold. He allocated the tangible assets sold by giving to the companies that sold them their book value, which he calculated to be $534.19 million. He valued the workforces sold by their cost that the selling companies would incur to replace them at $255.33 million and allocated those costs to the companies. He allocated the IP and customer relations by valuing what the licensed participants gave up to enable the sales on the basis that their license rights were limited to the “Products” “by or for the Participants” as defined in the MRDA, and allocated the balance of the sale proceeds to NNL as the “owner” of the IP. His resulting allocation was 54.8% or $1.58 billion to the Canadian Debtors, 10.4% or $300.97 million to the EMEA debtors and 34.7% or $1001.5 billion to the U.S. Debtors.

(i) Mr. Kinrich

[6] Mr. Kinrich’s view is that the RPEs that were licensed participants had all license and sublicense rights as owners. Assuming that to be the case for this analysis, I have some concerns with his analysis. Mr. Kinrich allocated all of the assets sold in the business sales on a revenue basis. He stated in his report that the value of the sold assets is reflected in the revenue generated by each entity that sold the assets. That is, value he said was reflected in revenue figures.

[7] Mr. Kinrich himself in his report said that financial economists agree that a discounted cash flow analysis is the preferred technique for asset valuation and that one of the requirements is to have projected future cash flows less costs. In his report he did not say why he had not done such an analysis when dealing with the business sales. At trial he relied on texts to support his use of a revenue approach in firms with losses, one of which is Valuing Small Businesses and Professional Practices, which suggests gross revenue multiples may be used in restricted situations, being to approximate a range of possible values with a minimum effort, conclude an estimate of value when other data are unavailable or inadequate or as one indicator of value used in conjunction with more rigorous valuation methods. The text also said that for companies with losses or erratic earnings, multiples of price to revenue for other comparative companies may give some indication of how others asses the future of the industry or profession. But that is not what Mr. Kinrich did. He did not look at revenue multiples from the sale of any comparable companies. I viewed his attempt to bolster his revenue approach by resort at trial to texts to be an attempt at ex post facto rationalization. It would have been a little more persuasive if these rationales has been provided in his report, particularly as in his report he did a sensitivity check based on gross margin (revenue less cost of goods sold) and contribution margin (revenue less selling, general and administrative expenses).

[8] Mr. Kinrich said at trial that he did not have available forecasts that would divide income streams by territory but that is beside the point so far as a gross revenue valuation is concerned. The issue is whether it would have been preferable to take costs into account in his revenue approach to allocation.

[9] Because the U.S. market had the highest revenues, it follows that using a revenue approach as Mr. Kinrich did will result in the high allocation of the business sale proceeds to the U.S. (70%) and a low allocation to Canada (11.9%). However, because revenue does not consider costs, this result ignores the way in which Nortel operated as a matrix structure and the reliance by all operating areas, including the U.S., on IP generated by R&D elsewhere. The 2009 revenue that Mr. Kinrich used in his analysis to compare revenues, the basis of which was assumed license rights under the MRDA, was subject to the obligation in the MRDA to make payments pursuant to the RPSM measured by R&D expenditures of each RPE. Mr. Kinrich did not take into account.

[10] Mr. Kinrich acknowledged on his cross-examination that while he assumed that the set of license rights as he saw them would continue in the future to exist, he gave no effect to sharing obligations that might arise under the MRDA, his reason being that he understood that the sharing provisions did not apply to sales proceeds. That in my view was no answer. What he undertook was to determine the relative value surrendered by each of the selling entities, including the RPEs. To determine the value of rights of each of the RPEs without taking into account the RPSM sharing obligations failed to properly determine relative values among the RPEs. I accept the opinion of others, including Dr. Bazelon and Mr. Green, on the point. The various businesses in Nortel historically operated on varied operating margins.

[11] Mr. Green pointed out, with reference to texts including those that Mr. Kinrich referred to at trial, that a disadvantage of focusing on revenues is that it can lull one into assigning high values to firms generating high revenue growth while losing money and that the method assumes that the businesses are equally profitable. His view was that a revenue based allocation was inappropriate in a matrix structure such as Nortel with interrelated operating businesses in which certain entities bore disproportionate shares of expenses like R&D which would be ignored.

[12] EMEA and the UKPC contend that Mr. Kinrich should not have used 2009 revenue figures as 2009 was an anomalous year for Nortel. Nortel filed for insolvency protection in January 2009 and from then on was operating under the supervision of courts in Canada, the US, the UK, and elsewhere. Throughout 2009, Nortel was actively engaged in selling its businesses, signing seven out of eight of its sale agreements in that year. All of this affected its ability to generate revenues. Four of the business sales were concluded before the year’s end. Because of these dispositions, complete financials for 2009 were not even available for certain businesses and revenues had to be estimated based upon performance prior to the sale closing date.

[13] Dr. Bazelon’s evidence was that NNI’s share of global revenue plateaued by 2008 at about 65% but in 2009 it increased by about 4%. In my view, EMEA and the UKPC have a valid point. 2009 was not a typical year for Nortel. While NNI contends that 2009 resembles the weighted average over the years 2001 to 2008, but that ignores the steadily declining trend from NNI having 74% in 2001 to about 65% in 2008. Using 2009 for his revenue analysis was overly aggressive. The effect was to shift about 4%, or $100 million, from EMEA to the U.S. Debtors.

(ii) Mr. Malackowski and Mr. Huffard

[14] The allocation of the proceeds of the LOB sales on behalf of the EMEA Debtors is based on the evidence of Mr. Malackowski who valued the IP sold at $765.2 million and on the evidence of Mr. Huffard who allocated all of the sales proceeds using different methods for each type of asset sold. There are problems with the EMEA allocation.

[15] Mr. Malackowski used a discounted cash flow analysis to value the IP. He said there was a defensive component and a synergistic component to the IP. To measure the defensive component, he took the revenue forecasts of Nortel in the “deal books”, market derived growth rates, and royalty rates from an IPCo model. For a discount rate he used an average of weighted average cost of capital rates of the industry in which the Nortel business operated. To measure the synergistic component, he used revenues of a hypothetical market participant for each line of business sold, market derived growth rates, and royalty rates derived from what he said was the implied rate paid by Ericcson as a member of the Rockstar consortium. He added 15% to the discount rate used in his defensive component.

[16] Mr. Malackowski’s valuation of the IP sold at $765.2 million, if accepted, means that the IP represented roughly 25 % of the total sales proceeds of $3.1 billion. Yet, the evidence is overwhelming that IP created by Nortel’s R&D was the driver of the profitability of the business. Even Mr. Huffard view was that within Nortel, IP was considered the driver of revenue in each of the businesses and purchasers of the businesses would have considered the acquisition of IP as a critical aspect. Mr. Britven, an expert called by the CCC, arrived at figures based on the purchase price allocations made by the purchasers that stated what the purchasers considered the fair value of the various acquired assets to be. Those figures put the percentage of the IP of the total business sale proceeds at 40%.

[17] In his rebuttal report, Mr. Malackowski in an attempt to show the size of what he considered to be a windfall if the position of Mr. Green were accepted, said that all of the Nortel the IP in total in the hands of Nortel could be worth $10.4 billion, of which he allocated $3.761 to the business sales and $6.6 billion to the residual sale to Rockstar. His reason for this extra value in his report was that some of the residual IP sold to Rockstar was encumbered by the non-exclusive licenses given to the purchasers of the lines of business. Rockstar paid $4.5 billion. If Mr. Malackowski’s figures are right, it means that the non-exclusive licenses given to the purchasers of the lines of business reduced the value of the residual IP sold to Rockstar from $6.6 billion to $4.5 billion, or by $2.1 billion. That reduction in value to Rockstar attributed to the non-exclusive licenses granted in the business sales means that those non-exclusive licenses were worth $2.1 billion, and it does not make sense that Mr. Malackowski valued both the outright licenses and the non-exclusive licenses given to the purchasers of the lines of business at only $765.2 billion.

[18] The Monitor points out that the royalty rates used by Mr. Malackowski in establishing revenues to be valued were taken from the IPCo litigation light model and that within that litigation light model he chose the lowest of three rates. He did not use any Nortel intercompany-stated royalty rates. The Monitor suggests that is an explanation why the IP valuation of Mr. Malackowski is too low. For certain the royalty rates charged directly affect the revenues and thus the value obtained by a DCF method of valuation. Whether it is the only reason for the low valuation is another matter. There are many inputs in a valuation.

[19] Mr. Huffard was the expert called by EMEA to opine on the allocation to be made of each component of the business sales. The Monitor is critical of his qualifications. Mr. Huffard is an investment banker with considerable experience advising distressed companies who has “led valuation analyses” for companies and their assets. He holds a Master of Management degree from Kellogg Graduate School of Management at Northwestern University. Mr. Huffard is not accredited as a valuator and said that in the field of investment banking that is typical.

[20] I must say that Mr. Huffard was not forthcoming in his evidence about his experience. When asked if he had done any valuations or the allocating of assets in connection with intellectual property companies, he said several times that he had trouble understanding what an intellectual property company was and asked if Nortel was an intellectual property company. Yet when asked on his deposition whether he had done any valuations or the allocating of assets in connection with intellectual property companies, he answered “Not in connection with intellectual property companies.” I think it fair to consider this answer in dealing with his evidence.

[21] Mr. Huffard believed that there were three classes of assets to be valued and examined in the business sales. The first is net tangible assets. The second is the IP. The third is customer-related assets and goodwill not otherwise encompassed in goodwill. Mr. Huffard did not do any valuation exercise for his third class. Rather he just took the balance of the purchase price and allocated it. The values attributed to the first two classes therefore directly affected the value of his third class.

