In 2007, more than a year before the near-collapse of America’s mighty capital markets triggered the worst global economic crisis since the Great Depression, Canada experienced a stunning financial meltdown of its own.
This episode is largely forgotten. Although it traced its roots to the same subprime cataclysm, you would have a hard time finding it mentioned even as a footnote in the multitude of accounts written about the Great Financial Crisis.
It would be fair to describe it as “tangential” to the larger catastrophe, which culminated in the bankruptcy of Lehman Brothers. And yet, calling it “tangential” belies the fact that real people were impacted – and in many cases, dramatically so. As The Globe And Mail put it, in a contemporaneous account published more than a dozen years ago, “at stake were the savings of a far-flung collection of investors stretching from the Yukon government to credit union branches in Chicoutimi.”
In 2012, I met one of the people whose life was forever changed by the episode. His name is Eric Ben-Artzi.
Eric holds a PhD in mathematics from NYU, is an expert in quantitative financial modeling and worked for major Wall Street firms, including Goldman Sachs, in roles ranging from trader and structurer to vice president of risk. In his role as a risk manager at Deutsche Bank, Eric protected the investing public by exposing accounting irregularities tied to how the bank valued complex illiquid positions related to a market-wide restructuring agreement aimed at resolving the crisis in Canada’s asset-backed commercial paper market.
In 2013, Eric and I wrote a book about Canada’s forgotten crisis. On countless occasions since it was (mostly) completed in 2014, we have debated the best route to go in publishing it. That debate is ongoing.
For now, Eric and I have decided to publish select chapters in these pages.
It is important to note that while he certainly has an important story to tell, the book is not about Eric. And it is not about Deutsche Bank, either. Rather, this is the story of the episode itself. It is a tale of hubris and the perils of financial engineering run amok. There are many such stories, and yet, no matter how many we hear, we never seem to learn the right lessons. If we did, there would not be so many such tales to tell.
Selected chapters will be published here over the next several weeks and months, in no particular order, although my goal is to start with chapters that will make for the most compelling reading. Please note: I am making no initial effort to provide a glossary of terms or to otherwise go out of my way to provide extensive background information. Over the next few weeks, as more chapters are published, I will create a dedicated section of the site aimed at facilitating a more thorough understanding for the uninitiated. In the interim, I am happy to field questions from readers via e-mail.
Those interested in obtaining the full manuscript, or anyone who wishes to inquire further about the work Eric and I have done, can contact me directly at: firstname.lastname@example.org
“By being open and honest about risks, the leaders of Coventree hope to create a culture that learns from minor missteps to prevent major errors.”
— Coventree Capital, 2006 Annual Information Form
Coventree was founded in 1998 by Geoffrey Cornish, David Ellins, and Dean Tai. Cornish, a securities lawyer by trade who practiced at Canadian business law firm Davies Ward & Beck with Ellins from 1985 until the founding of Coventree, served as the company’s president, director, and head of capital markets. Following Coventree’s initial public offering, Cornish controlled around 27% of the company. Dean Tai who, by some accounts was something of a foosball aficionado, served previously as COO at network computing services firm Align Service Delivery Corporation and prior to that was the Treasurer of the leasing division at IBM Canada. After the IPO, Tai served as Coventree’s CEO and, like Cornish, controlled around 27% of the company’s outstanding shares.
Coventree issued its first asset-backed commercial paper in 2000 via a conduit called “Rocket.” Over the years, the company would add eight more conduits to its stable: Apollo Trust, Aurora Trust, Comet Trust, Gemini Trust, Planet Trust, Slate Trust and Venus Trust. In addition to this cosmos-themed bunch, Coventree also operated a wholly-owned subsidiary called Nereus, which itself administered two trusts with far more terrestrial-sounding names: Structured Investment Trust III and Structured Asset Trust. Nereus specialized in credit arbitrage and, in the words of one bank securitizer, routinely took “strategic risks to create clients” for Coventree.
Indeed, Coventree was known for facilitating the creation of “innovative” debt products. According to one former employee, new ideas for asset classes were embraced with open arms and creativity was an important part of the firm’s success. This corporate culture helped Coventree become a pioneer in the Canadian market for credit arbitrage-backed commercial paper. During the three quarters leading up to the market collapse in August of 2007, fully 80% of Coventree’s revenues were derived from structured financial asset transactions. The company, which made just $34,000 in 2001, pulled in $24.4 million in profits in 2005 (the year before it went public) and boasted 56 employees by 2006.
