FOR AMANDA SMITH ONLY

Thomas MacMillan leaned back in his chair and glanced out of his office window down onto Bay Street, the epicenter of the Canadian financial industry. During his 10-year tenure as president and CEO of CIBC Mellon, MacMillan had presided over the dramatic growth of the jointly owned, Toronto-based asset servicing business of CIBC and The Bank of New York Mellon Corporation (BNY Mellon). However, now it was an overcast day in mid-September 2008 and MacMillan had a front-row seat to witness the onset of the worst financial crisis since the Great Depression. CIBC Mellon was facing this oncoming global financial storm with a solid balance sheet and was secure in the knowledge that both of its parents were also well capitalized. However, the well-publicized impending collapse of several long-standing financial titans threatened to impact all players in the financial services industry worldwide. Despite the fact that joint ventures (JVs) were uncommon in the financial sector, MacMillan believed that the CIBC Mellon JV was uniquely positioned to withstand the fallout associated with the financial crisis and that it would be able to weather the most significant risks facing the JV—execution risk and the potential exodus of assets and clients who were panicked by the wider financial pandemonium. MacMillan and his team recognized that it would be critical for the JV to continue to deliver a high level of client service and to avoid any major operational missteps. MacMillan’s moment of introspection was interrupted by a knock on the door. He was scheduled to meet with three members of the company’s executive management committee, Paul Marchand, Mark Hemingway and James Slater, to discuss two pressing issues facing the JV. First, they needed to discuss how to best manage any risks confronting the JV as a consequence of the financial crisis. Given the massive size and global reach of the largest financial service giants, and the likelihood that some of these behemoths might now be teetering on the edge of bankruptcy, CIBC Mellon, like other players in the financial services industry, would be forced to move adeptly to protect its operations from any potential exposure to the larger players’ fates. While the systems, structure and culture that prevailed at CIBC Mellon served as evidence of MacMillan and his team’s diligent efforts over the past 10 years to focus on risk management and to foster a culture of synergistic cooperation, the question remained—how could the policies and practices developed during the past decade be leveraged to sustain the JV through the broader financial crisis? Second, the four men were scheduled to continue discussions regarding options for refining CIBC Mellon’s strategic focus, so that the JV could emerge from the financial meltdown on even stronger footing. Notwithstanding the immediate urgency of the financial crisis, the JV’s management team recognized the need to continue to manage the business with a view towards future growth. asis was on asset servicing—the global custody business—which was generally viewed as “a dull business, with dull services, in a dull little corner of the financial services sector.” Asset servicing delivers securities- related administrative services to support the investment processes and goals of clients. Such services include global custody, securities lending, cash management, multicurrency accounting and reporting, global performance measurement and analytics, transition management, commission recapture and foreign exchange. Clients include pension plans, investment managers, mutual funds, insurance companies, and global financial institutions. The fees charged to provide such administrative services would typically be much less than one half of one per cent of the value of the asset being supported.

In 1996, CIBC was one of the big five Schedule 1 (domestic) banks in Canada. At that time, it had an average custodial operation, with approximately 14 competitors, principally trust companies and the security departments of the major banks. CIBC had $100 billion in custody assets and a handful of clients. Its technology platform was poor and needed significant investment. It had three choices to make:

1. Invest—theproblemtherewasthatitwouldhavehad to invest a lot of money ($300, $400, $500 million)— to come up with a world-class custodial system. It was concerned that the revenue potential from the Canadian marketplace would not have resulted in it receiving adequate returns for its investment.

2. Exit the business—as a lot of companies in Canada subsequently did.

3. Form a joint venture if it could find the right partner. CIBC believed there was the potential of creating a good, viable Canadian-based business, but it needed a partner.

In 1996, Pittsburgh-based Mellon Bank had a Schedule II (Canadian bank which was a subsidiary of a foreign bank) banking operation in Canada, which MacMillan ran. Its Canadian custody market share was about one per cent—specifically one client, Cdn$8 billion in assets under administration, which was being administered out of Boston. However, Mellon had a world-class technology platform, had scale (in a scale business) and was committed to growing its market share but was having difficulty breaking into the Canadian marketplace. It knew it would be a difficult

Globe: © Jan Rysavy/iStockphoto.com

and slow process to get established in Canada by setting up a greenfield operation.