[22] For the tangible assets, Mr. Huffard took the book values of the assets, which consisted of accounts receivable and prepaid expenses, inventory and fixed assets. This book value in his report totalled $403 million. At trial, in his demonstrative exhibit, his total was $361 million. Why the difference was not explained. From the book values, Mr. Huffard deducted liabilities assumed by the purchasers of the lines of business, the largest of which was deferred revenue. He viewed the assumed liabilities as a fourth asset class that resulted in an increase in the purchase price from the buyer’s perspective and a reduction from the seller’s perspective. They had to be deducted from the assets and he did this the tangible asset class. This resulted in net tangible assets in his report of $124 million, being $39 million for Canada, a negative $27 million for EMEA, $106 million for the U.S. and $6 million for Asia and the Caribbean. At trial, his demonstrative showed the net tangible assets at $118 million with the same net figures for Canada, EMEA and the U.S. as in his report. How this occurred on the different book values in his report and in his demonstrative was not explained.

[23] For the allocation of the IP, Mr. Huffard took Mr. Malackowski’s figure of $765.2 billion. He allocated them amongst the RPEs using Mr. Malackowski’s contribution approach using historical R&D spending from 1992 to 2008. His view was that the portions of the sale proceeds attributable to IP were, in effect, a capitalization of future revenues that would otherwise have been shared among the RPEs in accordance with the RPS methodology. This resulted in 40.8% or $312.2 million being allocated to Canada, 42.6% or $326 million being allocated to the U.S. and 16.5% or $126.2 million being allocated to EMEA. In his demonstrative at trial, these percentages were rounded, with 41% to Canada, 43% to the U.S. and 16% to EMEA.

[24] Mr. Huffard then included the balance of the sale proceeds of $2.198 billion into his third class of customer-related assets and goodwill. He did no analyses of the value of either the customer-related assets or of the goodwill and allocated them based on the revenue generated by each entity in fiscal year 2008. He said he did not use net revenues to allocate among the entities because in his view cash flows are influenced by transfer pricing and inter-company arrangements for tax purposes. Based on the revenues alone, he allocated 9.2% or $202 million to Canada, 62.6% or $1.375 million to the U.S., 18.6% or $155 million to EMEA, 2.6% or $57 million to the Caribbean and 7.1% or $155 million to Asia.

[25] While Mr. Huffard did not provide a total for the business sales allocation, by adding up the different classes, his total allocation to Canada was $553.2 million to Canada, $1.807 billion to the U.S. and $254.2 million to EMEA. This is somewhat less than the $2.85 billion available from the business sales proceeds.

[26] As Mr. Huffard did not undertake any valuation of his third residual category, his conclusion of the amount to be included in it is dependent upon the amount of his tangible asset valuation and Mr. Malackowski’s IP valuation. If Mr. Mr. Malackowski’s IP valuation is too low, then the amount in this residual class allocated by Mr. Huffard will be too high.

[27] There are problems with allocating this residual class entirely by revenues of each company or groupings of companies. Mr. Huffard described it as the value of customer-related assets and goodwill not otherwise associated with IP. He acknowledged that these assets, like any other assets, have their value fundamentally related to their ability to generate profits, and that while Nortel operated the businesses, it was not revenue that allocated those values but the RPS method of sharing profits after revenues and costs were calculated. This is the same criticism made of Mr. Kinrich in using a revenue tool to allocate the sales proceeds rather than a profitability tool.

[28] Mr. Huffard acknowledged that in circumstance where, because of decisions made and cost-effectiveness and historic reasons, sales and customer support was done in a country which had low domestic revenues, his revenue allocation method for the customer-related assets and goodwill category was not going to compensate that country because it had low revenues. This circumstance was commonplace in Nortel with its matrix structure, with customer support carried out in one country for sales in another. He acknowledged if a large percentage of the workforce is in a place like Canada, which does R&D and which does sales support and supports the global organization but doesn’t have a large native revenue stream, he was allocating the value of that workforce to the other entities where there is a revenue stream and not to Canada.

(iii) Mr. Green

[29] Mr. Green valued different asset classes differently. He first valued tangible assets by taking their book value and allocating them to the companies which owned them. This was the same method used by Mr. Huffard. However, different from Mr. Huffard, he did not deduct any deferred liabilities from the tangible asset amount. His evidence was that deferred liabilities are essentially amounts that would be due that are related to projects that have already been billed and perhaps paid for and so they didn’t offset the physical assets, the tangible assets such as the accounts receivable, the other things that were sold. He pointed out that by his not deducting deferred liabilities, it was to the relative benefit to the U.S. Debtors because they had the highest deferred revenues and, accordingly, deducting the liabilities would most significantly decrease the U.S. Debtors’ share of the value of net tangible assets. He also pointed out in his rebuttal report that not all liabilities recorded on the books of Nortel were assumed by the purchasers and for those that were it was not possible to determine on which entity’s books those liabilities were recorded. I accept this position of Mr. Green in not deducting assumed liabilities in valuing and allocating the tangible assets on the basis of book values recorded on each entity’s books.

[30] Mr. Green’s value for the tangible assets was $534.19 million, and he allocated $121.74 million to the Canadian Debtors, $317.59 million to the U.S. Debtors and $94.86 million to the EMEA Debtors.

[31] Mr. Green next valued the workforce in the lines of businesses that were transferred to purchasers. His opinion was that generally speaking, the cost approach is applied to the valuation of an assembled or in-place workforce. He valued the workforce by calculating the cost to replace the work force. He concluded that the total value of the work force was $255.33 million and he allocated $78.68 million to the Canadian Debtors, $134.74 million to the U.S. Debtors and $41.91 million to the EMEA Debtors.

[32] In the sale of the Enterprise business, several corporate entities owned by NNI were sold. Mr. Green valued these assets based on contemporaneous fair market valuations done at the time these businesses were sold. There is no evidence that these assets had a value other than as set out by Mr. Green. Mr. Green made an allocation to NNI of proceeds attributable to the sale of its subsidiaries in the amount of $110,970,000. No other valuer dealt with this asset.

[33] Mr. Green was of the view that once the tangible assets and the workforce were valued, the balance of the business sale proceeds was attributable to IP, the primary driver of Nortel’s value, and customer relationships. He valued and allocated the IP and customer relationships sold in the business sales by valuing the license rights of NNI and the EMEA RPEs surrendered by them to permit the sales to take place on the basis that their licenses were restricted to Products by or for the Participants as defined in the MRDA and as contended by the Monitor. The balance he attributed to NNL as the owner of the IP.

[34] Mr. Green performed a DCF valuation. He projected revenues and expenses for each business sold and for this to project the future revenues of the Nortel businesses he used forecasts prepared by Nortel that were referred to as “Retained by Nortel” forecasts. They projected the revenues that would have been earned and the expenses that would have been incurred, if the operating businesses had been retained by Nortel. After calculating the operating profits of each business sold, Mr. Green aggregated those profits and applied the RPSM on the assumption that the MRDA would have remained in place, using the capital stock percentage for the first quarter of 2010, which covered a rolling average from 2005 to 2009. He applied a discount rate of 12% for the operating profits derived from existing technology and 30% for operating profits to be derived from yet to be invented technology and thus more risky. He concluded that the value of the license rights surrendered by NNI was $438.2 million and by the EMEA RPEs was $164.2 million. The balance of his residual amount, being $1.379 billion was allocated to NNL.

[35] Mr. Green’s resulting allocation was 54.8% or $1.58 billion to the Canadian Debtors, 10.4% or $300.97 million to the EMEA debtors and 34.7% or $1001.5 billion to the U.S. Debtors.

[36] EMEA and the U.S. Debtors contend that a basic problem with Mr. Green's analysis is his conclusion or assumption that NNL was the owner of the IP and entitled to its residual value after deducting the license rights of EMEA and NNI which he limited to Nortel Products by or for the Participants. This is a basic legal issue.

[37] EMEA argues that customer relationships were very important to Nortel and that they should have been valued and allocated separately from IP and not included in Mr. Green's residual category. Mr. Green's explanation for not doing so was that customer intangibles represented historical relationships in which customer files and ongoing agreements exist, the value of which was represented in his revenue figures that he used and were thus subsumed in the IP license rights which he valued. He said that a separate valuation of customer relationships would be duplicative of the values of the license rights surrendered because it would be based on the same revenues and profits as used in the license rights valuation.

[38] Mr. Malackowski argued that the MRDA did not transfer customer relationships to NNL. This does not strike me as a valuation concept and one can argue, as the Monitor does, that NN Technology was owned by NNL and it included all intangibles.

[39] This is a valuation issue. There is no question that customer relationships were important to Nortel. However that is not the issue. The issue is how to value them. Mr. Berenblut was of the opinion that customer relations were co-mingled with IP rights because the value to use them depended on the ability to sell Nortel products and that their value would be included in the value of rights to sell Nortel products. Dr. Bazelon, an EMEA expert, agreed on cross-examination that goodwill and customer relationships are entangled with the IP and take their value from the IP, at least in part. Brian McFadden, the Chief Technology Officer at Nortel for some time, said that R&D was crucial in initiating relationships with and developing sales from customers for Nortel products. I accept Mr. Green's opinion that no separate valuation needed to be made for customer relationships.

[40] It is also argued that Mr. Green should have separately valued and allocated goodwill. Mr. Huffard included goodwill in his residual class, although he did not attempt to value it. Mr. Britven, called by the CCC, included a value for goodwill in his business sales analysis. He took what the purchasers had allocated in their PPAs as goodwill, and referred to it as Purchaser Goodwill.

[41] Mr. Green's response to this is that Nortel wrote off all of its acquired goodwill at the end of 2008. This indicated that, at the time, Nortel management did not believe it would be able to realize the value of the goodwill from these acquisitions in the future. As for its own “internal goodwill,” Nortel was suffering losses from its operations and was not generating positive cash flows. Thus, from an accounting and finance perspective, Nortel had no goodwill from its own operations. By classifying the residual value as goodwill, Mr. Huffard accounted for an asset that did not exist within Nortel and was not transferred to the buyers. By applying the buyer’s perspective, Mr. Huffard failed to answer the question of how to allocate the sales proceeds according to the value of the interests each of the Debtors transferred and rights each of them relinquished.

[42] There is actually support for Mr. Green's position in Mr. Britven’s report in which he included a value for goodwill taken from the purchasers’ PPAs. These purchaser allocations are done by purchasers for accounting purposes and usually are driven in part at least by tax considerations. Mr. Britven said that Nortel wrote off the value of substantially all of the goodwill that it had on its balance sheet. He said that Nortel did not have sufficient value to support any significant goodwill value and that the goodwill in the business sales related to the attributes of the buyer, not the attributes of Nortel. He said that any goodwill recognized by purchasers in their PPAs did not reflect amounts that could have been realized by the licensed participants through the continued operations of their lines of business.