To be sure, the company had big plans. Its “corporate objective,” as outlined in its 2006 annual report, was “to be a financial services company focused on targeted niches includ[ing] structured financing using securitization technology, conduit administration outsourcing services for third parties, asset management and private equity investments in synergistic enterprises, as well as the eventual entry into certain retail banking activities.”
By the time those words were written, the company had already filed an application with the Superintendent of Financial Institutions to incorporate a Schedule I bank in Canada. On top of its retail banking aspirations, Coventree also hoped to gain a foothold in the U.S. market and to start an “asset management business unit to manage fixed income products in addition to asset-backed commercial paper.”
Despite all of this, it wasn’t exponential growth, impassioned creativity, or lofty ambitions that inspired Coventree’s decision to go public. The Caisse de Depot et Placement du Quebec (CDPQ), the largest pension-fund manager in Canada, owned a large stake in Coventree prior to its public offering. At its discretion, CDPQ could force Coventree to buy back all of those shares via the exercise of a put option held by the pension-fund manager. That privilege would only cease to exist in the event Coventree became a public company.
On February 13, 2006, Dean Tai informed Coventree’s board of directors that he expected CDPQ to exercise that option within two years, an event that would have threatened the firm’s very existence. In order to avoid this, Coventree decided to take the company public.
Ultimately, the initial public offering was completed in November of 2006, and a secondary offering allowed CDPQ to sell 3.156 million of the shares it held, reducing the pension-fund manager’s stake to 10%. Incidentally, Coventree’s prospectus also reveals CDPQ pocketed around $1.25 million in dividends on its Coventree stake between September 29 and October 6. Tai and Cornish sold only 150,000 shares each in the IPO, and as such, the two controlled 54% of the company between them following the offering. CDPQ got a representative on Coventree’s board in Francois Maheu. Coventree was the only publicly traded third-party issuer of asset-backed commercial paper.
CDPQ, which invested on behalf of public sector pension plans in Quebec, was to the investor side of the equation what Coventree was to the issuer side. That is, CDPQ was the largest investor in third-party asset-backed commercial paper and Coventree was the largest issuer of such debt. By the time it went public in 2006, Coventree was the third-largest asset-backed commercial paper issuer of any kind in Canada. In other words, the company was big not only among third-party (i.e., non-bank) sponsors, but was in fact big in an absolute sense, ranking behind only Bank of Montreal and CIBC in terms of asset-backed commercial paper origination.
To be sure, CDPQ played a large part in Coventree’s success. In the company’s prospectus, under “risk factors,” Coventree noted that “a substantial minority percentage of the securities issued by [our] conduits [have] been purchased by one investor, CDPQ.”
In fact, during 2007, CDPQ held between 33 and 50% of all commercial paper issued by Coventree-sponsored trusts. This dependence on one buyer would come back to haunt Coventree when the market began to seize in July.
Less than two weeks after ratings agency DBRS issued a January 2007 press release effectively announcing the implementation of new rules that would cripple three quarters of Coventree’s business by making it virtually impossible to obtain a top notch credit rating for a CDO-backed third-party commercial paper deal, Dean Tai sent an e-mail to colleagues regarding the effect DBRS’s move would likely have.“I believe the situation we find ourselves in is analogous to a tsunami coming,” Tai said in the letter, dated February 1, 2007.
DBRS’s crackdown on structured finance transactions and Tai’s reaction to the ratings agency’s announcement both cast doubt upon the conventional wisdom surrounding the collapse of the market for Canadian third-party asset-backed commercial paper. The general consensus is that because most of the commercial paper issued by non-bank Canadian sponsors had a relatively low amount of exposure to U.S. subprime, previously esoteric concepts like the difference between global-style and Canadian-style liquidity provisions would have remained in relative obscurity and the market would have stayed afloat were it not for the rapid deterioration in the U.S. subprime mortgage market. A related argument says that were it not for the opaque nature of the structured finance deals (CDOs) that backed the commercial paper, Canadian investors might have known that only a small percentage of the assets backing the transactions were related to subprime.