CIBC approached several potential partners, but it was Mellon’s technology and people that impressed it the most. CIBC had a Canadian presence and a client base, but no technology and its service was average. Mellon had great technology, products and services, but no presence in Canada and few clients in the country. And it was receptive to CIBC’s overtures. It seemed the ideal circumstance for the birth of a joint venture—a great fit. Both parties needed each other and there was very little overlap. The opportunity to create a world- class Canadian asset servicing company—when not many existed in Canada at the time—was too enticing to pass up for both organizations. CIBC Mellon was the first significant JV for either parent.

Given the relative rarity of JVs in the financial services sector, the two sole shareholders in the proposed JV (CIBC and BNY Mellon) devoted a considerable amount of thought and planning during 1996 to the design and structure of the entity. A significant volume of legal agreements was negotiated to establish the parameters that would govern the relationship between the two shareholders. Buried within the reams of legal documents were provisions drafted to prohibit each of the shareholder parents from competing against the JV; to detail the limitations surrounding the JV’s use of the parent shareholders’ trademarks and other intellectual property; to outline the basis upon which each parent would provide services to the JV, including, in the case of BNY Mellon, the basis upon which it would provide and “Canadianize” its technology for the JV; and to require each of the shareholders to utilize the JV as a supplier of asset servicing and global custody services.

One of the most critical governance clauses pertained to voting rights. Under the JV’s shareholder agreement, each of the parties would enjoy a 50 per cent vote on every issue. In effect, this eliminated the need to vote on any issue—only consensus could yield a decision and the JV managers needed to secure the approval of both shareholders before taking any major issues to the board. Accordingly, at the outset MacMillan and his team recognized that in order for the JV to execute on its mission both strategically and operationally, it would be critical for the two shareholders, their employees and the JV’s employees to develop an acute understanding and respect for the unique capabilities that each shareholder brought to the JV. MacMillan acknowledged that “the governance processes developed for this JV effectively facilitated ought to the JV. MacMillan acknowledged that “the governance processes developed for this JV effectively facilitate.

markets, BNY Mellon had a strong understanding of credit in the global custody market and enjoyed a strong reputation with federal regulators in the United States. CIBC, on the other hand, had a strong understanding of credit in the Canadian marketplace and was known for its strong global trading platform (CIBC World Markets).

Equally important were provisions pertaining to risk management. The shareholders agreed to the formation and to the membership on the JV’s Asset & Liability Committee (ALCO). ALCO was tasked with the responsibility of overseeing the formulation of risk management policies and asset investment policies associated with the JV’s treasury and securities lending activities—principal activities under which financial services firms could become exposed to credit risk and market risk. This pivotal committee was populated by senior management from CIBC, BNY Mellon and the JV itself. MacMillan acknowledged that both shareholders had sought to structure the JV to develop a discrete, low-risk business and that risk tolerance would be maintained in the parents’ businesses. As such, the JV only engaged in very conservative transactions and did not engage in proprietary trading. Appropriately managing risk necessitated clear and constant communication in order to ensure that the JV was aligned with its shareholders. It also effectively positioned the JV’s management team to tap into the knowledge assets and accumulated experience of two major financial institutions.

When formed at the end of 1996, the JV had fewer than 200 employees, $110 billion in assets under administration, a market share of less than 10 per cent and revenues of about $25 million. However, over the next decade, the business grew dramatically. In 1997, it acquired the Canada Trust custody business. In 1999, it acquired the Bank of Montreal custody business, in 2002, the TD Bank third-party custody business, and in 2006, it was awarded the IG/Mackenzie custody business. By 2006, there were 1,400 employees and 1,140 custody clients. At this point, the asset servicing business offered a wide and integrated range of products and services from custody to risk management which could be grouped into two broad categories—core asset servicing functions and capital markets functions (see Exhibit 1). Historically, each of these two categories of business functions contributed approximately 50 per cent of the profits generated by the JV’s asset servicing business. While the core asset servicing business functions supplied a stream of recurring-fee revenue to the JV, the income stream generated by the capital markets functions could be more volatile, depending upon the state of the capital markets. The global