[43] I agree with Mr. Green's approach to goodwill and accept his opinion that there was no goodwill value in the Nortel businesses that were sold.

[44] Regarding the DCF method used by Mr. Green to value the U.S. and EMEA license rights, Mr. Kinrich was critical of the revenue forecasts used by Mr. Green and stated that he had not followed the International Financial Reporting Standards which state that in measuring value in use, an entity shall base cash flow projections on reasonable and supportable assumptions that represent management’s best estimate of the range of economic conditions that will exist over the useful life of the asset.

[45] This IFRS material was not put to Mr. Green on his cross-examination, which it should have been for this argument to be made. However, I do not think the criticism is justified. Mr. Green used projections made by Nortel. He used projections referred to as a “Retained by Nortel” scenario which projected what revenues and expenses would be either retained by Nortel or spun-out on its own as a stand-alone company. He declined to use Nortel’s “Safe Hands” projections for several reasons that he explained, including the fact that they forecasted the businesses in the hands of a well-capitalized third party who could invest adequate capital in the business and who could earn greater profits than if they remained in Nortel’s hands. Mr. Green did no DCF analysis as he allocated the business sales solely on revenues.

[46] Mr. Kinrich was also critical of Mr. Green for not including a terminal value in his DCF valuation. Mr. Green's explanation for this on his deposition was that in his present value analysis, at year nine the present value factors were close to zero. So even if there were a terminal value, it would be virtually of no value in a present value computation. In his report, he said he thought that to include potential profits after nine years was too speculative. There is no competing DCF valuation to indicate what Mr. Green did was wrong.

[47] Mr. Green’s analysis in part is dependent on the interpretation of the MRDA advanced by the Monitor on behalf of the Canadian Debtors. One cannot quarrel with the logic of it if that interpretation were to govern the allocation.

APPENDIX B

Residual IP proceeds allocation

[1] The residual IP was sold to Rockstar for $4.5 billion. After payment of a break fee and expense reimbursement to Google, the remaining net proceeds held for allocation amount to $4.45 billion.

[2] There is differing expert opinion as to how to allocate the proceeds of the Rockstar sale amongst the Canadian debtors, the U.S. debtors and the EMEA debtors.

[3] Mr. Green allocated the proceeds on the basis of his interpretation of the MRDA under which it was NNL that owned all of the patent rights that were sold to Rockstar.

[4] Mr. Kinrich for the U.S. debtors allocated the proceeds on a revenue approach on the basis that each Participant owned all of the economic rights to the patent rights sold to Rockstar in their exclusive jurisdictions and that their revenue streams that they gave up should be valued. Mr. Green as an alternative analysis for the Canadian debtors and Mr. Malackowsi as an alternative analysis for the EMEA debtors prepared valuations correcting what they saw as errors by Mr. Kinrich.

[5] Mr. Malackowski allocated the proceeds on a contribution basis by calculating what he saw as the contributions by each of the Participants to R&D over the life of the patents that were sold to Rockstar.

(i) Mr. Kinrich’s license approach to value

[6] Mr. Kinrich assumed that each of NNL, NNI and the EMEA debtors owned all of the economic benefits of the residual IP. He allocated all of the Rockstar proceeds to NNL, NNI and EMEA by taking what he said would be the revenue earned in each of those three geographical areas and then doing what he said was a discounted cash flow analysis (”DCF”) on those revenue streams. I have held that the Licensed Participants did not own all of the economic benefits of the residual IP. However, on the assumption that they did, I will consider Mr. Kinrich’s analysis.

[7] Mr. Kinrich obtained his revenue streams by taking one of the IPCo revenue model assumptions. He then apportioned the net revenues after costs and taxes to each of the three geographical areas by using those countries’ relative telecom infrastructure expenditures for six of the eight IPCo franchises that Nortel had and apportioning all of the net revenues after costs and taxes for two of the franchises (PC and Internet advertising) to the U.S. He then applied a discount rate to those net cash flows allocated to each country.

[8] I have considerable difficulty with a number of aspects of Mr. Kinrich’s analysis. If the value of the net cash flows as stated by Mr. Kinrich is overstated, the overstated amount would belong to NNL, as the amount of the sales proceeds from the Rockstar transaction would represent more than the value of the net cash flows, which on Mr. Kinrich’s assumption is what the Licensed Participants gave up in the Rockstar sale. The expert evidence called by the Monitor is exactly to that effect, contending that Rockstar paid more than the value of the cash flow projections from the IPCo model for other motives.

[9] Nortel had no material business licensing its IP or monetizing its technology by suing others, either before or after filing for protection from creditors in early 2009. Mr. John Veschi had been hired in July 2008 to take responsibility for Nortel’s IP group and to look at options for licensing its IP.

[10] Development of the IPCo option was led by Mr. Veschi after the insolvency filings. The premise of IPCo was that the residual patents would be monetized by the threat of patent infringement litigation and, if necessary, actual infringement proceedings against various technology companies in an attempt to force such companies to pay royalties to IPCo. It was considered important that IPCo not carry on any telecommunications or other technology business, because, if it did, it would be vulnerable to counterclaims for alleged infringement being brought by the targets of its infringement litigation, which would undercut its revenue generating ability.

[11] Over the course of 2009 and 2010, Mr. Veschi and his team, assisted by Lazard Frères & Co, Nortel’s financial advisor, and Global IPCo, a law firm specializing in patent sales, prepared several versions of a preliminary financial model, in an attempt to forecast the operating profit that could be earned by IPCo so that the potential economic benefits could be weighed against value expected to be received on a sale of the portfolio.

[12] The various versions of the preliminary financial model had three sub-models, with differing assumptions relating to how much litigation IPCo would pursue. The scenarios were dubbed “Harvest” (assuming very little litigation), “Litigation Light” and “Litigation Heavy”. More litigation resulted in greater forecast revenues, at greater forecast cost. Assumptions regarding litigation success of 60 percent, 70 percent and 100 percent were used. A wide variety of assumed net cash flows were used and a variety of discount rates to value the cash flows were used.

[13] There is a difference in the evidence of Sharon Hamilton, a partner of Ernst & Young, the Monitor, and Mr. John Ray, the principal officer of NNI, as to how reliable the IPCo forecasts were. Ms. Hamilton was of the view that the projected cash flows were largely guesswork, given that Nortel had little experience in licensing and there were no good precedents about the estimated cash flow. Mr. Ray was more confident of the forecasts taken the work that went into them.

[14] What is clear is that there were a number of different models. Version 1 was presented on March 10 2010, version 2 on April 27, 2010, version 2.2 on May 6, 2010, version 3 undated, version 3.1 on October 25, 2010 and version 4 on November 18, 2010. Each version had different cash flow forecasts.

[15] I think it fair to conclude that the forecasts were not considered to be in any way certain. There were many permutations and combinations, and at no time did Nortel agree that any one forecast was the appropriate one. The process never got that far before the decision was made not to operate IPCo but rather to sell the residual IP.

[16] Mr. Kinrich chose to use version 3.1, although he did not explain why. Version 3.1 had the highest cash flows of all versions. It is noteworthy that the latest version 4 had projected cash flow forecasts of approximately half of what was projected in the earlier version 3.1 used by Mr. Kinrich.

[17] Mr. Green, an expert valuer called by the Monitor, points out that version 3.1 itself was not a finalized document or accepted by Nortel or its advisors. Within it there were a number of scenarios and options still being explored. The unreliability of the forecasts in the various models can be seen by the wide disparity in discount rates used. Lazard used discounts of 25, 35 and 45% to value the various projected cash flows. These are very high discounts, as more than one expert testified, and indicated a high risk to the cash flows being achieved. Mr. Kinrich used much lower discount rates of 12% and 15%, which I will come back to, which did not reflect the risks in the IPCo forecasts and which caused a higher valuation of the cash flows than would be the case if the discounts used by Lazard in the IPCo models were used.

[18] While there were multiple scenarios in the version 3.1 model, Mr. Kinrich used only the most aggressive case that maximized revenue. Mr. Green’s view is that there is inadequate explanation by Mr. Kinrich why the specific scenarios of Version 3.1 were selected for the analysis as opposed to other lower cash flow scenarios or the later Version 4 model with lower cash flows and as the analysis is unsupported, it makes the valuation unreliable. I must say that in reviewing the details of Mr. Kinrich’s report it is not at all apparent what his justification was for using the cash flows that he did. It leaves an open question as to the reliability of what Mr. Kinrich was doing.

[19] The value allocated to each of the debtors by Mr. Kinrich is based on the attribution to the geographic regions of the debtors of the projected operating cash flows in the IPCo model chosen by Mr. Kinrich. Those cash flows projected royalty payments on a regional level, namely North America, EMEA and China.

[20] The IPCo model estimated North American licensing revenue based on sales in Canada and the U.S. Mr. Kinrich apportioned the revenue to Canada and the U.S. using those countries’ relative telecom infrastructure expenditures, saying that relative telecom expenditures were a reasonable basis on which to estimate relative market size and were consistent with the structure of the IPCo model that used market size as the driver of royalty revenues. He did the same thing for EMEA as the IPCo model estimated EMEA revenue based on sales in France, Germany and the U.K.

[21] Global IPCo, the IPCo law firm retained to assist in preparing the models, stated early on in their work that they had no opinion regarding the territorial split of patents or patent-related revenue. There was certainly no agreement by any of the Nortel entities as to how the projected IPCo cash flows would be split territorially.

[22] Mr. Kinrich then deducted costs from the revenue streams including a number of litigation costs. It is not possible from looking at his report to know exactly what level of litigation costs was assumed by him.

[23] After calculating the net cash flows for each country, Mr. Kinrich then said he did a discounted cash flow calculation to arrive at a valuation for each country. In my view, Mr. Kinrich did not carry out a valid DCF valuation. The discount rate he used was not appropriate and was not derived by any conventional valuation approach.