While it is certainly true that jitters regarding the impending collapse of the U.S. subprime mortgage market contributed mightily to the collapse of the market for third-party asset-backed commercial paper in Canada, DBRS’s actions in January of 2007 (and in November of 2006) demonstrate that there were questions as to the viability of CDO-backed commercial paper that pre-dated the climax of the subprime panic. In short, the greatest risk to investors may have been the complexity of the deals that backed the paper and the inability of ratings agencies to model the risk associated with those deals properly.
The market for asset-backed commercial paper in Canada expanded by some 50% from 2004 to 2006. This extraordinary growth was largely attributable to the increasing popularity of structured finance-backed transactions. In other words, the issuance of commercial paper backed by ‘traditional’ securitizations of assets such as credit card receivables, home mortgages, and auto loans, was out of style. The trend was towards commercial paper backed by collateralized debt obligations or, CDOs.
Indeed, by the fall of 2006, over a quarter of all Canadian asset-backed commercial paper was backed by structured finance deals, a remarkable shift over the course of just five years (Figures 1 & 2).
Figure 1: ABCP Composition (Canada, 2001), Figure 2: ABCP Composition (Canada, 2006)
(Source: Mark Adams, “Synthetic Securitization and CDOs,” presented at Osgoode Hall Law School)
To demonstrate the important role the deals played in fueling the industry’s growth, consider that Nereus (the Coventree subsidiary which dealt almost exclusively in structured finance) accounted for $1.2 billion of the total $4.2 billion of growth in commercial paper outstanding at Coventree from 2005 to 2006. So, 25% of Coventree’s growth in terms of total commercial paper outstanding was coming directly from what amounted to its structured finance arm.
On November 10, 2006, Huston Loke (at the time managing director at DBRS in charge of structured finance) sent an e-mail to the major players in the market for Canadian asset-backed commercial paper. Loke’s tone in the letter was cautious.
After briefly noting that the majority of the growth in the asset-backed commercial paper market was being fueled by the concurrent increase in the popularity of deals backed by structured finance or CDO-related transactions, Loke said he was concerned about the “near exclusive” use of leveraged super senior (LSS) technology in the new deals. This trend, Loke said, was having the effect of gradually embedding “contingent funding obligations in an increasingly larger portion of the market.” Loke also warned that the deals had the potential to “incorporate economic and risk characteristics that [were] quite different from previously structured transactions.”
Ultimately, Loke made it clear that from that point forward, DBRS would be taking a far more cautious approach to rating LSS and other CDO-backed deals: “Consistent with DBRS’ new approach to SFA transactions, market participants should… expect that the types of SFA transactions that will be approved going forward will be more restrictive than those approved in the past. This mandate will be effective immediately.”
Tai responded to DBRS’s letter in the only way he knew how: “Good note,” he said, in an e-mail response to Huston Loke.
But this was just for show. To be sure, there was nothing “good” about the situation for Coventree. In stark contrast to the response he sent to Loke, Tai expressed the utmost concern in an e-mail addressed to Nereus’s board dated November 17, 2006. “Given DBRS’ more recent letter regarding leveraged super senior transactions, and CDO [transactions] more generally, we are very uncertain about the long-term prospects of Nereus as a going concern,” Tai warned.
A week later, Tai’s frustration with Nereus was becoming immediately apparent: “Given that DBRS has now explicitly declared that they will be restricting [LSS transactions] to the point where they may be uneconomical, we would like to know why the Board has NOT insisted on Nereus diversifying away from these transactions or this business model in an aggressive fashion, despite warning at the inception and repeated warning by Coventree that this would occur.”
By contrast, Geoffrey Cornish did not believe the situation to be so dire. For him, DBRS’s November letter was just history repeating itself. The ratings agency was known for making bold statements and then retracting them once market participants voiced their opinions.
Cornish maintained this stance even after receiving a worried message on November 16 from Ken Toten, a Coventree executive in charge of the company’s structured finance unit. In an email to Cornish, Toten (a commercial mortgage-backed securities expert and former senior credit officer at Citibank) warned that Coventree would “not be able to achieve [its] funding goals for fiscal year 2007 based upon the user-unfriendly posture adopted by [DBRS]”.
Cornish would later testify that he believed Toten’s comments were simply an attempt to negotiate a higher salary. Nonetheless, Cornish quickly sent an e-mail to Tai : “We are going to have to get another ratings agency,” Cornish said.