securities lending component of the capital markets functions involved acting as an agent in facilitating the lending of debt and equities from the JV’s clients to other clients, who were typically brokers. While they did not disclose to CIBC Mellon why they were undertaking any particular loan, it could be expected that the brokers that borrowed the assets from CIBC Mellon would utilize the assets both for their own proprietary trading and to loan to the brokers’ clients, sometimes including hedge funds that pursued short positions in equities. Short positions were established by traders who sold equities that they did not currently own. In essence, short sales involved selling borrowed equity assets. Consistent with regulatory requirements and its low-risk culture, CIBC Mellon routinely secured the loans that it extended to its broker clients by requiring the borrowers to pledge high-quality assets in excess of the value of the underlying loans as collateral. Exhibit 2 illustrates the interactions that occurred between external parties and CIBC Mellon’s securities lending service.

By 2007, assets under administration for the JV’s asset servicing business exceeded $800 billion, and were growing. The JV had become the second-largest asset servicing business in Canada, with a market share over 30 per cent. It was settling 15,000 transactions each day. Total revenues for CIBC Mellon exceeded $350 million, and healthy quarterly dividends were being paid to each partner.

The Stock Transfer and Corporate Trust Businesses In 1997, the JV entered the trust services business through CIBC’s purchase of a 50 per cent interest in Mellon’s R-M Trust Company. The purchase was undertaken because the JV required a trust company as a deposit-taker for its asset servicing business and because R-M also had established stock transfer and corporate trust businesses. Through this business, Canadian companies that issued securities that traded on major stock exchanges relied on CIBC Mellon to manage administrative duties like security holder record keeping, securities transfers, investor communication, dividend payments and employee plan administration. The JV also acted as a corporate trustee for its trust clients’ assets. In its corporate trustee role, the JV acted as indenture trustee for a number of series of asset-backed commercial paper (ABCP). ABCPs were typically short-term commercial paper investments that were collateralized by other financial assets which were characterized by very low risk. As of 2006, the JV had 1,200 trust clients. CIBC Mellon did not borrow or lend any securities in connection with this line of business.

e felt many joint ventures failed (see Exhibit 3) and the reasons why CIBC Mellon had been so successful. Let me start with the one main reason that towers over all of them: our people. They are amazing and we have together somehow created an atmosphere where we can all thrive. Our people are our big differentiator. This can happen in any company. It has happened in our JV. We have succeeded because: ■■ The original business plan made sense. CIBC Mellon is profitable, growing, with good returns because the original business rationale was solid. It wasn’t two “lousy” businesses coming together. Outstanding Mellon technology and service was introduced into Canada relatively quickly through a JV that had CIBC in its name and made an immediate positive market impact. ■■ The parents receive benefit from the JV itself in the form of dividends but also from outside of the JV. Both parents make significant FX revenues for their own books from JV clients. We help Mellon win global custody bids, we help CIBC win additional banking business that is often tied to asset servicing, for example cash management. We contribute to the building of strong client relationships for both parents. Mellon gets to appropriately allocate costs to the joint venture connected with their technology spend—and this spend is significantly larger than what they could otherwise afford because of the JV. ■■ Both parents cooperate. I see it at every board meeting—they respect and appreciate the contribution of the others; work collaboratively to make the JV a success. Despite the historical, jurisdictional and managerial differences, we’ve managed to put these differences to the side to make the JV work. And when there are differences of opinion (and frankly there’s not that many) they work it out; and both organizations complement each other: CIBC defers to Mellon’s expertise in the global custody business; Mellon defers to CIBC’s knowledge/expertise relative to Canadian business and banking. I think it helps that both banks had a good solid friendship for many decades prior to the formation of the JV. They were (and are) comfortable with each other. They don’t really compete against each other in any major business lines. Canadians are generally comfortable working with Americans and vice versa. Both parents’ head offices are in the same time zone. ■■ Commitment of the parents. They have from the very beginning wanted to see the JV succeed and grow and they spent time making this happen. When more capital was required for acquisitions (the asset servicing business of Canada Trust, BMO, TD)—both parents were there. Early on in our history we had teething pains (our level of client service was not what it is today)—the parents didn’t waiver and constructively helped us overcome issues and push on. ■■ The JV has effectively leveraged the strength of both Mellon has 50 per cent ownership, but we benefit 100 per cent from their ongoing technology spend—over US$200 million a year. On a stand-alone basis, we couldn’t afford US$200 million each year for technology in support of asset servicing business. This is an enormous plus for us. From CIBC, we consistently leverage their client banking relationships to win new business for CIBC Mellon. We also leverage governance standards and risk management practices from both partners. ■■ And finally—our company is well managed. We have a strong board composed of mature, competent executives who can speak for their organizations. There has been minimal board turnover and when it has happened the transitions have been smooth. My lead board members from both organizations have been there since day one.