[24] Mr. Kinrich acknowledged in his evidence that a DCF analysis requires knowledge about the cash flows over time and requires a discount rate to take those cash flows over time and convert them to present value. He acknowledged in his report that typically a discount rate is derived from the cost of capital (the cost of debt and equity split on some basis), referred to by valuators as the weighted average cost of capital. However, he did not do this. Instead he said that the value of the residual IP was known from the $4.5 billion paid for it by Rockstar and by taking his projected cash flows that he used from the IPCo model, he could back into (or reverse-engineer) a discount rate, being 12.2% when China is not included and 15% when China is included.

[25] This analysis proceeds on the assumption that the amount paid by Rockstar was based on the revenues taken by Mr. Kinrich from the particular IPCo model that he used. However, neither Mr. Kinrich nor anyone else knew what revenue streams were used by Rockstar to base their purchase price on or indeed, if Rockstar based their purchase price solely on anticipated revenues they could earn from the patent portfolio they acquired. Without knowing that, it is not possible to say that the Rockstar purchase was based on a discount rate of 12.2% or 15%. A discount rate, as Mr. Kinrich conceded, should reflect the risk of the cash flows being achieved, but without knowing what cash flows Rockstar based its purchase price on, saying the Rockstar purchase reflected a certain discount rate is artificial. Rockstar did not even know what the various IPCo cash flow models were.

[26] Mr. Green, Mr. Berenblut and Dr. Cox and Mr. Malackowski, all expert valuers, were critical of the method used by Mr. Kinrich to arrive at his discount rates of 12.2% and 15%. I accept their criticism. These discount rates were much lower than the rates used by Lazard in the IPCo models, including the very net cash flow model used by Mr. Kinrich, of 25% to 45%. Mr. Berenblut testified that he would expect the range of discount rates to be between 30% and 70%, recognizing the fact that this was a contemplated rather than an established business and recognizing the risks associated with it.

[27] Mr. Malackowski used a discount rate of 30% in his analysis of the potential revenue from the residual IP portfolio. He derived that rate by examining risk-adjusted hurdle rates associated with implementation of technology-based IP. These rates account for a buyer’s required rate of return or the associated risk of commercializing a technology.

[28] Mr. Kinrich in his report stated that the inferred rates of 12.2% and 15% that he obtained were consistent with discount rates observed in the market place at the time, being the median weighted average cost of capital for communication equipment companies. However, even Mr. Kinrich noted in his report that IPCo would not have been a communications equipment manufacturer. There was no analysis by Mr. Kinrich to lead to a conclusion that the cost of capital for a start-up litigation and licensing business would be comparable to an established communications equipment manufacturer. Messrs. Berenblut and Cox in their reply report stated:

The Kinrich Report’s use of discount rates for established publicly traded companies in the communications industry as benchmarks for its selection of discount rates for its valuation of a yet-to-be established business to exploit the Residual IP is not supportable. A discount rate of 30 percent or more for this type of business is consistent with our understanding and experience and is also consistent with the discount rates used in the IP Co Model. The academic literature reports venture capital discount rates in the range of 30 to 70 percent.

[29] I accept the criticism of Messrs. Green, Berenblut and Malackowski that the discount rates obtained by Mr. Kinrich were too low. Had Mr. Kinrich used higher rates such as those used by Lazard in the IPCo models, or the rate used by Mr. Malackowski, the value of the revenues given up by the Licensed Participants, assuming they belonged to the Licensed Participants, would have been far less than opined by Mr. Kinrich.

[30] Mr. Green calculated the values from the IPCo models using the discount rates used by Lazard in the models. Taking the most optimistic cash flows from the IP Co. model, the lowest discount rate used by Nortel and its advisors, and a litigation success rate of 100%, the maximum DCF value of IP Co. is only $2.7 billion, compared to the $4.5 billion paid by Rockstar. Messrs. Berenblut and Cox calculated that if a 30 percent discount rate is used to discount the cash flows used by Mr. Kinrich, the resulting net present value of the expected cash flows from the IP Co Model is $1.8 billion. They think this figure is overstated because of the range of values for all of the various scenarios in the IPCo models with various discounts of 25 to 45% and litigation strategies and assumed success rates of the litigation strategies from 60 to 75 to 100%. That range went from $424 million to $2.7 billion. Mr. Green put the range of values in the IPCo models from $400 million to $2.7 billion.

[31] The report of Messrs. Berenblut and Cox explains why Rockstar would be likely to have paid more for the residual IP than Nortel could have made from it, that is, on the theory that the Licensed Participants owned all of the benefits sold to Rockstar, more than what the Licensed Participants gave up in the Rockstar transaction. The defensive value of the residual IP to the members of the Rockstar consortium made the residual IP far more valuable to Rockstar than it was in the hands of Nortel.

[32] As explained by them, Rockstar obtained ownership of the residual IP and each of the members of the consortium (including Apple, Microsoft, Ericsson and Blackberry) received a license to the residual IP. The structure enabled Rockstar to exercise all rights of ownership of the residual IP against third parties, while providing the individual consortium members with the defensive benefits to prevent others from suing them for patent infringement. As a single company, Nortel was less likely to be able to derive defensive benefits equal to the combined and cumulative defensive benefits that could be gained by several large companies with extensive product and service lines that ranged well beyond what Nortel offered. Several members participating in the Rockstar portfolio are more likely to find patents contained in the Residual IP that will be useful to responses to litigation. Furthermore, as a company in financial difficulty, Nortel was less likely to be an attractive target for patent litigation and therefore less in need of patents to assert in response.

[33] Mr. Green also made the same this point. He stated that the members of the Rockstar consortium purchased the residual IP portfolio, at least in part, as a defensive measure. It was his experience that having access to a large patent portfolio can help protect a large technology firm from lawsuits from other large companies. Access to a large patent portfolio, like the residual IPCo portfolio, can act as a deterrent because potential opposing parties must factor in the probability of a counter-suit. The defensive value of access to a significant patent portfolio is valuable to purchasers like the Rockstar consortium members, but would not be relevant to an entity like IPCo which intended to pursue an offensive licensing and litigation strategy, but had no operating business in the technology sector as all such businesses had been sold. The defensive value of such a portfolio to large companies is not measured exclusively by the present value of the cash flows from licensing.

[34] Dr. Catherine Tucker, an economist called by the U.S. Debtors with considerable technology experience, stated the same thing. In her report she said that patents are not just used in litigation to assert rights to a particular technology or domain. There is also the important role of a patent being used in a counter-suit should the company itself be sued for patent infringement. She referred to Kent Walker, Google's General Counsel, who wrote at the time of the Rockstar bid that it was supposed to create a disincentive for others to sue Google. This defensive attribute, of course, would not have been available to IPCo if it decided to operate a patent licensing business as it would not have been in a product producing business that would be vulnerable to patent suits.

[35] Mr. Green also expressed the view that the identity of the bidders themselves in the residual IP auction also illustrates that the basis on which value of the residual IP portfolio was determined is not consistent with that in the Kinrich report. The bidders included Google, Apple, Microsoft, Ericsson and other large technology companies with worldwide operations rather than companies whose primary business model was patent licensing and litigation. If the value of the residual IP sale was closely related to the cash flows from a licensing/litigation strategy, one would expect licensing/litigation businesses to have been bidders in the auction. Instead, the bidders in the auction were operating technology companies, which suggests that the value of the residual IP was determined in the market on some strategic basis in addition to the value of the IP in a licensing/litigation business.

[36] I accept the evidence of Messrs. Berenblut and Cox and Mr. Green that the approach of Mr. Kinrich of allocating proceeds based on cash flows from a licensing /litigation business model such as the IPCo models is inappropriate and that what Rockstar paid for was more than the value of the potential revenues from the business that was being considered by IPCo. That is, it was more than what the Licensed Participants gave up in the Rockstar sale, assuming it was theirs to give up.

[37] The U.S. Debtors contend that it is wrong to say that Rockstar paid more than the value of what the Licensed Participants gave up when they terminated their licenses in anticipation of the Rockstar sale and to say that the extra value belongs to NNL as the owner of the NN Technology. They say that NNL could not transfer its rights without the consent of NNI and the EMEA Licensed Participants, just as NNI and the EMEA Licensed Participants required the consent of NNL to do so. They say that all parties consented to the transfer of their MRDA interests as part of the Rockstar sale, effectively agreeing to the assignment of their rights under article 14(e) of the MRDA which permitted an assignment of a party’s rights under the MRDA only with the consent of all of the other parties.

[38] I do not accept that contention. The MRDA did provide in article 14(a) that the MRDA could not be assigned by any licensed participant without the consent of the other Licensed Participants. But neither the MRDA nor the licenses of the Licensed Participants were assigned to Rockstar. Rockstar would not have taken an assignment of the MRDA with its obligations and duties amongst the participants. I accept the evidence of Mr. Britven, an expert valuer and the national intellectual property consulting practice leader with Duff & Phelps in Houston, that no third party would want to step into the shoes of a Licensed Participant by taking a transfer of the MRDA with its obligations to share profits and transfer ownership of patents to NNL, among other things. Even Mr. Ray eventually admitted that there was no transfer of license rights to Rockstar.

[39] What occurred was a sale of the residual IP to Rockstar with NNI and the EMEA debtors terminating their licenses under the MRDA as a condition precedent to the sale. What is at issue is the value of those licenses that were terminated. If the value of what could be earned from the licenses was less than Rockstar paid for the residual IP, the difference would belong to NNL, the legal owner of that IP.

[40] Mr. Green did an alternative valuation on the assumption, with which he disagreed, that IPCo would have operated on a stand-alone business and that the licenses surrendered by U.S. Debtors and EMEA debtors would have included the rights to the residual IP portfolio. He used version 3.1 of the IPCo model, as Mr. Kinrich had, but made some changes. He used the three discount rates that had been used by Lazard in the various IPCo models and used the three assumptions in the IPCo models as to the anticipated success in litigation against infringing third parties. He also deducted from the revenue streams going out to 2020 the RPS percentages for 2010 under the MRDA on the theory that if the Licensed Participants had rights under their licenses to earn the revenues proposed in the IPCo models, those licenses came with an obligation to make RPS adjustments in favour of the other Licensed Participants. Any gain on the sale above the DCF valuations on the revenue streams was allocated to Canada.