On July 11, 2007, Cornish sent an e-mail to Tai, Judie Dalton, and David Allan who, earlier in the year, had taken over for Cornish as Coventree’s head of capital markets.
Cornish had just left an S&P conference in which the ratings agency outlined its decision to downgrade billions in subprime debt. Although S&P’s actions did not affect any of Coventree’s paper, Cornish learned that investors were demanding information about the extent of issuers’ subprime exposure. He told his colleagues that Coventree should consider disclosing information about its conduits’ subprime exposure now, rather than wait for the proverbial barbarians to reach the gate: “[We should prepare] information that Coventree could either issue as a press release or give as an info sheet to our commercial paper dealers on where we are at – clearly investors are going to be asking a lot of detailed questions on what we have in our conduits in this asset class in the near future.”
Dalton, head of the company’s funding group, told Cornish that the company had already been asked about its subprime exposure and noted that “as a public company, [Coventree] may not have the luxury to go anywhere but” a press release. Whatever we do, Dalton said, “we need to send this out tomorrow at the latest.”
Later that day, Doug Paul, a member of Coventree’s funding group who reported to Dalton, sent an e-mail to National Bank Financial (the head of Coventree’s dealer syndicate) outlining the firm’s exposure to the subprime mortgage market, broken down by conduit and note series within each conduit.
On July 20, Scotia Capital, the lead dealer for paper issued by Coventree’s Nereus subsidiary, e-mailed Dalton and encouraged her to publicly disclose the extent of Coventree’s subprime exposure in order to head off rumors about the health of the third-party asset backed commercial paper market. Four days later, on July 24, Dalton sent the subprime breakdown to every bank in the dealer syndicate (Figure 3).
“We have given this to everyone in our Dealer group so they can use their own judgement to send it to clients with the sophistication to use the information (we figure you have the best read on this),” Dalton said. “Otherwise, the Dealers are armed with enough information to provide timely and accurate replies when asked.”
Figure 3: Coventree Subprime Exposure by Conduit and Note Series
That same day, a large investor sold $100 million of third-party asset-backed commercial paper back to CIBC (another member of Coventree’s dealer syndicate) prior to maturity. This indicated the market was getting nervous about non-bank sponsored notes and set-off alarm bells for CIBC. The bank’s money market desk was instructed to cease purchasing Coventree-sponsored notes until the risk management department had time to conduct a full review of commercial paper-related credit limits. Two days later, when the review was complete, CIBC lowered the credit limits associated with third-party asset-backed commercial paper. Over the course of the next four days, CIBC elected not to roll its maturing Coventree-sponsored notes.
Despite this, CIBC continued to place the paper for Coventree. From July 25 through August 3, CIBC sold some $245 million of third-party asset-backed commercial paper to investors even as the bank wasn’t rolling its own paper. It seems clear that those purchasing the notes from CIBC during late July and early August were not made aware that the market was effectively shutting down. It also appears that CIBC benefited directly from its informational advantage — $22 million of the paper it sold from July 25 to August 3 came from its own inventory.
On July 25, Dalton got a call from Luc Verville, the man in charge of the CDPQ’s asset-backed commercial paper purchases. Verville wanted to know more about Coventree’s subprime exposure. That same day, Verville told Allan that CDPQ had sold some Coventree-sponsored paper and, instead of rolling into more of the company’s notes, had bought bank-sponsored paper instead. Given Coventree’s dependence on CDPQ, this must have felt to Allan like being told your significant other has been unfaithful.
As it turned out, Verville was seeking to cut the CDPQ’s asset-backed commercial paper exposure by $1 billion. “Well, If you have to lighten up, lighten up from those that have more subprime exposure,” Verville told Allan. Allan suggested that Coventree might be willing to work on a rebalancing plan under which the firm would scatter the subprime exposure that was concentrated in a few specific notes over all note series in all conduits. Verville listened politely, but never expressed much interest in the scheme following the phone call.
That day, Verville also called Huston Loke at DBRS and asked how Coventree’s paper with subprime exposure was performing in the market. In fact, none of Coventree’s paper was performing well.