But equally important, I am blessed with an extremely strong executive management team. This team gets direction from the board and we run with it. They are also very skillful at working with both shareholders to ensure that all interests are balanced and satisfied and the strengths of both parents are fully realized. This is a skill requirement unique to joint ventures. For example:

■■ Leveraging technology development at Mellon and sales development at both parents.

■■ Working with both parents’ risk, audit and compliance.

■■ It is a skill to unleash the power of our parents with- out being overwhelmed by them.

And while we are proactive in leveraging the strengths of our shareholders, we never forget we are a stand-alone organization that needs to be managed effectively. We have developed a strong internal culture quite independent of our parents, including our own strategy, brand, vision and core values. And we’ve shown enormous skills in imitating and successfully concluding major acquisitions at critical times.

The markets have been generally favourable the last 10 years. There has been tremendous growth in the mutual fund industry, a major sector for us. There has been an accelerating trend to globalization of the capital markets, including increased complexity of financial instruments, and heightened requirements for reporting and transparency and real time information. These all play to our strengths. Not all of this was anticipated back in 1996. Over the past 10 years, we’ve had a good tailwind. You are better to be lucky than good, but we have been good.

2007–2008: From Tailwind to Headwind

The growth that had characterized the global financial sector up until 2006 began to materially change in 2007. A rapid series of problems began to either emerge or become more widely acknowledged. Fundamental differences that existed between the Canadian and U.S. banking sectors posed a unique set of concerns for financial institutions with operations on both sides of the border. Discrepancies in consumer debt and equity levels, divergent banking regulations and differences in the structure of each country’s mortgage security industry comprised some of the most significant concerns.

The Canadian Financial Sectori

Canada as a whole was entering the crisis with a strong balance sheet and economic position. Consumers had lower debt and more savings than in the United States.

Mortgages were originated and held by Canadian banks, not packaged up and sold as securities. Canadian mortgages were generally five years or less, and mortgage interest was not tax deductible in Canada, so homebuyers were not encouraged to buy beyond their means. There were no 40-year terms, and buyers had to be able to have a down payment. Canadian Banks could not lend more than 80 per cent of the value of a house without mortgage insurance from the Canada Mortgage and Housing Corporation.

Canadian banks were large, stable and sophisticated national entities (an oligopoly). With branches across the country and often in other countries, Canadian bank risk was dispersed. Canadian bankers tended to be more risk- averse than their U.S. and international counterparts. Canadian banks were required to maintain a tier one capital ratio of seven per cent and generally exceeded it. They had to cap overall leverage at 20× capital.

iThis section is from “You Can Take It to the Bank,” Ivey inTouch Magazine, Fall 2009, p. 14.

Canadian banks were regulated by a single piece of legislation, the Bank Act, which was reviewed every five years, and one national body, the Office of the Superintendant of Financial Institutions (OSFI). OSFI had broad oversight—there was no “shadow banking system” that fell outside the regulations. Canada also had strong monetary policy set by the Bank of Canada and the Department of Finance.

In Canada, most investment banks were owned by commercial banks, providing them with access to capital during a crisis.

The U.S. Financial Sector

Despite the close geographic proximity, the nature of the United States banking and mortgage industry differed significantly from the system that prevailed in Canada.

Decades ago, the U.S. government launched two agencies to promote home ownership in the United States—the Federal National Mortgage Association (“Fannie Mae”) and the Federal Home Loan Mortgage Corporation (“Freddie Mac”). These agencies were designed to increase the availability of funds for originating mortgages and to encourage the emergence of a secondary market for mortgages. Subsequently, mortgages could be traded without the involvement of either the original borrower or the original lender.