[41] If one assumed the median discount rate (of 35%) and the median litigation success rate (of 70%), and excluding the revenues from China, then Mr. Kinrich’s allocation of the Rockstar Sale proceeds, as adjusted by Mr. Green, would be as follows. Also shown is the allocation advocated by Mr. Kinrich.

Adjusted Kinrich Allocation of Rockstar Sale Proceeds

Kinrich’s Actual Proposed Allocation of Rockstar Sale Proceeds

Canada

$4,003.06 million

$430 million

U.S.

$346.12 million

$3,310 million

EMEA

$105.19 million

$710 million

Total

$4,454.37 million

[42] If revenues from China were included, the results would be an allocation of $3905.44 million to Canada, $420.99 million to the U.S. and $127.94 million to EMEA.

[43] The U.S. Debtors contend that Mr. Green was wrong to apply the RPSM to the value of the cash flows. They say firstly that the MRDA expressly provided in the third addendum signed in December 2008 that it does not apply to the sale of a business. What that amendment provided was that the operating income or loss used to calculate the RPSM was to exclude “gain/loss on the sale of business”. That is not a reference to the proceeds of the sale of a business, but rather a reference to the gain or loss, presumably capital gain or loss, recognized on a sale of a business. That makes sense because the RPSM was dealing with the split of profits or losses from operating earnings to be allocated to the participants under the MRDA. Ordinarily the gain or loss on the sale of capital assets would be recognized in an earnings statement but the parties to the MRDA did not want that taken into account in the RPSM.

[44] However, what Mr. Green was valuing in this analysis was the annual profits that would be earned by the Licensed Participants from operating IPCo in the future, assuming the Licensed Participants had the right to do so under their licenses. He was assuming that the profits would be split in accordance with the RPSM in the MRDA. I agree with the theory that if one is to value the benefits that could have been earned by the Licensed Participants if they had operated IPCo, which is what the U.S. Debtors say they would have done but for the Rockstar sale, the Licensed Participants would have been subject to some profit split.

[45] The U.S. debtors point out that what the profit split would be is a matter of conjecture and that it is not possible to assume, as Mr. Green did, that it would be the same in the future. The RPSM under the MRDA was based on the amount of R&D spend each year by Nortel and the Licensed Participants. After Nortel became insolvent, the R&D expenditures essentially stopped after 2009 and there is no evidence of what R&D would have been undertaken if IPCo had been run as a business by Nortel.

[46] Certainly there would have had to be some transfer pricing in place if Nortel had run IPCo as a business. What the parties would have worked out is unknown. The tax authorities would certainly have been interested in the transfer pricing associated with the running of the IPCo had that occurred and it does not mean that the parties would not have had to agree on a profit split of some sort. They would have been required to do so.

[47] It is perhaps fair to be critical of Mr. Green for assuming the transfer pricing would continue to be the same under an IPCo business run by Nortel as it had been before. It is also fair, however, to ask that if the U.S. debtors contend, as they do, that they are entitled to be paid for what they gave up in the Rockstar sale and that the present value of the anticipated net cash flows is what they gave up, one may have expected them to lead some transfer pricing evidence as to what transfer pricing would have been appropriate.

[48] The assumption that the transfer pricing that the parties would have worked out in the event that Nortel operated IPCo would have been the same as provided in the MRDA has some logic to it. The residual IP was created by R&D conducted by the parties, at least in part, during the MRDA that split profits on the basis of the R&D expenditures of NNL and the Licensed Participants. R&D was the driver of the profitability of Nortel and the RPSM was chosen at the request of the tax authorities as the most appropriate method for determining the compensation to each of the participants for the R&D performed by them. The profits to be earned from operating IPCo could perhaps be seen to be an extension of the results of the R&D that had been spent.

[49] The lack of transfer pricing evidence and analysis on the point, however, as to how the profits would be split in an IPCo business casts some doubt on the accuracy of Mr. Green’s alternative analysis. It is not a basis, however, to reject it out of hand as contended by the U.S. debtors.

[50] Mr. Malackowski’s preferred allocation approach is a contributions approach based on R&D expenditures made by each of the participants to the MRDA. He prepared an alternative revenue or licensed based allocation which contained dramatically different results from his contributions approach. His revenue approach allocated 33.6% of the Rockstar sale proceeds to the EMEA debtors versus 17.6% using his contribution approach. It allocated 11% to the Canadian debtors versus 39.5% using his contribution approach and it allocated 55.4% to the U.S. debtors versus 42.9% using his contribution approach.

[51] For his revenue or license approach, Mr. Malackowski used the data generated as a result of his valuation methodology to allocate the proceeds of the Residual IP Sale. He valued the Residual IP Portfolio by determining what revenues were expected to be generated by a worldwide licensing strategy in specific geographic territories and allocating the values to those territories. He estimated global revenues for the business areas in which the technology was used, royalty rates, licensing expenses, tax and discount rates. Mr. Malackowski concluded that the value of the residual IP was $3.570 billion, approximately one billion less than actually paid by Rockstar. He then “reconciled” this value with the actual purchase price of $4.5 billion by increasing pro rata the values he had calculated for each business franchise.

[52] For the exclusive territories of Canada, United States, Britain, Ireland and France, he allocated all of the value for those territories to each of the countries. For the rest of the world (“ROW”) he allocated 20% to each of the countries. It was this latter allocation of ROW that was the main cause of the increase in the allocation to EMEA as it had what he called “three seats at the table of five”.

[53] I have difficulty with Mr. Malackowski’s revenue or license model of allocating the Rockstar sale proceeds. The first is that there is no explanation by Mr. Malackowski why his market based valuation was $1 billion less than the actual sale proceeds. Rather than simply grossing his value up to “reconcile” it with the actual proceeds, it seems to me that his valuation was an indication that Rockstar paid for more than what could be achieved in revenues from the acquired IP portfolio. Mr. Green expressed the opinion that the adjustment was inappropriate and unsupported by valuation principles, and assumed that Rockstar just used different royalty or revenue assumptions. I accept that criticism.

[54] Mr. Green also expressed other criticisms of Mr. Malackowski’s calculations, all of which appear logical and which I accept. For example:

(i) Mr. Malackowski assumed all revenues for a country should be included in the royalty base, whereas he should have considered that only revenues from products and not services on which no patent royalty would likely be available.

(ii) Mr. Malackowski assumed that revenues from all licensees will begin to be earned in 2011 i.e. he assumed that all licensing efforts against dozens of targets across multiple jurisdictions would be 100% successful within a few months of the portfolio being sold. Mr. Green’s view is that his assumption is hard to credit and is inconsistent with the fact that the royalty rates selected by Mr. Malackowski are the IPCo “litigation light” rates, which would, by definition, require at least some form of enforcement action, which would necessarily delay the receipt of royalty payments.

(iii) Mr. Malackowski assumed increasing royalties through 2022 without considering that the patents and technologies are wasting assets and many are likely to expire before the end of the period used by Mr. Malackowski.

(iv) Mr. Malackowski deducted costs of 20% of royalty revenues, stating that he based the rate on the observed financial performance of sophisticated non-practicing entities such as Acacia Research Group. Mr. Green reviewed Acacia’s public filings and those of other licensing entities and have found a significant discrepancy between their reported costs and those that the Malackowski Report asserts are representative. The Acacia public filings disclosed that the company’s costs of operation from 2005 through 2012 have ranged from 112% of revenue in 2005 to a low of 52% of revenue in 2012. Other licensing entities, such as Interdigital and Rambus, report operating costs from 2005 to 2012 ranging from a low of 28% of revenues to as much as 164% of revenues.

[55] These errors lead to the conclusion that Mr. Malackowski’s valuation of $3.570 billion of the residual IP sold to Rockstar was likely overstated, indicating an even greater discrepancy between his value and the actual sale price. It also indicates issues with the territorial split of the revenues. The assumption of Mr. Malackowski that the entire sale proceeds were based on revenue forecasts by Rockstar, permitting him to simply increase his $3.570 billion value by another $1 billion without analyses ignores the likelihood that Rockstar paid what it did in part as a defensive move for its participants to protect their operating businesses, which Nortel no longer had. I do not have confidence in using Mr. Malackowski’s analysis to allocate the proceeds of the Rockstar sale on a license or revenue basis.

[56] In the end, I also cannot accept Mr. Kinrich’s calculation of the amounts from the Rockstar sale to be allocated to NNL, NNI and EMEA. Assuming the Licensed Participants had a right to the value of the residual IP that Nortel could have achieved, and looking at the various scenarios in the IPCo models, I would recalculate those values and allocate the proceeds by adjusting the calculations of Mr. Kinrich and averaging them with the calculations of Mr. Green in his alternative approach.

[57] I would take the mid-point between the low value of $400 million to $2.7 billion, or $1.5 billion using the discount rates of Mr. Green and Messrs. Berenblut and Cox. Using the same split as Mr. Kinrich, on the assumption that value would not be realized in China, would result in an allocation of 9.3% or $139.5 million to the Canadian debtors, 14% or $210 million to EMEA and 76.7% or $1.15 billion to the U.S. debtors. The balance of the $4.45 billion, or $2.9 billion, would be allocated to Canada. On the assumption that value could be realized in China, the resulting allocation would be 11.1% or $166.5 million to the Canadian debtors, 22% or $330 million to EMEA and 66.9% or $1.0 billion to the U.S. debtors. The balance of the $4.45 billion, or $2.9 billion, would be allocated to the Canadian debtors.

[58] I would then average these allocations with the allocations arrived at by Mr. Green in his alternative analysis, set out in paragraphs 358 and 359 above, which were based on the median discount rates and litigation success rates used in the IPCo models.

[59] The results of that allocation, assuming the revenues from China are included, would be an allocation to Canada of $3,485.97 million, to EMEA of $228.97 million and to the U.S. of $710.5 million, or a total of $4,425.44 million. I would round these figures up on a pro rate basis to arrive at the proceeds available of $4,454.37.