In the past, Coventree’s dealers generally completed the sale of the debt by 9:00 a.m. each morning. On July 25 however, Scotia Capital sent some of its allotment of commercial paper back to National Bank unsold at 9:05. Some four months earlier, Scotia became aware that some of the paper issued by Coventree-sponsored conduits was exposed to the U.S. subprime mortgage market. Subsequently, the bank’s management directed its staff to begin reducing its inventory of asset-backed commercial paper over the course of the summer. Despite the fact that this directive clearly illustrated the firm’s misgivings about the market, Scotia continued to place the paper for Coventree through August 10. In fact, one secondary dealer who bought the paper from Scotia and marketed it to retail clients would later claim that while Scotia was paring-down its own inventory of non-bank commercial paper, the firm was still “actively and aggressively marketing [the paper] by means of frequent or daily written and oral solicitations and communications.”
Meanwhile, RBC Dominion Securities’ head of Canadian fixed income Peter Dymott told a colleague that if CDPQ was getting nervous, he wanted out of Coventree’s dealer syndicate. “Caisse isn’t rolling all of their maturities,” Dymott said. “The model is dead for Coventree, I want out.”
This was a potential landmine. RBC was the largest (and probably the most influential) commercial paper dealer in Canada. If they abandoned ship it could, in Geoffrey Cornish’s words, “spook the market further at a critical time.”
In fact, RBC hadn’t sold any Coventree paper with subprime exposure in at least a week. The bank claimed it was being forced to hold onto the notes because no one wanted them and it simply could no longer accommodate the paper in its inventory. On July 27, unnerved by Coventree’s suggestion that disclosures about its exposure to subprime should be distributed only to “clients with the sophistication to use the information,” RBC resigned from Coventree’s dealer syndicate.
Each day, dealers took longer and longer to report back as they tried in vain to sell their allotment of Coventree-sponsored paper. On July 26, the full allotment remained unsold as of 10:19. Exasperated, National Bank’s Bob Courchesne told Coventree’s Doug Paul that National was reaching its limit in terms of how much unsold Coventree paper it could place in inventory.
“I can’t buy all the Planet A on earth here,” Courchesne told Paul, referring to series A paper issued by Coventree’s Planet trust, which had some 29% of subprime exposure. “You don’t seem to have a dealer group.”
The following Monday (July 30), Coventree began to notice that spreads for its paper were creeping higher. Typically, Coventree’s A notes paid between 1 and 4 basis points above CDOR. By August 1, that spread had widened to nearly 13 basis points. On August 2, Allan and Dalton held a conference call with CDPQ’s Verville. Allan came away worried. “The only answer seems to be to convince him that our subprime exposure is not a risk concern,” Allan told Cornish. “We have to get that information to him FAST.”
Less than 24 hours later, at 1:00 p.m. on Friday, August 3, Bob Courchesne from National Bank informed Dalton that RBC and Toronto-Dominion Bank had just sponsored a new deal. Not only was this a bank-sponsored deal, it was also a traditional securitization and it was in the amount of $3 billion. For Coventree, it was devastating. Recent market activity certainly indicated that supply exceeded demand for non-bank asset-backed paper and with the new RBC deal coming to market, it was a virtual certainty that no one would be buying Coventree’s paper.
“There already was an imbalance between supply and demand before this additional $3 billion came into the marketplace,” Cornish said. “This simply exacerbates the problem.” Allan concurred: “[That] might well have been liquidity that otherwise would have gone to rollovers in the non-bank sector.”
Three hours passed. Just after 4 p.m., Doug Paul reminded Cornish, Allan, and Tai that on Tuesday, $625 million in Coventree-sponsored paper was coming due, including well over $100 million in the problem Planet A notes. Dalton spoke with Natalie Davidson at CIBC shortly thereafter. Davidson said that none of the large dealers could afford to inventory any more of Coventree’s unsold paper. Davidson also told Dalton that in all likelihood, Coventree would default the following week.
“From my understanding, the street cannot inventory any more of your paper including National due to trading lines,” Davidson warned. “Which means that you will have to redeem unfunded commercial paper or the trust(s) may be put into default. This is why I’m trying to reach David [Allan] as well – I’m not sure if your lead dealer has conveyed how grave the situation now is.”