In the 1990s, to further encourage home ownership in the United States, policymakers lowered the amount of equity that homebuyers were required to invest in the purchase of a home. As a consequence of this policy shift, borrowers who were previously unable to secure a mortgage were able to enter the housing market. Further, the overall degree of leverage in the U.S. housing market increased substantially and a housing bubble emerged as homeowners began to speculate by moving into more expensive homes.

The coincident emergence of three financial innovations in the United States—interest-only mortgages, asset securitizations and credit default swaps—ultimately set the stage for the perfect storm that had converged over the U.S. financial system by 2007.

Unlike self-amortizing mortgages in which the mortgage principal was retired through regular payments of principal and interest over the life of a mortgage, interest-only mortgages were mortgages in which the borrower was given the opportunity to pay only the interest portion of a regularly scheduled mortgage payment. Interest-only mortgages were designed to open home ownership to low-income earners who demonstrated enhanced future earning potential, at which point their mortgage would be converted into a self-amortizing mortgage. Interest-only mortgages benefitted these low-income homeowners by facilitating their entry into the housing market though payments which were lower than the payment under a self-amortizing mortgage.However, the emergence of interest-only mortgages also contributed to speculation in the housing market, as some investors purchased homes, made the interest payments while waiting for the value of their homes to increase and then sold the homes, paying back the mortgage principal with the proceeds from the home sale and pocketing the surplus.

Asset securitization involved aggregating a series of future cash flows into a security which was then sold to investors. Mortgage-backed securities (MBSs) were a type of asset securitization in which the underlying asset backing the security was a mortgage which generated cash flows from the interest payments. A securitization was a structured finance product that was originally designed to distribute risk. In fact, when conceived, MBSs were regarded as low-risk investments because they were backed by mortgages and mortgage defaults were relatively rare occurrences.

Credit default swaps (CDSs) RESEMBLES INSURANCE POLICIES in the sense that one party paid a series of cash flows to a counter-party in exchange for the promise that the counter-party would reimburse the payer if the underlying asset defaulted. A significant portion of the market for CDSs was built around MBSs. Investors in asset-backed securities such as MBSs regularly insured their investments by purchasing CDSs. The premium revenue stream associated with a CDS on an MBS was considered particularly attractive due to the low level of perceived risk, again due to the relatively rare occurrence of mortgage defaults. Despite their resemblance to insurance policies, CDSs were traded as contracts in the derivatives markets and were free from insurance industry regulations. Consequently, the relative ease with which CDSs could be issued, coupled with the fact that it was not necessary to own the underlying asset in order to purchase a CDS, effectiv

Two phenomena associated with these three financial innovations further compromised the precarious foundation upon which the U.S. banking and mortgage industry was perched—subprime mortgages and individual compensation systems prevailing in the financial sector. While mortgages issued to creditworthy borrowers were known as prime mortgages, subprime mortgages were issued to borrowers with poor credit. MBSs based on subprime mortgages became particularly attractive investment vehicles due to their high returns and low levels of perceived risk (due to the assumption that widespread mortgage defaults were highly unlikely). At the same time, mortgage originators

and derivative traders were being compensated on the volume of mortgages originated and derivatives sold (MBSs and CDSs). Increased trading volumes in these assets were fueled by the fact that compensation was rarely adjusted to the riskiness of either the borrower or the underlying asset.

By 2007, the robust growth in U.S. home prices slowed dramatically. As home prices began to decline, the value of mortgages began to exceed the market value of many homes. A flood of mortgage defaults ensued to the point that mortgage-backed securities began to decline in value. The complex nature of these securities further undermined their value. Given that it was not possible to link an MBS to specific properties, investors could not evaluate the risk of default on specific MBSs and, therefore, were unable to ascertain market values for these MBSs. The secondary market for mortgages was near collapse.