[60] The results of that allocation, assuming the revenues from China are not included, would be an allocation to Canada of $3,521.28 million, to EMEA of $157.6 million and to the U.S. of $748.06, or a total of $4,426.94 million. I would round these figures up on a pro rate basis to arrive at the proceeds available of $4,454.37.

[61] The U.S interests assert that on a license or revenue analysis, very little revenue should be attributed to China. They assert that the IPCo models included both a “China in” and “China out” option. I must say I have carefully looked at the IPCo model 3.1 used by Mr. Kinrich and I cannot find a China out option. On cross-examination of Mr. Malackowski, who thinks China revenues should be included, it was put to him that the IPCo model had a “toggle” for China, which I take to be a sheet with revenues for China.

[62] In any event, Mr. Kinrich testified that he at first took the mid-point of the particular China forecasts he used after doing an economic literature search on patent value and speaking with Mr. Zenkich, who told him that the market would pay little to nothing for a China patent, he reduced his revenues for China downward more towards the US in some qualitative fashion. He reduced then by 75%. Mr. Zenkich, an expert in valuing patents, testified that in 2009-2010 participants in the market for patent portfolios assigned little to no value to Chinese patents.

[63] The thinking of Nortel’s patent people changed over time. In December 2000, Angela Anderson, Director, Intellectual Property Law in the U.K stated that China was a sizeable and growing market accessible at moderate cost. She said that the target filing % (3% of cases) would be higher but for enforcement issues. “Show the flag, but don't over-invest.” She testified that at that time, it was clear that China was going to become more of a potential marketplace for Nortel products. In addition, the patent system was starting to look like a real patent system, so it made sense to start using the patent system in China at that time.

[64] By 2006, Nortel intended to file far more patents in China. The plan was to file up to 30% of the top patents in China and in 18 months’ time raise this to up to 50%, selecting those having the highest commercial potential. In the IPCo model of May, 2010 that included revenues from China, it stated that early 2010 modelling did not include China in its royalty base but the new plan included China but only in the years 2015 to 2020. It stated that 80% of its patents and 70 % of the applications in China were for wireless 4G technology. The logic of waiting until 2015 was the time for 4G market maturity. EMEA contends, and I have no reason to question it, that the assumptions in the IPCo model regarding China were conservative.

[65] Mr. Malackowski’s view was that in doing a revenue or license approach, it would be wrong to exclude China revenues. His reasoning was that Nortel had decided to file high interest patents in China, that patent protection was improving in China and had improved over the past five to ten years and that China was a very important and large market. He has had experience in China. His firm has a partner in Shenzhen for addressing the work they do in China.

[66] Mr. Zenkich testified that the basis for his conclusion that no one would pay anything for a Chinese patent was based on his business of being a patent broker. He testified that when his clients had large patent portfolios, there was no interest expressed in the Chinese patents that were part of those portfolios. Similarly, they were never asked by clients who looked to purchase patents to identify Chinese assets for purchase. I take this to be no evidence of knowledge of values that could be achieved for a Chinese patent, but only that Mr. Zenkich had no knowledge of a client being interested in in a Chinese patent. Included in material referred to in his report was a 2011 report entitled "China's Emerging Patent Trading Market” that referred to a patent auction in China in 2010 which sold 38 lots and the intention of the seller to hold another auction in 2011. The article also referred to efforts being made to set up an exchange in China with the support of governments that would facilitate transactions. That article was contradictory of the view expressed by Mr. Zenkich.

[67] Mr. Zenkich referred to a 2012 publication by the U.S. Patent Office that referred to comments it had received to the effect that there were difficulties with enforcing Chinese patents. That is certainly anecdotal evidence of statements made by others, and it cannot be belittled. How accurate are all of the statements is perhaps a matter of some debate. For example, a comment by one person as to the cap on damages in China was shown during the evidence to be incorrect. While Mr. Zenkich had stated in his report his belief that that significant interest in patent granting activity in China over the last ten years has increased the risk that patents may be challenged as invalid, even if granted, he acknowledged on cross-examination that he had no experience in trying to enforce patents in China and that his company had no experience in trying to enforce a patent anywhere in the world. He also acknowledged that he did not independently conduct surveys or seek out patent data of this kind of activity and that he was unable to identify a single instance where a Chinese patent was found invalid and its US or European counterpart was not. One of the documents cited by Mr. Zenkich in his report was a publication by a Beijing law firm of October 2009 that stated that the major cities, in particular Beijing, Shanghai and Guangzhou, can be considered as a reliable forum for patent infringement actions. Mr. Zenkich chose instead to rely on the U.S. Patent Office document that contained comments regarding the difficulty of enforcing patents in China.

[68] I am afraid that I cannot put a great deal of reliance on Mr. Zenkich’s evidence of the unreliability of the Chinese patent system. I accept he may be of the view that it is unreliable, but his view was not supported by any cogent, reliable and admissible evidence. The views of Mr. Kinrich are also not supported by any cogent evidence. He appears to have largely relied on Mr. Zenkich.

[69] In my view, if a license or revenue approach to value is to be used to value the residual IP, it should include revenues from China that were used in the IPCo model, mainly for the reasons expressed by Mr. Malackowski and the fact that the projections were somewhat conservative.

[70] The conclusion I come to, if an allocation of the proceeds of the Rockstar sale were to be based on a license or revenue approach, would be an allocation to Canada of $3,485.97 million, to EMEA of $228.97 million and to the U.S. of $710.495 million, or a total of $4,425.435 million. I would round these figures up slightly on a pro rate basis to equate to the proceeds available of $4,454.37.

(ii) Mr. Malackowski’s contribution approach to value

[71] The EMEA debtors contend that the allocation of the proceeds of the Rockstar sale should be made on the basis of the contribution to R&D made by each of the RPE entities that created the residual IP sold to Rockstar. They contend that the contributions by each RPE to measure this should not be the contributions made during the five year look-back period used to allocate the residual profits under the MRDA but rather the contributions made during the period of time that the residual IP that was sold to Rockstar was invented. Based on the evidence of Mr. Malackowski, they say the look-back period should be from 1991 to 2006[^26].

[72] There are two fundamental issues that have been raised to the calculations if the contribution approach to allocation is to be used. The Canadian Debtors contend that there is no basis to use a contribution approach to allocate the proceeds of the Rockstar sale or the business line sales. They say that if a contribution approach is nevertheless used, the look-back period for looking at R&D contributions should be the five year look-back period under the MRDA from 2005 to 2009. The U.S. Debtors also disagree that a contribution approach should be used to allocate the Rockstar and business line sale, but contend that if a contribution approach is used, they agree with the EMEA debtors as to the length of look-back period but contend that all R&D spending must be taken into account. They contend that what must be taken into account is not only the R&D costs incurred by each RPE in their own exclusive territory, but also all transfer pricing adjustments made by an RPE, particularly the adjustments made under the CSA agreements prior to the MRDA coming into force.

[73] Mr. Malackowski said in his report that to measure contribution, ideally, the contributions of the RPE’s labs to the development of the patented technologies could be fully and accurately determined by interviewing all of the firm’s R&D staff, and by reviewing all the documentation related to the firm’s research (e.g. lab notebooks, invention disclosures, meeting minutes, research presentations etc.). This approach was not possible for Nortel’s IP due to the size of the portfolio, the limitations on time and the availability of information. Mr. Malackowski did not have access to lab notebooks and R&D staff. Moreover, as R&D was organized across the Nortel Group and carried out in a highly coordinated and integrated manner across the various RPEs, it was even more difficult to separate out the distinct contributions of the various RPEs. In these circumstances he said he had to select a proxy data that reasonably reflected the research efforts of the various RPE’s labs.

[74] Mr. Malackowski chose to measure contributions to the development of the IP by measuring each RPE’s spending on R&D. He stated that in a large organization, where R&D funding supports a large number of R&D personnel and results in a large number of patents over time, this funding can be valid and indeed the most accurate proxy measurement for determining the contribution of each research group to the development of IP. He stated that it is common practice to regard each dollar spent on R&D as fungible for the purposes of measuring relative contribution to R&D in a group, as Nortel did under the RPSM.

[75] Mr. Malackowski stated that in his experience, in large IP portfolios the vast majority of the value of the portfolio is usually derived from a minority of the patents. This is due in part to the fact that technology IP can be overlapping and duplicative. Value is often derived from a relatively small number of patents that are essential to industry standard technology or that cover an essential process or solution to a common problem. Mr. Malackowski expressed the view that the patents that were categorized as high interest by Global IP likely represented the vast majority of the value of the residual patent portfolio. Approximately 37% of the total residual patent portfolio was identified as high interest.

[76] The evidence was that it generally took one year for Nortel R&D spending to result in a patent application for an invention. He therefore thought it appropriate to determine contribution to the creation of Nortel’s IP by measuring R&D spending starting the year before the filing of the earliest unexpired patent categorized by Global IP as high interest, i.e. in 1991. He stated that the most logical end point was in 2006, the year before the last high interest patent was filed. He provided calculations for four look back periods produced by two different start points and end points. His two start points were 1991, reflecting the year before the earliest unexpired high interest patent in the residual patent portfolio, and 2001. His two end points were 2006, representing the year before the last high interest patent in the residual patent portfolio, and 2008, representing the last year of ordinary course operations[^27]. 2001 was the start of the MRDA.

[77] By looking at the expenditures on R&D for this period from 1991 to 2006, Mr. Malackowski allocated 39.5% or $1.777 billion to Canada, 42.9% or $1.930 billion to U.S. and 17.6% or $793 million to EMEA. For the period 1991 to 2008, he allocated 40.6% or $1.827 billion to Canada, 43% or $1.935 billion to U.S. and 16.4% or $738 million to EMEA.

[78] The effect of using the longer look-back period substantially reduces the amount allocated to Canada, the reason being that the R&D expenditures from 2005 to 2009 during the five year RPSM were proportionally done more by Canada than EMEA and the U.S. The percentages from 2005 to 2009 were 49.5 for Canada, 38.8 for the U.S. and 11.7 for EMEA.