Allan called Davidson on Saturday and in a testy exchange, tried to make it clear that he was not in denial about the seriousness of the situation. “Natalie, don’t worry – I am well aware of the gravity of Tuesday’s situation. We will be prepared for the worst,” Allan said, before making it clear that he expected some defaults on Coventree’s paper.
By Sunday, Davidson and CIBC had heard enough. On a call with Allan she said the bank was ready to follow in RBC’s footsteps and resign as a member of Coventree’s dealer syndicate. Ultimately, the bank stopped selling Coventree-sponsored notes as of Tuesday, August 7.
That Saturday was the first time senior management at CIBC was made aware of just how urgent the situation had become. This belated escalation of the issue led directly to what the Ontario Securities Commission would later call “a delay in engaging appropriate processes and in assessing the impact” of the sea changes taking place in the market. This failure to inform the appropriate personnel ended up costing the bank $21.7 million in fines.
By Monday morning, Judie Dalton knew what she had to do. Dalton called Verville at CDPQ in a last ditch effort to convince Coventree’s closest ally to step in and help the company place its paper the next day. That afternoon at 3:15, Verville placed a call to Davidson at CIBC and reassured her that there would be no market disruption on Tuesday. An hour later, Allan followed up on Verville’s call with Davidson. “Natalie, Luc is stepping back in to hold this market up,” he promised.
That evening, Tai informed the board that CDPQ was coming to the rescue. On Tuesday, Verville kept the promise – CDPQ rolled its Coventree paper for 90 days and bought additional notes as well. The support however, came with a price. Spreads on Coventree A notes widened to 25 basis points above CDOR reflecting, in Cornish’s words, “the price that CDPQ… wanted to extract… so to speak, for [its] continued support of the marketplace.”
The relief was short-lived. Shortly after 9 a.m. on Thursday morning, National Bank told Coventree that things were “going very bad.” The market was on high alert after BNP Paribas’ announcement that it was suspending redemptions in three funds linked to asset-backed securities. Deutsche Bank, CIBC, and HSBC all failed to sell their allotment of Coventree-sponsored paper. “Thank God the Caisse is still there,” National Bank said. “I don’t know what would have happened” without them.
Coventree rolled its paper on Friday with the help of CDPQ, but the game was up. Verville held a meeting that day with all of the major players in the Canadian asset-backed commercial paper market. Over the weekend, CDPQ attempted to convince the Bank of Canada to accept third-party commercial paper as collateral for cash loans — theoretically, this would boost demand for the notes as banks would be more willing to hold the paper if they knew they could borrow central bank cash against it. Ultimately, the effort was unsuccessful. The central bank quite logically observed that there simply was no way to assign concrete values to the individual issues. On top of that, the Bank of Canada did not wish to be seen as propping up certain market sectors in times of distress.
On Monday, August 13, the bottom finally fell out. Coventree failed to roll all of its paper and its shares were halted on the Toronto Stock Exchange.
By August 22 – 9 days after the meltdown –the spread between Coventree’s Series A notes and CDOR had risen to an unprecedented 50 basis points. At that juncture, Coventree’s business model was completely wiped out. The company derived 80% of its revenues from credit arbitrage-backed transactions. This meant that it was dependent for its very survival on the spread between the rate it earned from its structured finance deals and the interest rate it paid to investors who bought the commercial paper backed by those deals.
With rates on its paper at 50 basis points (a half percentage point) above CDOR, there was nothing left for Coventree in terms of revenue – the spread between what it earned on its deals and what it paid out to investors had vanished.
“At these spread levels, Coventree’s revenues from credit arbitrage transactions will in the short-term be reduced to zero,” the company said in a press release.
In the interim at least, it was game over for Cornish, Tai, Dalton, and Allan.
Sources for Chapter 3
 Ontario Securities Commission (2011). Reasons for Decision: In the Matter of Coventree Inc. et al. Retrieved from: https://www.osc.gov.on.ca/en/Proceedings_rad_20110928_coventree.htm
 Erman, B., McNish, J., Perkins, T., & Scoffield, H. (2018, April 26). The ABCP black box explodes. Retrieved from: https://www.theglobeandmail.com/report-on-business/the-abcp-black-box-explodes/article22625269/
 Coventree. (2006). Annual Report. Retrieved from: http://globaldocuments.morningstar.com/documentlibrary/document/26bee4750d96a67a.msdoc/original