The difficulty associated with valuing these securities proved to be particularly problematic for financial institutions that owned the devalued MBSs and for financial institutions facing insurance-like claims on the CDSs they had written on the bet that widespread mortgage defaults would never occur. Consequently, these financial institutions were required to raise more capital to shore up their capital ratios. However, the increasing pervasiveness of uncertainty effectively turned off the taps in both credit and capital markets, making the task of raising capital almost impossible. As the cost of capital skyrocketed and credit stopped flowing in the United States, financial institutions began to fail. Several “runs on the bank” were triggered in which customers lined up to fully withdraw their deposits. In June 2008, panicked customers of IndyMac Bank in the United States withdrew $1.5 billion in deposits (approximately 7.5 per cent of total bank deposits). Similarly, over the course of ten days in September 2008, customers withdrew more than $16 billion from Washington Mutual Bank (totaling nine per cent of total bank deposits). The uncertainty spilled over U.S. borders, triggering bank runs and failures overseas as well Most notable was the bank run and subsequent failure of the U.K.–based Northern Rock bank, which was subsequently nationalized, in part, to subdue the panic.

Conclusion

The Challenge of Refining the Future Direction of the Joint Venture As early as the summer of 2007, credit spreads for certain financial companies and instruments widened dramatically. In Canada, the market place for ABCP began to show signs of stress. The JV’s ALCO Committee ON RECOMMENDATION OF CIBE Mellon’s risk management group and leveraging the respective credit market specialties of both shareholders, directed the JV to refrain from using any of CIBC Mellon’s treasury or client funds (the latter in the form of cash collateral for securities lending transactions) to purchase non-bank-owned ABCP. Eventually, in August 2007, the $30 billion market for non-bank-owned ABCP essentially froze. The looming financial crisis did not portend a quick or strong recovery in the credit markets, particularly in the ABCP market.

MacMillan and his team suspected that the future growth potential for the stock transfer and corporate trust business segments was more limited than it was for the asset servicing business. Notwithstanding the onset of the financial crisis, MacMillan debated whether the JV should retain or divest these business lines in order to focus more intensely on the asset servicing business for which the JV was formed.

Challenge of Avoiding Major Operational Missteps The brewing financial storm became fodder for the media and it started to rattle financial markets. Despite the fact that there were reasons to believe that the impending financial crisis might not be as bad in Canada as it was likely to be in the United States and elsewhere, numerous challenges remained. Not least of these was the fact that the crisis would likely bring out the worst in many long-term business relationships. As liquidity was tightening, many financial sector lenders, borrowers and partners alike were putting aside years, even decades, of cooperation in order to ensure their own survival. This was in contrast to the approach adopted by CIBC Mellon in the months leading up to and during the crisis. It retained its long-standing practice of emphasizing very extensive communication with its clients and its shareholders, ensuring shared understanding of issues, including having representatives from both parents on ALCO and maintaining transparency.

By mid-September 20028, the most significant risks facing CIBC Mellon were credit risk, operational risks and market risk, as well as the potential exodus of assets and clients who were panicked by the wider financial chaos. MacMillan and his team recognized that it would be critical for the JV to continue to deliver a high level of client service and to avoid any major operational missteps. A key challenge facing the JV pertained to efforts to remain loyal to both long-time and newer business clients, while not exposing the JV to excessive risk in the context of an increasingly volatile market. While CIBC Mellon’s global securities lending operations had extended considerable credit to some of the now more precariously perched financial giants, the JV was comfortable that these loans were adequately collateralized. Nevertheless, in order to ensure that the loaned assets were not subsumed into any debtors’ possible bankruptcy proceedings, the JV would need to execute against legal agreements with rigour, to preserve its legal rights, including, if necessary, taking possession of collateral assets and then liquidating these assets in an increasingly turbulent market. Critical decisions were faced by MacMillan, Marchand, Hemingway and Slater, ranging from short-run decisions such as how to determine when to call in credit extended to some of the JV’s global securities lending clients and how to liquidate any collateral that the JV was forced to take into possession, to longer-run decisions surrounding how to manage the JV’s relationships with its solvent clients, so as to stem any risk of client or asset flight.

MacMillan closed the door to his office. Notwith- standing the 110 years of collective experience between MacMillan, Marchand, Hemingway and Slater, the four men recognized that the markets were headed for uncharted waters. MacMillan opened the meeting, reminding the group, “Gentlemen, now more than ever, we need to leverage our JV’s administrative heritage, the guidance of our shareholders and the respective strengths of the parents to move through these unprecedented times.