[79] Mr. Malackowski’s report contains discussion why he looked at a long period back to 1991 to measure R&D spending. He said that old patents maybe more valuable than recently filed ones. He said that technologies are adopted by the market slowly over time and do not realize their full value until later in the life of the patent. He did recognize that newer patents will have longer life before they expire and they may have favour due to technological obsolescence, but pointed out that there is risk in newer technologies that they may not be accepted by the market. Based on these considerations he concluded that he should take into account R&D spending from the year before the first high interest patent.

[80] Mr. Malackowski did not consider what Nortel’s thinking was about the life to its technology. In the first version of the MRDA the R&D spending used to split residual profits was calculated using an amortized 30% rate, with expenditures from any one year declining by 30% in the following years. In Nortel’s response to questions from the tax authorities in 2003 in connection with its request for an APA for that MRDA , Nortel stated:

It is difficult to ascertain the exact useful life of R&D developed at Nortel; however, Nortel's analyses indicated that a 30% amortization was conservative yet reasonable. Numerous sources suggest that the useful life of telecommunications R&D is short; however, there is no one definitive external source that explicitly determines that a 30% amortization rate is correct.

[81] The tax authorities did query this response in a question that referred to information from Nortel that it said seemed to suggest that the useful life of R&D is equivalent to product useful life. “However, isn't it the case that benefits from R&D may persist beyond product useful life? For instance, value may result from further developing the intangible.”

[82] In preparation for APA negotiations with the tax authorities, Gilles Fortier, NNL’s taxation manager for transfer pricing, circulated a document among Nortel tax executives dated May 10, 2002 summarizing the “key drivers” for Nortel, on the one hand, and the tax authorities, on the other, with regard to the APA. The position of the tax authorities was stated to be that the life of Nortel’s intellectual property was 7-10 years or more whereas Nortel was suggesting 4-7 years. This position of Nortel was consistent with using a 30% amortization rate for R&D spending in allocating profits under the CSA. Nortel wanted a shorter period because using a longer period would increase the profits in NNI for tax purposes that Nortel did not want. Canada had a lower tax rate due to its generous research and development policies.

[83] A later application by NNL and NNI for an APA with the tax authorities for the years 2007 to 2011, in which a straight five year R&D expenditure would be used to allocate profits, indicated that NNL and NNI thought that the useful life of the Nortel intangibles was estimated to be approximately five years with a gestation lag of one year. Included in the APA request was the following:

The economic life of technology is difficult to measure because as long as the technology is being sold, it is also being continuously updated and enhanced. Indeed, software and hardware development in the telecommunications industry is widely understood to be an iterative process, because of the tendency to superimpose improvements upon older versions of the technology. Therefore, any discussion of product useful life must consider when an individual product was originated, how to apportion the impact of successive improvements, and when the product was completely superseded.

Nortel's telecommunications technology consists of hardware and software, and it continues to grow and change as demand for bandwidth and functionality grows. As a result, there has been an evolution in the commercial and economic life span of technologies from longer to shorter cycles.

Nortel's Chief Technology Office estimated that a dollar spent on R&D typically has a shelf life of about five years, and additionally, the time from when the investment in the R&D is made to the time when revenue can be generated from the investment ranges from about 6 to 12 months.

Recognizing the difficulties inherent in estimating the useful life, based on information obtained in our discussion with Nortel management, and our review of the R&D policy documents, the useful life of the Nortel intangibles is estimated to be approximately five years with a gestation lag of one year.

[84] The evidence from Mr. Malackowski’s report is that 99% of the high-interest patents sold to Rockstar had an invention date prior to 2006 and the bulk were from 1995 to 2004. This is considerable evidence that what Nortel was telling the tax authorities did not turn out to be the case. This is not to suggest that Nortel did not believe what it was representing to the tax authorities, or perhaps more appropriately put, that Nortel’s transfer pricing tax people did not think that a legitimate tax case could be asserted supporting its 30% declining amortization calculation in the first MRDA and then its five year look-back period in the second version of the MRDA. It is clear, however, that Nortel expected negotiations with the tax authorities would take place that could alter the 30% amortization rate and the later five year flat rate, and the MRDA expressly contemplated that in Schedule A. It cannot be said that Nortel as an enterprise conclusively concluded that its profit allocation keys of 30% or five years were necessarily correct. It was a tax position prepared by Nortel and its advisors.

[85] If a contribution theory is to be used to measure the value of what the parties gave up, I think it inevitable that a longer look-back period would be appropriate. The market has indicated that. However, I would lengthen the time to be taken into account. One of the weaknesses of using a contribution approach is that not every dollar spent results in valuable technology. The theory then must be that what one loses in the corners is gained in the straights. That being the case, I see no reason to disregard the R&D expenditures in 2007 to 2009. They were real and cannot be said to have contributed to the residual IP sold to Rockstar[^28]. The fact that Rockstar has started out by enforcing earlier patents does not mean that later patents or patent applications will not be of value or that Rockstar did not pay anything for them.

[86] I would take the R&D expenditures from 1991 to 2009. The data is available from exhibit B.1.7.1 of Mr. Malackowski's report. The resulting percentage of expenditures is 40.93% for Canada, 42.87% for the U.S. and 16.2% for EMEA.

[87] The U.S. Debtors contend that because under the CSA agreement NNI was required to allocate transfer payments to other RPEs, those payments should be included in what is considered to have been contributed to R&D. They rely on upon the opinion of Laureen Ryan, a forensic accountant who went through the transfer pricing worksheets and calculated $4.4 billion allocated to other RPEs under the CSA agreement. On her figures, the percentages for R&D expenditures for 1989 to 2000 would be 21% for Canada, 6% for EMEA and 73% for the U.S.

[88] There is a problem with Ms. Ryan’s evidence. The first is that she did no cash analysis to determine if NNI actually paid out any cash to any other RPE as part of its transfer pricing requirements under the CSA and later MRDA. There is no evidence in the record that anything allocated to any party was actually transferred by way of cash and Ms. Ryan conceded that she could not say if anything was actually paid. She did speak to her general understanding that money was transferred by NNI to NNL but I take that to be hearsay evidence and not any cogent evidence that any funds were transferred in fact. Just as important, there was no evidence as to how cash transferred from NNI or any other RPE was actually used. Cash was moved throughout the Nortel Group as required, but what those requirements were at any time is not a matter of record or available evidence. Ms. Ryan also conceded that she was not able to say where any of the money came from to actually do the R&D spending, whether from customers, governments, shareholders or other Nortel entities.

[89] While Ms. Ryan in her report and evidence calculated what she said were allocations for R&D made by NNI to the other RPEs under the MRDA, the U.S. Debtors made no argument in their closing briefs that these payments should be attributed to NNI. One problem with the evidence on this point is that Ms. Ryan assumed that the RPEs used transfer pricing adjustments for only for only two types of expenses: direct R&D spending figures, and sales, general, and administrative costs. Ms. Ryan pro-rated the intercompany funding between those two expenses. That assumption was obviously incorrect because, as Ms. Ryan conceded, it ignores very significant additional costs incurred by the RPEs, including restructuring costs, costs of revenues, manufacturing, and distribution. The very need for an assumption to be made was because Nortel never kept records of what transferred cash from one Nortel company to another was used for. Ms. Ryan also erred in failing to deduct the $2 billion settlement with the IRS and CRA regarding the $2 billion that was deemed to be a dividend paid by NNI to NNL. She also failed to take into account the sale of Nortel’s UMTS business to Alcatel.

[90] As stated above, Mr. Malackowski thought that ideally to determine contribution to R&D by any particular RPE, he would need to have access to lab notebooks and other records and to Nortel R&D personnel. As he did not have that he had to select a proxy data that reasonably reflected the research efforts of the various RPE’s labs. He chose to measure contributions to the development of the IP by measuring each RPE’s spending on R&D. He testified that this would be reflective of the types of activities that we know lead directly to the inventive process. It is the engineering time and the related expenses that result in the innovation. He testified that a transfer pricing adjustment is an allocation that is done for other purposes, specifically tax efficiency, not for recording the matching between the inventive nature of contribution and results, and he viewed it as inappropriate.

[91] Ms. Ryan is a specialist in accounting and forensic investigations. I prefer the evidence of Mr. Malackowski on this point that for his contribution analysis, it is not appropriate to add to any RPE’s contribution amounts that were allocated from that RPE under the transfer pricing regimes in the CSA or MRDA.

[92] Mr. Malackowski did an “inventorship” analysis in his reply report of the countries in which the inventors of the residual patent portfolio resided. He stated that while he did not consider inventorship to be the appropriate basis for allocation, it was a useful metric for testing the allocations of the various parties.

[93] The results of Mr. Malackowski’s analysis indicated that for the high interest patents, 46.3% were from Canada, 33% from the U.S., 18.7% from EMEA and 2.6% from ROW. For the entire portfolio, 51.9% were from Canada, 27.4% were from the U.S., 17.7% were from EMEA and 2.9% were from ROW. Using the percentages for the entire residual patent portfolio, which is what was sold, and allocating ROW equally to the others, would give Canada 52.9% of $4.45 billion or $2.35 billion, U.S. 28.4% or $1.26 billion and EMEA 18.7% or $832 million.

[94] Mr. Britven, an expert called by the Monitor, while of the opinion that a contribution allocation theory was not correct, also did an inventor based analysis. That analysis allocated 51.3% to Canada, 28.9% to the U.S., 18.2% to EMEA and 1.6% to others. That is very close to the figures from Mr. Malackowski’s inventorship analysis

[95] I conclude that if the contribution allocation theory asserted by the EMEA debtors is accepted, the percentage allocation of the residual IP sold to Rockstar of $4.45 billion is 40.93% or $1.82 billion for Canada, 42.87% or $1.92 billion for the U.S. and 16.2% or $720 million for EMEA to be rounded down pro rate to get a total of $4.45 billion.

(iii) Mr. Green’s approach

[96] Mr. Green allocated virtually all of the proceeds of the Rockstar sale to Canada.[^29] There were two categories of patents involved in the sale:

patents that had been used in several business lines and in respect of which non-exclusive licenses had been granted to the business line purchasers; and
the remaining patents, which had not been used in any Nortel business.

[97] For the group of patents identified in (1) i.e. patents that had been used in several business lines and in respect of which non-exclusive licenses had been granted to the business line purchasers, the value of the U.S. and EMEA Debtors’ licenses with respect to those patents (which is the value that they would have earned had they continued to operate the businesses) was determined by Mr. Green and allocated to them as part of his allocation of the business line sale proceeds.

[98] With respect to those patents described in (2) that were not used in any of Nortel’s operating businesses, Mr. Green considered whether there was any evidence that the U.S. and EMEA Debtors had any prospect of generating earnings through the exercise of their license rights in connection with those patents. He concluded that they did not because the U.S. and EMEA Debtors’ license rights were limited to the right to make Products – i.e. products made or designed (or proposed to be made or designed) by or for a Participant, embodying or using the Nortel IP. This was consistent with the position taken by the Monitor in this case. Thus he allocated none of the proceeds of the Rockstar sale to the U.S. and EMEA Debtors and all of the proceeds to Canada.

[99] Mr. Green’s valuation is a straight result of the interpretation put on the MRDA by the Monitor. One cannot quarrel with the logic of it if that interpretation were to govern the allocation.

CITATION: Nortel Networks Corporation (Re), 2015 ONSC 2987

COURT FILE NO.: 09-CL-7950

DATE: 20150512

ONTARIO

SUPERIOR COURT OF JUSTICE

COMMERCIAL LIST

IN THE MATTER OF THE COMPANIES' CREDITORS ARRANGEMENT ACT, R.S.C. 1985, c. c-36, AS AMENDED

APPLICATION UNDER THE COMPANIES' CREDITORS ARRANGEMENT ACT, R.S.C. 1985, c. C-36, AS AMENDED

REASONS FOR JUDGMENT

Newbould J.

Released: May 12, 2015

[1] EMEA is an acronym for 19 Nortel subsidiaries in Europe, the Middle East and Africa.

[2] All reference to dollars is to U.S. currency.

3 Judge Kevin Gross is the U.S. bankruptcy judge.

[4] See Nortel Networks Corp. (Re), (2013), 2 C.B.R. (6th) 1;aff’d (2013), 2013 ONCA 427, 5 C.B.R. (6th) 254 (Ont. C.A.); 2013 WL 1385271; aff’d 737 F.3d 265.

[5] A later Allocation Protocol which set out procedural matters to govern the allocation hearing was made and approved by orders of both Courts in May, 2013.

[6] Unless otherwise indicated, statements of fact in these reasons are findings of fact.

[7] Nortel Networks Australia was also a RPE until December 31, 2007.

[8] This was an alternative argument for the CCC to its first argument that the MRDA should govern the allocation.

[9] There were different CSAs for different types of costs. The relevant CSAs were the R&D CSAs that provided for the sharing of costs of the R&D carried out by the Nortel entities doing R&D. NNL made a separate CSA with each of those entities.

[10] I prefer this test to that articulated in Ventas, Inc. v. Sunrise Senior Living Real Estate Investment Trust (2007), 2007 ONCA 205, 85 O.R. (3d) 254 (C.A.), in which it was said that interpreting a contract that accords with sound commercial principles is limited to situations in which there is some ambiguity. I do not think that is correct and it is not what other cases of appellate authority have stated. See my comments in Thomas Cook Canada Inc. v. Skyservice Airlines Inc. (2011), 83 C.B.R. (5th) 106 at para. 13 and Oncap L.P. v. Computershare Trust Co. of Canada (2011), 94 B.L.R. (4th) 314 at paras. 21 to 24. See also Geoff R. Hall, Canadian Contractual Interpretation Law, 2nd ed. (Markham Ont.:LexisNexis 2012 at p. 46 fn. 191.

[11] There was a separate R&D CSA made with each participant. They were the same. Reference during argument was to the CSA made between Northern Telecom Limited [now NNL] and Northern Telecom Inc. [now NNI], and I refer to it in these reasons.

[12] The amended Schedule A was effective January 1, 2006 and reflected a change in the calculation of the amount spent on R&D by each participant.

[13] The NN Technology in the MRDA was called the NT Technology in the CSA as the parties at the time of the CSA in 1992 were Northern Telecom, later changed to Nortel Networks.

[14] Rulings on admissibility of evidence were left for decision to be made after argument at the conclusion of the trial.

[15] At page 30 of the report, Horst Frisch, in referring to intercompany transactions between participants under a RPSM allocation, state-“The old CSPs possess and will continue to possess valuable intangible property.” What property is being referred to is not stated. It could be a reference to license rights.

[16] C.I. Covington Fund Inc. v. White, 2000 CanLII 22676 (ON SC), [2000] O.J. No. 4589 paras. 38–39 (S.C.J. (Commercial List)), aff'd 2001 CanLII 28384 (ON SCDC), [2001] O.J. No. 3918 (Div. Ct.). G.D. Searle & Co. v. Novopharm Ltd., 2007 FCA 173, [2007] F.C.J. No. 625 (C.A.), leave to appeal to SCC refused [2007] S.C.C.A. No. 340 . Patchett v. Sterling Engineering Coy. Ltd. (1955), 72 R.P.C. 50 (H.L.). This has now been codified in section 39 of the Patents Act 1977 (U.K.), c. 37.

[17] Board of Trustees of Leland Stanford Junior Univ. v. Roche Molecular Sys., Inc., 131 S. Ct. 2188, 98 USPQ (2d) 1761 (2011) at p. 2195-2196, quoting United States v. Dubilier Condenser Corp., 289 U.S. 178 (1933) at p. 189.

[18] See the affidavit of Peter Currie sworn April 11, 2014 for a full description of Nortel’s matrix structure and operations.

[19] Early in these proceedings, on the motion in 2009 to approve the IFSA, counsel to the U.S. Debtors stated in its written brief that NNL owned the IP. The report of the administrators for the EMEA Debtors of June 14, 2009 stated that all IP rights belonged to NNL. Once the size of the sale proceeds became known, these positions of the U.S. Debtors and EMEA Debtors changed.

[20] In In re Owens Corning, 419 F. 3d 196 at 205 (3rd Cir. 2005) the U.S. Court of Appeals observed that substantive consolidation “treats separate legal entities as if they were merged into a single survivor left with all the cumulative assets and liabilities, (save for inter-entity liabilities which are erased). The result is that claims of creditors against separate debtors morph to claims against the consolidated survivor.”

[21] The projected cash on hand in all of the Nortel entities as of June 30, 2014 after payment of secured creditors was $1.525 billion, being $343 million in the Canadian Debtors, $744 million in the U.S. Debtors and $438 million in the EMEA Debtors. See schedule 5 of the Britven report, ex. 45.

[22] Mr. Kilimnik prepared an expert report on which he was deposed prior to the trial. At the opening of the trial, counsel for the ad hoc group of bondholders said that Mr. Kilimnik would be called as a witness. However, on the day before he was scheduled to testify, his report was withdrawn by the bondholders and he was not called as a witness at the trial.

[23] I also prefer the evidence of Mr. Kilimnik and Mr. Binning as to the data in exhibit 58 that compared Nortel bond spreads to government yields and what could be drawn from it. Professor McConnell said he could not draw an inference from the data but also said that he was not contradicting Mr. Binning.

[24] There was one series of bonds for $200 million issued by NNL with a NNC guarantee but no guarantee by NNI.

[25] The CCC contended for an “ownership” allocation very similar to the Monitor, being $5.805 billion to the Canadian Debtors, $1.009 billion to the U.S. Debtors and $488 million to the EMEA Debtors.

[26] For the IP sold in the business line sales, EMEA says that the look-back period should be from 1991 to 2008, two years longer than for the Rockstar sale.

[27] Mr. Malackowski said he did not think it appropriate to look at 2009 R&D expenditures post-filing as he understood that little basic research was being performed during this time given that R&D spending was cut dramatically and none of the patents designated as high interest by Global IP were filed during this time period. The R&D expenditures in 2008 were $1.458 billion and in 2009 were $1.076 billion. Mr. Malackowski also said an appropriate look-back period for the business sales would be 2001 to 2008.

[28] The Canadian expenditure in 2009 was not just to preserve the business lines as asserted by EMEA. Canada spent $564 million in 2009 on R&D, far more than the $180 million spent on the CDMA and LTE businesses.

[29] He allocated $426,097 to the U.S. representing the value of the workforce transferred to Rockstar, being very few people.

CITATION: Nortel Networks Corporation (Re), 2015 ONSC 2987

SUPERIOR COURT OF JUSTICE – ONTARIO

COMMERCIAL LIST

COUNSEL:

Prologue

Nortel history and its matrix structure

Bankruptcy filings

Decision to liquidate

Interim Funding and Settlement Agreement (“IFSA”)

Asset sales

Position of the parties

The MRDA

(i) Governing law of the construction of the contract

(ii) Position of the parties

(iii) Analysis

(a) The meaning of the exclusive license

(b) The right to sue for infringement

(iv) Surrounding circumstances or the factual matrix

(a) 1996 APA

(b) Sub-licences

(c) Representations to tax authorities

(d) Avoiding permanent establishment status in the U.S.

(e) Patent litigation

(f) Conclusion of factual matrix evidence

(v) Commercial reasonableness

(vi) Conclusions on the meaning of the MRDA

Applicability of the MRDA to the allocation issues

EMEA position on ownership of the Nortel IP

Appropriate method to allocate the proceeds of sale

Appropriate pro rata allocation method

Allocation on a basis other than pro rata

Conclusion

Epilogue

APPENDIX A

(i) Mr. Kinrich

(ii) Mr. Malackowski and Mr. Huffard

(iii) Mr. Green

APPENDIX B

(i) Mr. Kinrich’s license approach to value

(ii) Mr. Malackowski’s contribution approach to value

(iii) Mr. Green’s approach

CITATION: Nortel Networks Corporation (Re), 2015 ONSC 2987

Judge

Counsel

Parties

Areas of Law

Fact Patterns

Expert Witnesses

Financial Outcomes

Issues

Statutory Provisions 1

Legal Tests 3

Legislation 1